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/7-1906/v1
|
07 Dec 18
|
{
"type": "Research Article",
"title": "Hub genes in a pan-cancer co-expression network show potential for predicting drug responses",
"authors": [
"Francisco Azuaje",
"Tony Kaoma",
"Céline Jeanty",
"Petr V. Nazarov",
"Arnaud Muller",
"Sang-Yoon Kim",
"Gunnar Dittmar",
"Anna Golebiewska",
"Simone P. Niclou",
"Tony Kaoma",
"Céline Jeanty",
"Petr V. Nazarov",
"Arnaud Muller",
"Sang-Yoon Kim",
"Gunnar Dittmar",
"Anna Golebiewska",
"Simone P. Niclou"
],
"abstract": "Background: The topological analysis of networks extracted from different types of “omics” data is a useful strategy for characterizing biologically meaningful properties of the complex systems underlying these networks. In particular, the biological significance of highly connected genes in diverse molecular networks has been previously determined using data from several model organisms and phenotypes. Despite such insights, the predictive potential of candidate hubs in gene co-expression networks in the specific context of cancer-related drug experiments remains to be deeply investigated. The examination of such associations may offer opportunities for the accurate prediction of anticancer drug responses. Methods: Here, we address this problem by: a) analyzing a co-expression network obtained from thousands of cancer cell lines, b) detecting significant network hubs, and c) assessing their capacity to predict drug sensitivity using data from thousands of drug experiments. We investigated the prediction capability of those genes using a multiple linear regression model, independent datasets, comparisons with other models and our own in vitro experiments. Results: These analyses led to the identification of 47 hub genes, which are implicated in a diverse range of cancer-relevant processes and pathways. Overall, encouraging agreements between predicted and observed drug sensitivities were observed in public datasets, as well as in our in vitro validations for four glioblastoma cell lines and four drugs. To facilitate further research, we share our hub-based drug sensitivity prediction model as an online tool. Conclusions: Our research shows that co-expression network hubs are biologically interesting and exhibit potential for predicting drug responses in vitro. These findings motivate further investigations about the relevance and application of our unbiased discovery approach in pre-clinical, translationally-oriented research.",
"keywords": [
"co-expression networks",
"network hubs",
"drug sensitivity prediction",
"anticancer drugs",
"transational bioinformatics",
"systems biomedicine",
"biological networks"
],
"content": "Introduction\n\nThe analysis of networks extracted from different types of “omics” data is a useful strategy to enable the characterization and prediction of meaningful properties of the underlying complex biological systems1–3. Measures of the centrality of genes or proteins in such networks have been shown to be indicators of biological function4–7. Specifically, the biological significance of highly connected genes, i.e., hubs, in different molecular association networks has been determined using data from several model organisms, molecular interaction types, phenotypes and pre-clinical research applications5,8–10. Other research, however, has shown that hub genes in (patient-derived) gene co-expression networks may not have sufficient prognostic value in a few selected classes of cancer11. Despite such insights, the predictive potential of candidate hubs in gene co-expression networks in the specific context of cancer-related drug experiments remains to be thoroughly investigated. An examination of such associations may offer novel opportunities for the accurate prediction and understanding of anticancer drug responses.\n\nAddressing the above-mentioned challenge is now possible thanks to the availability of large collections of data originating from thousands of drug experiments in cancer cell lines. Over the past few years, the investigation of cell line-based computational models for anti-cancer drug sensitivity prediction has been accelerated by publicly-funded efforts of large research consortia. In particular, the Cancer Cell Line Encyclopedia (CCLE)12 and the Genomics of Drug Sensitivity in Cancer (GDSC)13,14 projects represented significant steps forward for the oncology and pharmacogenomics research communities. These projects have generated genomic and transcriptomic data from thousands of (untreated) cancer cell lines and their accompanying treatment sensitivity measurements for hundreds of experimental and clinically-approved drugs. Using these datasets, computational models for predicting anticancer drug sensitivity based on the analysis of transcriptomic and other types of “omics” data have shown to be useful in the selection and prioritization of candidate compounds for pre-clinical research15–18.\n\nHere, we investigate the relationship between significant co-expression network hubs and drug responses (Figure 1). We identified 47 genes representing “hubs” in a pan-cancer transcriptomic network extracted from more than 1000 (untreated) cell lines. These hubs are substantially implicated in a diversity of cancer-related biological processes, and their individual expressions (in the untreated cell lines) are correlated with drug sensitivity. Next, we validated such findings using an independent dataset that also comprises thousands of cell line-drug experiments. We observed that a relatively simple model, based on multiple linear regression, can make predictions that are concordant with the actual drug sensitivity values observed in vitro. Moreover, although we do not claim that our model clearly outperforms more complex techniques, its prediction performance is comparable to, and in some cases improves on, previously published models. This is particularly interesting because, unlike prior work, we followed an unbiased discovery approach, i.e.: we did not seek, up-front, a specific set of genes to optimize such a prediction task.\n\nMotivated by these findings, we used our 47 hub-based model to predict sensitivity scores for four glioblastoma (GBM) cell lines, including three (stem-like) cell lines that were not included in the discovery and validation datasets, against 24 drugs. We selected the top three drugs predicted as highly effective together with a drug predicted as lowly effective (negative control), and performed in vitro tests on the 4 cell lines. The sensitivity scores predicted by the hub genes tend to be concordant with the observed in vitro responses. Lastly, to facilitate future research, we offer a Web-based interface that allows users to predict drug sensitivity scores for their own samples and expression data with our 47-hubs-based model.\n\n\nMethods\n\nThe published pre-processed CCLE (microarray) gene expression and drug sensitivity datasets were obtained from the CCLE website. In the gene expression dataset, we focused on genes with symbols, calculated their standard deviation (SD) across all samples (1037 untreated cell lines) and ranked them based on their SD. For further analyses, we selected the most variable genes: 177 genes with SD values above the 99th percentile of the SD value distribution. The 99th percentile was chosen as a stringent data filtering threshold that allowed us to focus on the most highly variable genes in the dataset. This threshold also resulted in a number of genes that was suitable for both computational analysis and post-processing expert interpretations. We computed the gene-gene (Pearson) correlation coefficients between all the 177 genes and merged them into a single gene expression correlation network. We applied WiPer19 to this fully-connected weighted network to detect highly connected nodes (hub genes). This method was selected because: a) it was developed in our team; b) unlike other methods, it offers strict statistical support, i.e., corrected P-values, for each weighted degree value estimated in the network; c) we, and others elsewhere, have previously shown its usefulness for making biologically-relevant predictions20–22. For each network node, WiPer computes the weighted degree and a corresponding P-value to assess the significance of the observed values, and adjusts it for multiple testing. Genes exhibiting (Bonferroni-adjusted) P<0.05 (100K random network samples for WiPer permutation test) were considered hubs (47 genes) (Dataset 1)23. Drug sensitivity information was not used to select hubs. The resulting 47 genes were examined with different Gene Ontology (GO) and biological pathway analysis tools (below). For each hub gene, we estimated the correlation of its expression profile (across all samples) with the activity area (AA) values available from all sample-drug combinations. The CCLE used the AA as indicator of drug sensitivity. It has been shown that the AA is: a. an accurate estimator of drug efficacy and potency, and b. negatively correlated with the half-maximal inhibitory concentration (IC50), which is an alternative measure of drug sensitivity12. We compared hubs and non-hubs on the basis of such individual expression-sensitivity correlations.\n\nWe represented each CCLE sample (cell line-drug combination) with the expression values of the 47 hub genes and their corresponding AA values. The full list of CCLE drugs and their annotations are available in the Supplementary Information of12. We focused on samples with complete expression and AA data. The resulting set of 10,981 (cell line-treatment) samples was used for training and testing regression models. The dataset was standardized by re-scaling each gene so that each gene has mean and standard deviation of 0 and 1 respectively. For each model, we implemented 10-fold cross-validation (CV) for separating training from testing and for assessing prediction performance. We also used leave-one-out CV (LOOCV) and similar prediction performance results were obtained. Diverse regression techniques with different levels of complexity were investigated. We focused on a multiple linear regression model with Ridge regularization (Ridge parameter = 1E-08) because its performance (regression errors) was better than or comparable to those obtained with other techniques, such as support vector machines and k-nearest neighbors, and because of its interpretability in comparison to relatively more complex models. Moreover, we applied ridge regression to achieve a balance between model simplicity, interpretability and prediction power. As in the case of other regularization techniques, by introducing such a ridge penalty, we aimed to reduce the risk of overfitting. Although lasso or elastic net regularizations are also suitable approaches, they would have required the estimation of additional learning parameters and the removal of genes, which were deemed biologically interesting before model training. Moreover, ridge regression allows us to address the problem of multiple collinearity. This is particularly relevant to our research problem as our genes converge to different cancer-related pathways and their expression correlations offer complementary predictive information.\n\nThe accuracy of model predictions was assessed by measuring their (Pearson, Spearman and Kendall) correlations with the observed values in the CCLE and the concordance index (CI). The CI approximates, for a random pair of samples, the probability of correctly predicting which sample is more (or less) sensitivity than the other24. A CI equal to 0.5 indicates that the model’s performance is comparable to that from a random predictor, while an index equal to 1 represents the perfect predictor.\n\nRaw expression data were obtained from the ArrayExpress database (accession number E-MTAB-3610) and drug sensitivity (natural logarithm of the IC50 in μM, LNIC50) were downloaded from GDSC database (release 5.0). We normalized raw expression data with the RMA function of the R oligo v.1.42.0 package25. Then we averaged the resulting log2 probe-set intensities to estimate the expression of each gene. Associations between probe-sets and gene symbols were obtained through the hgu219.db v. 3.2.3 annotation package26. For each cell line-drug experiment available (sample), we retrieved the expression data for the 47 genes used as inputs to our prediction model and retrieved the corresponding drug sensitivity values. We focused on the 16 drugs found in both this and the CCLE dataset. This resulted in a dataset consisting of 9984 samples, each one represented by 47 gene expression values and one LNIC50 value. We standardized expression data as in the case of the CCLE dataset, reformatted the file and input it to the CCLE-derived prediction model (further information below). For each sample in the dataset, the model predicted a drug sensitivity score (approximation of AA). We compared predicted vs. observed values using the indicators applied to the CCLE dataset analysis. We adapted the CI to account for the fact that AA and LNIC50 are expected to be inversely correlated, i.e., for a given sample, concordance is achieved when a high (predicted) AA value corresponds to a low (observed) LNIC50 value, and vice versa.\n\nAccess to the CCLE and GDSC datasets, including extensive documentation, are provided in their respective original publications and data websites.\n\nFor CCLE RNA-Seq analyses, the original data were downloaded from the CCLE website. Ensembl gene IDs were annotated by gene symbols (GRCh37.69), which were used as unique identifiers. We intersected features (rows) and experiments (columns) of microarray and RNS-Seq datasets and thus obtained two expression matrices of the same size with 16744 rows and 970 columns. RPKM values of RNA-Seq dataset were additionally log2-transformed: expression = log2(1+RPKM). Next, Spearman correlation was calculated between gene expression profiles corresponding to the same samples. Drug sensitivity prediction model was trained and tested as done with the microarray data. We investigated gene length as a potential source of bias in our analysis as done in 27. As such, we used the maximal transcript length of a gene based on the GRCh37.69 annotation.\n\nU87 cells, initially obtained from the ATCC (Rockville, USA), were kindly provided by Prof. Rolf Bjerkvig (Department of Biomedicine, University of Bergen, Norway), and were cultured as monolayers in DMEM containing 10% FBS, 2 mM L-Glutamine and 100 U/ml Pen-Strep (Lonza). GBM stem-like cultures (NCH421k, NCH601 and NCH644) were kindly provided by Christel Herold-Mende (University of Heidelberg, Germany) and were cultured as 3D non-adherent spheres as previously described28,29.\n\nWe measured the (baseline) gene expression of four GBM cell lines using GeneChip Human Gene 1.0 ST Arrays (6 U87, 6 NCH421k, 3 NCH644 and 3 NCH601 biological replicates), as reported29. For our model’s 47 genes, we also validated gene expression measurements using quantitative PCR (qPCR) for U87, NCH421k and NCH644 cell lines (each one in triplicate). To this aim, RNA was extracted from 1x106 cells using TRI Reagent® (Sigma-Aldrich). RNA isolated in the aqueous phase with a Phase lock gel-Heavy (5 Prime) was precipitated with 100% isopropanol and purified using RNeasy® Mini kit combined with an on-column DNase treatment (Qiagen). For the qPCR, RNA was reverse-transcribed into cDNA using Superscript III™ (Invitrogen) following manufacturer’s instructions. qPCR was performed in 96-well plates using SYBR® Green Master Mix (Bio-Rad) and CFX-96 thermal cycler (Bio-Rad). Normalized gene expression levels were calculated using the CFX manager 3.1 software (Bio-Rad) via the delta-delta Cq method with “Hspcb, Rps13, 18sRNA” as reference genes and taking into account the calculated amplification efficiency for each primers pair. We provide a MIQE-compliance checklist table and details of procedures in Dataset 223.\n\nThe gene expression dataset was standardized as above. Each sample, represented by a 47-gene expression profile, was input to the prediction model and a drug sensitivity value was predicted for each one of them (18 samples in total), for each of the 24 drugs included in the model. Predicted values were compared between them to determine their relative differences in terms of cell lines and drugs. Next, these predictions were compared to the in vitro sensitivity values that were obtained as follows. We tested four drugs: paclitaxel (Sigma-Aldrich), panobinostat, 17-AAG and erlotinib (all Selleck Chemicals) independently on the selected four GBM cell lines with eight drug concentrations (details below and in Dataset 3)23. For each cell line and dose, we performed treatment experiments in triplicate (i.e., 3 treated biological replicates / dose). As a measurement of drug sensitivity, WST-1 (Sigma-Aldrich) cell viability assays were implemented. U87, NCH421k, NCH644 and NCH601 cell lines were seeded into 96-well plates at densities of 1,500, 5000, 4000 and 6000 cells per well, in appropriate culture medium29. Cells were incubated, 24h hours after seeding, with the 8 different drug concentrations ranging from 10 µM to 6.1x10-4 µM, with a final volume of DMSO not exceeding 0.1% and each condition was tested with six technical replicates. After a 72-h incubation, WST-1 reagent was added in medium to a final concentration of 10%. The adherent cell line (U87) was incubated at 37°C for 2 hours and 3D sphere stem-like cell lines (NCH421k, NCH644 and NCH601) were incubated at 37°C for 6–8 h. Absorbance was measured against a background control at 450 nm on a FLUOstar OPTIMA Microplate Reader (BMG LABTECH). Using the normalized viability measurements, we generated drug dose-response curves and estimated IC50 values (μM) for each sample-drug combination. The dose-response curves were fitted with a four-parameter logistic regression model, whose parameters were calculated using GraphPad Prism 7 (GraphPad).\n\nWe performed multiple comparisons of our hub-based prediction model versus other approaches, including published research. To compare our results with those reported previously30, we implemented an elastic net model. The elastic net model selected has λ and α parameters equal to 0.00105 and 0.95, respectively. The λ value was estimated using the cv.glmnet function (λ value reporting the lowest MSE in a 10-fold cross-validation) in R. The models were trained and tested using 5-fold cross-validation, and were compared on the basis of the CI between the predicted and observed activity areas. To compare our results with those reported previously31, we implemented a SVM using the R package e1071 v. 1.6.8 with default settings excepted for gamma. For this parameter, we used the optimal values determined by Dong et al. for each drug31.\n\nLASSO models that optimize drug sensitivity estimation were also investigated. Such models were generated in R using the glmnet v. 2.0.16 package (α = 1). We built models and evaluated prediction performance using a nested CV procedure, and CIs between predicted and observed sensitivity values were reported. We ensured that each of the 10-folds had the same proportion and distribution of sensitivity values for each drug. Within each CV iteration, the cv.glmnet function was used to determine the optimal lambda (using 10-fold CV and based on the minimum RMSE). For model applied to our 47 genes: Optimal λ mean value = 0.0004 (range: [0.00037, 0.0007]).\n\nWe used the R statistical environment for data analysis and visualization (www.r-project.org), packages: ggplot2 v.2.2.1, pheatmap v. 1.0.10, ComplexHeatmap v.1.17.1 and SNFtool v.2.3.032. Concordance indexes24 were calculated based on rescaled Kendall rank correlation coefficients, which were also used to estimate confidence intervals (by Fisher’s transformation). For network analyses, we applied Cytoscape for visualization33, MINE for similarity exploration34 and WiPer for network hub identification19. REViGO35 and g:Profiler36 were applied for biological process and pathway enrichment analyses. The Weka workbench was used for building and testing regression models37,38, and GraphPad Prism 7 for analyzing drug response curves. A two-tailed, Student’s t-Test was used to estimate statistical differences between correlation values from hubs and non-hub genes. We provide researchers with a Web-based application to enable them to predict anticancer drug sensitivity using their own (47-gene) transcriptomic data (Results). The tool is based on the R Shiny package. Although this package offers useful functionality for generating an interactive user interface, we customized available code using the R Shinyjs package. Users can input pre-processed expression datasets. Alternatively, our application can also implement z-score rescaling of the input data. Figures containing the prediction results can be downloaded and stored as either .png or .jpeg files. Results are also shown as tables with sample-specific predictions (in rows) with their corresponding drugs (in columns), and may be stored as either .csv or .tsv files.\n\n\nResults\n\nOur hypothesis was that genes highly connected within co-expression networks, i.e., hubs, may be reflective of molecular activity relevant to drug response, across biological processes and tissue sites. To test this hypothesis, we analyzed the CCLE gene expression dataset, which was derived from 1037 (untreated) cell lines representing different cancer types from 18 tissue sites. To reduce network complexity while aiming at preserving potentially relevant information across all samples, we selected genes with highly variable expression pattern across cell lines (i.e., 177 genes with standard deviation of expression values across cell lines located above the 99th percentile). Using the pan-cancer expression profiles from these genes, we calculated all the between-gene (Pearson) correlation values and merged them into a fully-connected weighted network (Figure 2A), which included 177 nodes and more than 15K edges, i.e., correlations (Dataset 1)23.\n\n(A) Snapshot of a (fully connected) weighted gene correlation network from untreated cell lines. Nodes and edges representing genes and their correlations respectively. Network hubs and non-hubs are colored in red and white respectively. A zoom-in view of examples of hub and non-hub nodes. The color intensity of edges reflect the expression correlations between such nodes and others in the network. (B) Graphical summary of (non-redundant) Gene Ontology terms statistically over-represented in the list of 47 hub genes. Terms are projected onto a semantic similarity space with REViGO35, in which similar terms are positioned closer to each other. Each term is represented by a bubble with color and size indicating the term’s level of statistical enrichment in our list and frequency in the GO database, respectively. (C) Comparison of hubs vs. non-hubs on the basis of their individual associations with drug sensitivity (P < 0.0001, two-tailed, Student’s t-test). The boxplot depicts the mean correlation between the gene expression and the Activity Area (AA) values across CCLE cell lines. Box notches indicate 95% confidence interval for each median value. Non-overlapping notches indicates a significant difference at the 95% level. (D) Cell line-drug experiments are visualized in terms of the 47-gene expression data. The panel above the gene expression heat map illustrates the AA values observed for selected sets of cancer cell lines (grouped by tissue site) and two compound examples (erlotinib and paclitaxel) for illustration purposes.\n\nWe identified network hubs by extracting those genes with statistically detectable connectivity scores (i.e., weighted degree values) using WiPer19. This resulted in 47 hubs (WiPer-adjusted P < 0.05, Supplementary Data S1), one of which (ANAX1) is illustrated in Figure 2A together with an example of a non-hub node (HCLS1). A hub is distinguished by the weighted degree, i.e., sum of the edge weights linked to the gene, together with its associated statistical significance (Methods). In Figure 2A, this is in part illustrated by the intensity of the edges (i.e., HCLS1’s edges are lighter than ANAX1’s edges). The 47 hub genes are significantly implicated in a wide diversity of biological processes and pathways of relevance to cancer progression and therapeutic response. They include cell proliferation, death, migration, adhesion, angiogenesis, kinase signaling and the extracellular matrix (Figure 2B and Supplementary Figure S1 in Dataset 3)23.\n\nWe also investigated the connections between the enriched biological processes (Figure 2B; GO terms) and known drug targets. Genes associated with a particular GO term were matched to known drug targets annotated in the DGIdB database39. We found that within each biological process term, different genes are known targets of different drugs, though the majority of them are not known to be targets for the drugs investigated here (Supplementary Figure S2 in in Dataset 3)23. Furthermore, only four genes in our list of hubs are known drug targets: DKK1 (Irinotecan), MYB, SPARC and TUBB6 (the latter three targeted by paclitaxel).\n\nA GO enrichment analysis of all the genes in the network reported a larger number of statistically enriched GO terms in comparison to the analysis focused on the 47 hubs (biological processes: 196 vs. 74 terms). This may be explained by the increased in the number of genes analyzed. Both gene sets shared in common several significantly enriched processes, including: cell adhesion, proliferation and death. However, there are biological processes that were statistically overrepresented in the 47 hubs exclusively, including endocytosis and several processes specialized in responses to different biological stimuli. These results underscore the significant implication of the 47 hubs in a wide range of cancer-relevant biological processes.\n\nNext, we analyzed the drug sensitivity data (activity areas (AA)) available for these cell lines (11670 cell line-drug experiments) in the CCLE. The AA, which is inversely correlated with the IC50, was defined by the CCLE to approximate the efficacy and potency of a drug simultaneously12. We stress that such data were not considered during the network generation and analysis steps outlined above. For each gene in the network, we calculated the correlation between gene expression and AA across all available (cell line-drug) data, and observed that: a. the expression of hub genes tend to be anti-correlated with drug sensitivity, and b. although such correlations are weak, they are stronger than in the case of non-hub genes (Figure 2C, P < 0.0001, two-tailed, Student’s t-Test). The 47 hub genes did not include previously reported markers of drug sensitivity, e.g., ALK, BRAF, ERBB2, EGFR, HGF, NQO1, MDM2, MET and VEGFRs12,40. A possible explanation is that our discovery strategy was not oriented or biased to specific drugs or target families. Moreover, different genes may be associated with a specific drug response without actually representing known targets for the drug.\n\nTo further illuminate the information encoded by the 47 hubs, we clustered the samples (available cell line-drug experiment data) based on their (baseline) expression profiles (Figure 2D). Although, this analysis is based on a simple hierarchical clustering technique and the genes do not clearly segregate all samples in terms of drug responses, these results illustrate the heterogeneity of gene expression profiles and motivated us to further investigate their predictive potential. Using an alternative visualization and (unsupervised) clustering technique, a similar observation could be made (Supplementary Figure S3 in Dataset 3)23. Overall, these results suggest that our 47 hubs represent a novel, biologically meaningful gene set with drug sensitivity prediction potential.\n\nWe used the expression values from the 47 network hubs and drug sensitivity data (n = 10,981, untreated cell line-drug experiments, i.e., samples, with full expression and AA data available in the CCLE) to generate a drug sensitivity prediction model based on multiple linear regression. For a given sample (47-gene expression profile) and drug (identity of one of the 24 CCLE drugs), the model estimates a sensitivity score that approximates the AA values observed in the CCLE. For model training and testing, we used separate datasets respectively through a 10-fold cross-validation sampling procedure. Prediction capability was evaluated with multiple performance indicators that compare the predicted and observed sensitivity values: Pearson, Spearman and Kendall correlations, and a concordance index (CI) (Figure 3). The R code specifying our prediction model is available on Zenodo41.\n\n(A) Density plot of predicted vs. actual sensitivity values (n=10981). Pearson, Spearman and Kendall correlation coefficients: 0.86, 0.73 and 0.54, respectively. (B) Focused view of the predicted vs. actual sensitivity for panobinostat, one of the drugs displaying the highest (actual and predicted) activity area (AA) values. Additional examples in Supplementary Figure S5. (C) Concordance indices between the predicted and the observed AA values for a selected set of drugs. An index value = 0.5 is the expected value from random prediction. Error bars: 95% confidence interval of the estimated concordance index.\n\nFigure 3A and Supplementary Figure S4 (Dataset 3)23 show that the predicted and actual AA values are positively correlated (Pearson, Spearman and Kendall, correlations coefficients: 0.86, 0.73 and 0.54 respectively). In Figure 3A, it is also possible to distinguish a number of clusters that are linked to several drugs with different observed (and predicted) drug sensitivities (Supplementary Figure S5 in Dataset 3)23. For example, the cluster located on the top-right of the plot corresponds to Paclitaxel, followed by a cluster associated with panobinostat, and a third cluster consisting of a mixture of samples tested with 17−AAG, Irinotecan and topotecan. Figure 3B includes a focused view of the predicted vs. actual sensitivity for panobinostat, one of the drugs displaying the highest (observed and predicted) AA values. This plot and others in Supplementary Figure S5 (in Dataset 3)23 indicate that there are drugs for which our model can make relatively accurate sensitivity predictions in comparison to other drugs in this dataset.\n\nTo provide further insights into our model’s prediction capacity, Figure 3C displays the CI for a selected set of drugs. For a random pair of samples, the CI estimates the probability of correctly predicting the relative sensitivities of such samples (e.g., sample X is more sensitive than sample Y) in relation to the observed relative sensitivities. Perfect and random prediction performances are indicated by concordance indices equal to 1 and 0.5 respectively. Our model reported concordance indices with median values above 0.5. Altogether, these results suggest that our 47 hubs are linked to drug responses in vitro, and that their predictive potential deserves further investigation.\n\nWe compared our results to those obtained from the CCLE’s RNA-Seq dataset, which was made publicly available last year. First, we investigated the similarity of the (original) microarray and RNA-Seq datasets and observed a high level of concordance between these datasets, with mean Spearman correlation between gene expressions profiles of 0.87 (confidence interval at 95%: 0.870-0.871). The correlations for our network hubs was even higher: 0.94 (global expression of 47 genes among all cell lines) (Supplementary Figure S6 in Dataset 3)23. Also we generated a mean-standard deviation representation of the genes characterized by both techniques (Supplementary Figure S7 in Dataset 3)23. In both platforms, the 47 genes show high variability and moderate average expression, and none of them was lowly expressed. These observations indicate the inter-platform robustness of the network hubs in terms of their gene expression.\n\nWe also investigated the predictive performance of our model when RNA-Seq data were used instead of microarrays. The overall prediction performance obtained in both application scenarios was almost identical. CI: 0.772 vs. 0.770, and Spearman correlation: 0.728 vs.0.725 (microarray and sequencing data respectively). Lastly, we further compared the connectivity of our 47 genes in networks generated with data from the two platforms independently. We regenerated gene networks based on microarray and sequencing data, and this time considered the sum of R2 as a measure of degree of each gene (node) and visualized the distribution for all genes and the 47 hubs. We observed that, in both platforms, our 47 genes are shown as top hub genes (Supplementary Figure S8 in Dataset 3)23. These analyses corroborate the robustness of the gene expression profiles and predictive properties of our hub-based signature in microarray and RNA-Seq platforms. Also we assessed the length of the genes in our signature and found that 42 of 47 genes were longer than 2000 nt. Based on our previous experience27, we should not expect negative effects switching from arrays to sequencing for the vast majority of the genes.\n\nWe tested the drug sensitivity prediction capacity of our 47 hubs on the 2016 release of the GDSC dataset, which partially overlaps with the CCLE in terms of cell lines and drugs42. To allow our CCLE-derived model to make predictions on this dataset, we focused on the 16 drugs that are found in both datasets. First, as in the case of the CCLE data, we show that the (baseline) expression profiles of these 47 genes are diverse across samples and drugs (Figure 4A, and Supplementary Figure S3 in Dataset 3)23. Note that in the GDSC dataset drug sensitivity is represented as the logarithm of IC50 (LNIC50) values (AA values were not provided in this dataset).\n\n(A) Cell line-drug experiments are visualized in terms of the 47-gene expression data. The panel above the gene expression heat map illustrates the natural logarithm of half-maximal inhibitory concentration LNIC50 (μM) values observed for selected sets of cancer cell lines (grouped by tissue site) and two compounds (erlotinib and paclitaxel). (B) Application of CCLE-derived model to the GDSC data. Density plot of predicted (activity area (AA)) vs. actual sensitivity (LNIC50) values for drugs that are common between the CCLE and GDSC (n = 9984). Pearson, Spearman and Kendall, correlations coefficients: -0.72, -0.71 and -0.50 respectively. (C) Concordance indices between the predicted and the observed sensitivity values. An index value = 0.5 is the expected value from random prediction. Indices are corrected to account for the notion that higher concordance is reached when high AA (prediction) corresponds to a low LNIC50 (observed) values, and vice versa. Error bars: 95% confidence interval of the estimated concordance index.\n\nNext, we applied our (CCLE-derived) prediction model to the GDSC data and made sensitivity predictions (AA values) for all the samples (cell line-drug experiments) available (Methods). The resulting predictions were then compared with the actual sensitivity values in the GDSC dataset (Figure 4B, and Supplementary Figure S4 in Dataset 3)23. As required, the predicted (AA) and actual sensitivity (LNIC50) values for these samples (n = 9984) are anti-correlated (Pearson, Spearman and Kendall, correlations coefficients: -0.72, -0.71 and -0.50, respectively). This indicates that our 47-hub-based model is, in general, estimating sensitivity values that are in agreement with those observed in a test dataset, i.e., higher predictive agreement is reached when high AA (prediction) relates to a low LNIC50 (actual) values, and vice versa.\n\nFigure 4C summarizes the assessment of our model’s predictive performance on the GDSC dataset based on (drug-specific) CIs, as done for the CCLE dataset (Figure 3). Concordance indices > 0.5 were obtained for 12 out of the 16 drugs, and (among those 12 drugs) concordance estimates for 9 drugs can be reliably interpreted as larger than 0.5 (95% confidence intervals of the estimated indices). The predictive performances for several drugs (e.g., Nilotinib, Nutlin-3 and Sorafenib) are very similar to those estimated in the CCLE dataset. As in the CCLE dataset, the sensitivity observed in samples treated with AZD0530 and Lapatinib proved to be more difficult to accurately predict. Although concordance indices > 0.5 were obtained for irinotecan and paclitaxel predictions, this represented a reduction of prediction performance in comparison to the predictions made for CCLE samples. The prediction performance of 17-AAG, PD-0325901 and TAE684 were also diminished. Overall, our findings further suggest that our network hubs are relevant for predicting drug sensitivity, and highlight challenges in a drug-specific context.\n\nAs the GDSC dataset shares different cell lines in common with the CCLE, we also assessed the prediction performance of our hub-based prediction model on GDSC cell lines that are not included in the CCLE. To do this, we applied our CCLE-based model on the GDSC dataset and made a distinction between predictions for overlapping and unique cell lines. When focused on experiments with cell lines found in both data sets, we obtained the following correlations between predicted (AA) and observed sensitivity values (IC50): -0.73 (Pearson), - 0.72 (Spearman) and -0.52 (Kendall). For cell lines uniquely represented in the GDSC, we obtained the following correlations: -0.72, -0.68 and -0.48. Although a slight reduction in prediction performance is observed, these results are comparable and stress the robustness of our prediction results for different types of cell lines, including those not included in our hub discovery and model training dataset.\n\nAlso we investigated the stability of hubs across the CCLE and GDSC datasets. To do this, we repeated the network generation and hub identification procedures on the GDSC with WiPer (Methods). This analysis resulted in the detection of 69 network hubs (as before, WiPer adjusted P-value < 0.05). Among such genes, 23 genes are also found in our 47-gene signature, such as: VEGFC, CAV2, MYOF, CAV1 and TM4SF1. Although this overlap does not include the full set of hubs obtained in our CCLE analysis, it gives an indication of the robustness of a set of such genes despite the important differences between the datasets in terms of the numbers and types of cell lines.\n\nTo further demonstrate the robustness of our predictions, we implemented multiple runs (or iterations) of the 10-fold cross-validation (CCLE data) and assessed their reported performances. For 100 independent (10-fold) cross-validations, the prediction performance is very similar: all iterations reporting CIs between 0.765 and 0.77, and a coefficient of variation = 0.026% (Supplementary Figure S9 in Dataset 3)23.\n\nUsing our 47-hub signature, we also investigated (multiple linear regression) models trained on the GDSC and tested on the CCLE datasets. Although comparable with the cross-validation results obtained with the CCLE dataset, the GDSC-based cross-validation showed an overall improvement in drug sensitivity prediction performance: CI = 0.82, rs = 0.82 and rk = 0.64 (Supplementary Figure S10 in Dataset 3)23. Next, we applied the resulting model to the CCLE dataset. The prediction performance is similar to that obtained with the CCLE-derived model (Supplementary Figure S11 in Dataset 3)23. Moreover, as in the CCLE-derived model, we observed that the predictive quality is relatively higher or deteriorated according to specific drugs.\n\nWe also investigated the impact of reducing the 47-gene set on prediction performance. We used our 47 genes as inputs to LASSO modeling, and we observed that is possible to generate models with an average of 44.6 genes (range: 43 to 46 genes). However, LASSO-based models offered very similar prediction performance in comparison with our 47-gene model (CCLE, using a nested 10-fold CV, mean CI: 0.77 ± 0.004).\n\nWe also implemented a drug sensitivity prediction model based on LASSO using all gene expression features as inputs to the model. The resulting model consisted of 605 genes, which did not include any of our 47 hubs. When comparing the prediction performance of our 47-gene model vs. the 605-gene LASSO model, we did not observe significant differences, though the latter offered a slightly higher prediction performance (CCLE, nested 5-fold CV, CI: 0.77 vs 0.80). This relative improvement in performance is not surprising as the LASSO model, unlike our hub discovery strategy, explicitly sought to identify the best set of genes for optimizing this specific regression task.\n\nTo assess the effect of network size on the identification of hubs, we applied our hub detection analysis to a larger network consisting of 530 genes. These genes were selected with a more flexible filtering criterion (Methods): Genes showing SD of expression above the 97th percentile of the SD value distribution. As expected, a larger number of significant hubs were detected in this network (203 hubs, at corrected P-value < 0.05). Among them, our original set of 47 hubs were included, which reiterates their statistical significance and robustness of our analysis. This was also observed when repeating the analysis using a far less stringent procedure for estimating statistically significant hubs, i.e., P-value estimation. Using only 1000K permutations to estimate P-values, we detected 212 candidate hubs (corrected P-values < 0.05) that also included our original set of 47 hubs.\n\nWe re-implemented models previously reported30,31, and compared their performance with our model. We chose these works because of their model coverage and analytical depth using different supervised prediction techniques. However, note that unlike our discovery strategy, their models were based on input genes that were explicitly sought to optimize drug sensitivity prediction. Also, unlike our model, Dong et al.31 considered prediction of drug sensitivity as a classification problem. Given a gene expression dataset, their approach aimed at assigning each sample to one of two pre-established response classes: resistant and sensitive. They used CCLE data to build their models. For each drug, they started by discretizing a “scaled AA” (sAA) into three categories: resistant if sAA < -0.8 SD (standard deviation which is equal to 1), sensitive if sAA > 0.8 SD and intermediate otherwise. After removing samples with an intermediate response, they focused on the classification of the extreme response classes (resistant vs, sensitive). Their drug-specific models were based on a support vector machine (SVM) and recursive feature selection using gene expression data. They reported an accuracy of 0.81 (on average) when their models were cross-validated on the CCLE. The performance was considerably reduced when tested on GDSC data (only 3 out of 11 drug models reported an average AUC equal to or above 0.69).\n\nTherefore, to directly compare Dong et al.’s models with ours, we had to re-specify and re-implement our drug sensitivity prediction approach. This is needed because our approach is defined as a regression problem and is not constrained to predetermined sensitivity classes. Hence, we first labeled the samples as sensitive and resistant as done by Dong et al. We then tested whether the predicted sensitivities (predicted AA values from our model) correctly assign each sample to the “right” sensitive and resistant classes. The predictive performances of our and Dong et al.’s models are comparable with a small advantage for Dong et al.’s models (average AUCs = 0.79 vs. 0.73, Supplementary Figure S12 in Dataset 3)23. However, this advantage is not surprising since Dong et al.’s models optimizes the separation of two well-separated sensitivity classes. Our predictions are obtained from a regression model trained and tested on all samples with all available sensitivity values. Despite such a caveat, the prediction performance achieved by our 47-hub model was very similar to the performance from Dong’s drug-specific models except for five drugs (AZD0530, erlotinib, lapatinib, LBW242, PD-0325901) out of 21 models (drugs), and our model clearly outperformed their model for one drug (PD-0332991).\n\nIn the comprehensive study by Jang et al.30, thousands of models were compared and the authors concluded that an elastic net-based model was the best choice. Therefore, we trained and tested an elastic net model, and compared it to our model. The models were trained and tested using 5-fold cross-validation, and were compared on the basis of the concordance between the predicted and observed activity areas. The elastic net model, overall, outperformed our 47-hub model (average CI of 0.81 vs. 0.77). However, the elastic net model required 613 genes as input features to achieve this performance. As the difference in concordance between these models was only 0.04 on average, we also compared the individual predicted sensitivity values generated by the two models. We found that their predicted sensitivity values are highly correlated (0.99 of correlation and average difference of 0.02). These results, which are graphically illustrated in Supplementary Figure S13 (in Dataset 3)23, indicate that these models’ prediction performances are comparable.\n\nAdditionally, we implemented prediction models based on the gene expression of well-known markers for drugs used in clinical practice, and which were also included in our datasets. Here we report results for two such markers: PDGFR (a target of Sorafenib) and EGFR (and target of Erlotinib), which were used as inputs to prediction (linear regression) models. To make an unbiased comparison, we compared prediction performances specific to each drug. For both drugs, we found that models built with our 47 hub genes outperformed models built with the gene expression of these targets. For erlotinib, our model reported a CI = 0.62, while the EGFR-based model showed a CI = 0.57. The difference was more significant for sorafenib: Our model reached a CI = 0.57, whereas models built with either PDGFRA or PDGFRB reported CIs below 0.5 (0.48 and 0.47 respectively).\n\nTo further validate the prediction potential of our network hubs on independently-generated data, we made predictions and performed in vitro tests for several GBM cell lines and compounds in our lab. First, we measured the (baseline) expression profiles of four (untreated) GBM cell lines that have been well-characterized in our lab: U87, NCH644, NCH601 and NCH421k. While the CCLE and GDSC datasets included U87, the latter three are stem-like GBM cell lines that were not included in our previous analyses.\n\nAlthough genome-wide expression data can appropriately cluster multiple samples (biological replicates) from such cell lines, we found that the expression profile of our 47 genes are sufficient to achieve the same biologically-meaningful segregation while offering a clearer, fine-grained view of their differences (Supplementary Figure S14 in Dataset 3)23. We also verified the platform-independent replicability of these results with another 47-gene expression dataset derived from three of these cell lines measured with qPCR (Supplementary Figure S14 in Dataset 3)23. These results show the biologically-relevant discriminatory capacity and reproducibility of our 47-hub expression profiles in our set of brain cancer cell lines using microarrays and qPCR. Raw qPCR Cq values are available on Zenodo23.\n\nNext, our model predicted the sensitivity of the four GBM cell lines (18 samples in total, Methods) against the 24 drugs included in the model. The baseline 47-hub expression profiles of these cells were input to the prediction model (six U87, three NCH644, three NCH601 and six NCH421k gene expression profiles). Figure 5A summarizes the 432 predicted sensitivity (AA) values according to drug (18 predictions per drug). To investigate such predictions in vitro, we focused on the top-3 drugs associated with the highest predicted sensitivities (paclitaxel, panobinostat and 17-AAG), as well as on erlotinib, which was predicted as an ineffective compound. The main reason for the selection of these compounds was our interest in investigating compounds predicted to be highly active (3-top drugs) together with a “negative” control that was predicted, and expected, to have lower activity (Erlotinib). Moreover, these drugs correspond to four different drug classes: cytotoxic, histone deacetylase inhibitor, antibiotic derivative and an EGFR inhibitor respectively. In the case of erlotinib, the predictions are consistent with the fact that the tested cells do not (NCH644, NCH421k) or very lowly (U87, NCH601) express EGFR. Figure 5B and Supplementary Figure S15 (in Dataset 3)23 show a more focused view of the predicted sensitivity values for our samples against these four drugs.\n\n(A) Sensitivity predictions (horizontal axis) for 24 drugs (vertical axis). Box plot summarizes the (432) predicted sensitivity (activity area (AA), as defined in the prediction model) values for four glioblastoma cell lines: U87, NCH644, NCH601 and NCH421k. Only the U87 cell line was included in the model learning phase. The 47-gene expression profiles of multiple biological replicates (18 samples in total) were input to the prediction model (six U87, three NCH644, three NCH601 and six NCH421k samples). (B) Alternative boxplot summary of the prediction results for four drugs (erlotinib, 17-AAG, panobinostat and paclitaxel) and the different cell lines. These drugs, which were selected for subsequent in vitro tests, were predicted to be relatively highly (17-AAG, panobinostat and paclitaxel) and lowly (erlotinib) effective against the four cell lines. (C) Summary of in vitro test results. The selected drugs were tested on each cell line in triplicates, relative viability (vs. vehicle-treated samples) was measured for eight drug concentration values (µM) and half-maximal inhibitory concentration (IC50) values were estimated for each drug-sample experiment. The boxplot shows the resulting natural logarithm of IC50 (LNIC50) values obtained. Drug response data for NCH601 samples and erlotinib are not available, and for NCH644 samples and erlotinib not shown because of lack of effect. Boxes show the median, the 25th and 75th percentiles (lower and upper hinges), and (1.5x) interquartile ranges.\n\nWe tested the selected drugs on each cell line, in triplicates, and measured their response based on their relative viability (i.e., normalized to vehicle-treated samples) for eight drug concentration values (µM). For each treated cell line, we estimated the IC50 values and compared them on the basis of cell line and drug groups. Figure 5C summarizes the results with boxplots showing the LNIC50 values. Drug response data for NCH601 samples and erlotinib were not available (not tested), and data for NCH644 samples and erlotinib are not shown due to lack of effect. Supplementary Figure S16 (in Dataset 3)23 includes all the drug response curves and additional details.\n\nAs predicted by our model, all our cell lines exhibited the lowest sensitivity, i.e., the highest IC50 values, when treated with erlotinib (median LNIC50 = 0.74 µM). Overall, U87 tended to be the least sensitive cell line in relation to all four drugs (median LNIC50 = -1.27 µM across all sample-drug experiments), though it did not show the lowest sensitivity for every single compound or biological replicate. Our model consistently predicted NCH601 as the most sensitive cell line against all drugs (Supplementary Figure S15 in Dataset 3)23. Our in vitro tests showed that NCH421k tended to be more sensitive than NCH601 (median logIC50: -1.64 vs. -1.54 µM). Despite this particular discrepancy, we found a global agreement between predicted and observed sensitivities on the basis of cell type (Spearman correlation between the median sensitivity values, predicted (AA) vs. observed (LNIC50) in the four cell line groups: -0.40).\n\nIn accordance with the predictions, Paclitaxel was the most effective drug across all treated samples (median LNIC50 = -2.35 µM). Lesser agreement between predicted and observed sensitivities were obtained in the case of the remaining two drugs. For all samples, our model predicted overall higher sensitivity for panobinostat than for 17-AAG (Figure 5B). Relatively similar responses were obtained, in vitro, for panobinostat (median LNIC50 = -1.29) and 17-AAG (median LNIC50 = -1.33 µM), though a larger variability of sensitivity values was observed in the former case. Nevertheless, predictions and in vitro tests concordantly showed that NCH421k and U87 samples treated with panobinostat were consistently more sensitive than all samples treated with 17-AAG (Figure 5C and Supplementary Figure S16 in Dataset 3)23.\n\nWe had a closer look at topotecan, a drug that may be expected to exhibit differential activity for at least one (but not all) of the cell lines investigated. This drug is known to target TOP1 (DNA Topoisomerase I). In our set of GBM cell lines selected for validation, TOP1 is relatively highly expressed in NCH601 and weakly expressed in U87. Moreover, this target is not included in our 47-gene signature. As illustrated in Supplementary Figure S17 (Dataset 3)23, our model predicted relatively higher sensitivity values for NCH601 than for U87. Furthermore, Topotecan is predicted to be more effective than Erlotinib in all 4 cell lines. Taken together, these results provide further evidence of the potential of our 47-hub-based model for predicting drug sensitivity in vitro, and will encourage future investigations.\n\nTo enable further research, we developed a web-accessible tool that allows researchers to upload their own gene expression data, make sensitivity predictions and visualize results in a few steps (Figure 6). We term this tool: Dr.Paso (Drug Response Prediction and Analysis System for Oncology Research)41. The Help section of the website offers a guided application example using CCLE data. Users provide their input data as a text file containing the (baseline) 47-gene expression for different samples, and then can select all or specific drugs for making predictions (Figure 6A). Dataset re-scaling (feature standardization with means and standard deviations equal to 0 and 1, respectively) can be applied to harmonize the input dataset with the feature representation used in our model. Prediction results are presented with graphical displays and tables in different panels. Moreover, users can control the amount and focus of information at the drug and sample levels (Figures 6B–D). Results can be saved in different graphical and tabular file formats. The tool is freely available at www.drpaso.lu.\n\nScreenshots of: (A) Main page with user input and analysis options; (B) Global view of predicted sensitivity values for a given input gene expression dataset and all drugs available in the CCLE; (C) Alternative view of predictions focused on a specific input sample and all drugs; (D) Tabular-based view of results. All views can be selected and downloaded according to user requirements.\n\n\nDiscussion\n\nThe biological relevance of hubs in different types of molecular networks has been previously investigated, e.g., in the context of gene lethality. The predictive potential of candidate hubs in gene co-expression networks in the specific context of cancer-related drug experiments deserve deeper investigations. This is important not only for further understanding the biological roles of network hubs, but also because such knowledge may offer new opportunities for the accurate prediction of anticancer drug responses. Here we investigated the relationship between hubs detected in a pan-cancer co-expression network and drug sensitivity in vitro.\n\nThe development of computational models for estimating drug sensitivity based on gene expression data from large collections of cancer cell lines is important to support pre-clinical research, and provides a basis for future clinically-oriented applications. Our research offers insights into such challenge through the integration of network-based and statistical modeling approaches. For a given drug, we showed that in principle it is possible to predict anti-cancer drug sensitivity based on the gene expression profile of 47 genes, which represent significant hubs in a pan-cancer transcriptomic network and are prominently implicated in a variety of cancer-relevant biological processes. This is particularly appealing because at the start of our investigation we did not aim to select a specific set of genes that could optimize the supervised prediction of drug sensitivity. We implemented an unbiased discovery approach, which was motivated by the hypothesis that co-expression network hubs encode useful information for investigating drug response in vitro.\n\nThe prediction model resulting from our network hub analysis is not proposed as a competitor for existing approaches for drug sensitivity prediction. Nevertheless, our study and other previous research highlight the challenges and complementary predictive capacity exhibited by different modeling approaches15,43. No single model can consistently make accurate predictions for all drugs and cell lines available in the CCLE and GDSC datasets, including models that include genomic data and more complex learning parameters30,44,45. Different models can offer more, or less, accurate predictions for certain drugs, and there is no conclusive evidence about the dominance of a particular modeling technique46. For example, our model makes good predictions for irinotecan, panobinostat and PF2341066, all of them with AUC > 0.85 and CI > 0.6. Moreover, these examples are highly comparable with the performance obtained by previous work, e.g., in 31. Also in comparison to Dong et al.31, our model made more accurate predictions for PD-0332991 (AUC=0.84 vs. 0.75), but weaker predictions for lapatinib (AUC = 0.62 vs. 0.74). Such limitations may be partially explained by a lack of sufficient molecular information to account for the complexity of cell lines and their drug responses, choice of surrogate measures of drug sensitivity and inconsistencies of sensitivity data between the CCLE and GDSC40,47,48. The latter may also partly explain the overall degradation of predictive performance when training models on the CCLE and testing them on the GDSC.\n\nThe predictive capacity of our 47-hub model is grounded in an unbiased network-guided selection of model inputs prior to the fitting of a regression model. Future investigations, motivated by new datasets and clinically oriented questions, are certainly envisaged and are expected to include new biomarker discovery and prediction modeling strategies. There is a need, for example, for additional research on the connection between network hubs and drug sensitivity with a focus on particular cancer types or drugs. Our analyses indicate that on the basis of tissue sites, the top-3 cancer types for which our model makes relatively highly accurate predictions are: thyroid, pancreas and prostate cancers, with CIs = 0.8, 0.86 and 0.86 respectively (CCLE data and using 10-fold cross-validation). Predictions for breast-derived samples reported lower performance (CI = 0.74).\n\nOur investigation was limited to the drugs available in two well-established datasets. As larger collections of data from drug experiments become publicly available, it will be possible to develop more extensive analyses for newly approved or experimental compounds. Although we provided evidence of the robustness of our analyses when using microarray, RNA-Seq and qPCR data, the impact of expression data generation platforms on drug sensitivity prediction deserves further research. Also the analysis of larger networks, including those generated using different data filtering methods, is an interesting topic that deserves future research.\n\nHere, we focused on gene expression data for two reasons: i) Our network-based biomarker discovery strategy is based on the analysis of gene expression data; and ii) previous research (using CCLE and GDSC datasets) has indicated that, although mutation and copy number alterations can be informative, the most powerful prediction models are those based on gene expression data12,13,17,42. Nevertheless, future work could benefit from the incorporation of other “omics” data types to investigate different types of networks and hubs. Although we did not identify major effects when using the latest version of the CCLE gene expression data (RNA-Seq), future work could include additional analyses and models based on such a dataset. In this article, we started using the microarray version because it was the only gene expression dataset available at the beginning of our project. Future work may also be motivated by the fact that the CCLE RNA-Seq dataset could allow the analysis of transcript-level (gene isoform) data for predicting drug response. Such information has been recently shown to be a useful source of features for drug sensitivity prediction49. Moreover, the investigation of the biological role of hubs in gene isoform networks may open new directions for drug sensitivity research and other applications.\n\nInconsistencies in drug sensitivity as measured for the same cell lines across different studies, i.e., independent datasets, can also limit the application of insights derived from a single dataset. We expect that in the future we can address this point by either: a) weighing sensitivity values according to the available experimental evidence derived from multiple datasets, b) building global models that can generate predictions in an integrated fashion using multiple, independent datasets, or c) investigating models based on harmonized versions of datasets obtained from different studies50. Another limitation of our study is the use of two drug sensitivity measures, AA and IC50, as provided by the CCLE and GDSC datasets, to assess prediction performance. Further investigations will involve prediction performance analysis based on common measures of sensitivity. Such analyses will, nevertheless, be limited by potential inconsistencies in experimental sensitivity measurements across studies, as reported in the case of the CCLE and GDSC data40. Therefore, future work will require the incorporation of harmonized versions of such and other datasets, such as those recently generated by the PharmacoDB project50.\n\nBeyond a connectivity-centric interpretation of hubs, an interpretation of their potential functional roles in co-expression networks is not straightforward. Based on their implication in different cancer-related biological processes and their high expression correlations with many genes involved in different pathways, it is reasonable to postulate that our 47 hubs may have relevant mechanistic roles in the drug response context. Moreover, we found that these genes are related to different known drug targets via multiple biological processes, which may offer clues about the potential signaling controlling role of the hubs. However, these and alternative interpretations will require further investigations.\n\nOverall, while further investigations are needed, our study offers evidence of the relevance of gene co-expression network hubs in the context of drug sensitivity and cancer research. We hope that our findings will enable deeper investigations and pre-clinical research applications.\n\n\nData availability\n\nFull qPCR data (including raw Cq values) are available on Zenodo23.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).\n\nExtended data associated with this study are available on Zenodo23.\n\nDataset 1. Gene co-expression network data. It contains network nodes, weighted network and list of hubs.\n\nDataset 2. qPCR data from independent validation, including MIQE and additional information.\n\nDataset 3. Supplementary Figures. Legends are included under each figure.\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\nSoftware available from: www.drpaso.lu.\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.168997941.\n\nLicense: MIT license.",
"appendix": "Grant information\n\nThis research was supported by funds from the Luxembourg’s Ministry of Higher Education and Research (MESR), funding received by F. Azuaje, G. Dittmar and S.P. Niclou. GD was supported by FNR grant PEARL-CPIL.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nFor technical guidance to C.J., we thank H. Erasimus and S. Fritah (drug experiments) and V. Barthelemy, A. Bernard, J. Bohler and A. Dirkse (cell line manipulation) at the LIH NorLux Neuro-Oncology Laboratory. For helpful feedback on the manuscript, we thank L.C. Tranchevent at LIH.\n\n\nReferences\n\nCamacho D, Vera Licona P, Mendes P, et al.: Comparison of reverse-engineering methods using an in silico network. Ann N Y Acad Sci. 2007; 1115: 73–89. PubMed Abstract | Publisher Full Text\n\nEmmert-Streib F, Glazko GV, Altay G, et al.: Statistical inference and reverse engineering of gene regulatory networks from observational expression data. Front Genet. 2012; 3: 8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChai LE, Loh SK, Low ST, et al.: A review on the computational approaches for gene regulatory network construction. Comput Biol Med. 2014; 48: 55–65. PubMed Abstract | Publisher Full Text\n\nJalili M, Salehzadeh-Yazdi A, Gupta S, et al.: Evolution of Centrality Measurements for the Detection of Essential Proteins in Biological Networks. Front Physiol. 2016; 7: 375. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJeong H, Mason SP, Barabasi AL, et al.: Lethality and centrality in protein networks. Nature. 2001; 411(6833): 41–2. PubMed Abstract | Publisher Full Text\n\nYu H, Kim PM, Sprecher E, et al.: The importance of bottlenecks in protein networks: correlation with gene essentiality and expression dynamics. PLoS Comput Biol. 2007; 3(4): e59. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYu H, Greenbaum D, Xin Lu H, et al.: Genomic analysis of essentiality within protein networks. Trends Genet. 2004; 20(6): 227–31. PubMed Abstract | Publisher Full Text\n\nLi Z, Ivanov AA, Su R, et al.: The OncoPPi network of cancer-focused protein-protein interactions to inform biological insights and therapeutic strategies. Nat Commun. 2017; 8: 14356. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAhmed H, Howton TC, Sun Y, et al.: Network biology discovers pathogen contact points in host protein-protein interactomes. Nat Commun. 2018; 9(1): 2312. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZotenko E, Mestre J, O'Leary DP, et al.: Why do hubs in the yeast protein interaction network tend to be essential: reexamining the connection between the network topology and essentiality. PLoS Comput Biol. 2008; 4(8): e1000140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang Y, Han L, Yuan Y, et al.: Gene co-expression network analysis reveals common system-level properties of prognostic genes across cancer types. Nat Commun. 2014; 5: 3231. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarretina J, Caponigro G, Stransky N, et al.: The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity. Nature. 2012; 483(7391): 603–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang W, Soares J, Greninger P, et al.: Genomics of Drug Sensitivity in Cancer (GDSC): a resource for therapeutic biomarker discovery in cancer cells. Nucleic Acids Res. 2013; 41(Database issue): D955–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarnett MJ, Edelman EJ, Heidorn SJ, et al.: Systematic identification of genomic markers of drug sensitivity in cancer cells. Nature. 2012; 483(7391): 570–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCostello JC, Heiser LM, Georgii E, et al.: A community effort to assess and improve drug sensitivity prediction algorithms. Nat Biotechnol. 2014; 32(12): 1202–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStetson LC, Pearl T, Chen Y, et al.: Computational identification of multi-omic correlates of anticancer therapeutic response. BMC Genomics. 2014; 15 Suppl 7: S2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReinhold WC, Sunshine M, Liu H, et al.: CellMiner: a web-based suite of genomic and pharmacologic tools to explore transcript and drug patterns in the NCI-60 cell line set. Cancer Res. 2012; 72(14): 3499–511. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRees MG, Seashore-Ludlow B, Cheah JH, et al.: Correlating chemical sensitivity and basal gene expression reveals mechanism of action. Nat Chem Biol. 2016; 12(2): 109–16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAzuaje FJ: Selecting biologically informative genes in co-expression networks with a centrality score. Biol Direct. 2014; 9: 12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCoker OO, Dai Z, Nie Y, et al.: Mucosal microbiome dysbiosis in gastric carcinogenesis. Gut. 2018; 67(6): 1024–32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang F, Li Y, Wu X, et al.: Transcriptome analysis of coding and long non-coding RNAs highlights the regulatory network of cascade initiation of permanent molars in miniature pigs. BMC Genomics. 2017; 18(1): 148. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRodius S, Androsova G, Gotz L, et al.: Analysis of the dynamic co-expression network of heart regeneration in the zebrafish. Sci Rep. 2016; 6: 26822. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAzuaje F, Kaoma T, Jeanty C, et al.: Hub genes in a pan-cancer co-expression network show potential for predicting drug responses. Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1494802\n\nHarrell FE Jr, Lee KL, Mark DB: Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med. 1996; 15(4): 361–87. PubMed Abstract | Publisher Full Text\n\nCarvalho BS, Irizarry RA: A framework for oligonucleotide microarray preprocessing. Bioinformatics. 2010; 26(19): 2363–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarlson M: hgu219.db: Affymetrix Human Genome 219 Plate annotation data (chip hgu219). R package version 3.2.3. 2016. Publisher Full Text\n\nNazarov PV, Muller A, Kaoma T, et al.: RNA sequencing and transcriptome arrays analyses show opposing results for alternative splicing in patient derived samples. BMC Genomics. 2017; 18(1): 443. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCampos B, Wan F, Farhadi M, et al.: Differentiation therapy exerts antitumor effects on stem-like glioma cells. Clin Cancer Res. 2010; 16(10): 2715–28. PubMed Abstract | Publisher Full Text\n\nSanzey M, Abdul Rahim SA, Oudin A, et al.: Comprehensive analysis of glycolytic enzymes as therapeutic targets in the treatment of glioblastoma. PLoS One. 2015; 10(5): e0123544. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJang IS, Neto EC, Guinney J, et al.: Systematic assessment of analytical methods for drug sensitivity prediction from cancer cell line data. Pac Symp Biocomput. 2014; 63–74. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDong Z, Zhang N, Li C, et al.: Anticancer drug sensitivity prediction in cell lines from baseline gene expression through recursive feature selection. BMC Cancer. 2015; 15: 489. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang B, Mezlini AM, Demir F, et al.: Similarity network fusion for aggregating data types on a genomic scale. Nat Methods. 2014; 11(3): 333–7. PubMed Abstract | Publisher 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): 2498–504. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReshef DN, Reshef YA, Finucane HK, et al.: Detecting novel associations in large data sets. Science. 2011; 334(6062): 1518–24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSupek F, Bošnjak M, Škunca N, et al.: REVIGO summarizes and visualizes long lists of gene ontology terms. PLoS One. 2011; 6(7): e21800. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReimand J, Kull M, Peterson H, et al.: g:Profiler--a web-based toolset for functional profiling of gene lists from large-scale experiments. Nucleic Acids Res. 2007; 35(Web Server issue): W193–200. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHall M, Frank E, Holmes G, et al.: The WEKA Data Mining Software: An Update. SIGKDD Explorations. 2009; 11(1): 10–18. Publisher Full Text\n\nFrank E, Hall MA, Witten IH: The WEKA Workbench. Online Appendix for \"Data Mining: Practical Machine Learning Tools and Techniques\". Fourth Edition, ed: Morgan Kaufmann; 2016. Reference Source\n\nCotto KC, Wagner AH, Feng YY, et al.: DGIdb 3.0: a redesign and expansion of the drug-gene interaction database. Nucleic Acids Res. 2018; 46(D1): D1068–D1073. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSafikhani Z, Smirnov P, Freeman M, et al.: Revisiting inconsistency in large pharmacogenomic studies [version 3; referees: 2 approved, 1 approved with reservations]. F1000Res. 2017; 5: 2333. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAzuaje F, Kaoma T, Jeanty C, et al.: Hub genes in a pan-cancer co-expression network show potential for predicting drug responses. Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1689980\n\nIorio F, Knijnenburg TA, Vis DJ, et al.: A Landscape of Pharmacogenomic Interactions in Cancer. Cell. 2016; 166(3): 740–54. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPapillon-Cavanagh S, De Jay N, Hachem N, et al.: Comparison and validation of genomic predictors for anticancer drug sensitivity. J Am Med Inform Assoc. 2013; 20(4): 597–602. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMenden MP, Iorio F, Garnett M, et al.: Machine learning prediction of cancer cell sensitivity to drugs based on genomic and chemical properties. PLoS One. 2013; 8(4): e61318. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGupta S, Chaudhary K, Kumar R, et al.: Prioritization of anticancer drugs against a cancer using genomic features of cancer cells: A step towards personalized medicine. Sci Rep. 2016; 6: 23857. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAzuaje F: Computational models for predicting drug responses in cancer research. Brief Bioinform. 2017; 18(5): 820–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHaverty PM, Lin E, Tan J, et al.: Reproducible pharmacogenomic profiling of cancer cell line panels. Nature. 2016; 533(7603): 333–7. PubMed Abstract | Publisher Full Text\n\nCancer Cell Line Encyclopedia Consortium, Genomics of Drug Sensitivity in Cancer Consortium: Pharmacogenomic agreement between two cancer cell line data sets. Nature. 2015; 528(7580): 84–7. PubMed Abstract | Publisher Full Text\n\nSafikhani Z, Smirnov P, Thu KL, et al.: Gene isoforms as expression-based biomarkers predictive of drug response in vitro. Nat Commun. 2017; 8(1): 1126. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmirnov P, Kofia V, Maru A, et al.: PharmacoDB: an integrative database for mining in vitro anticancer drug screening studies. Nucleic Acids Res. 2018; 46(D1): D994–D1002. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "42324",
"date": "08 Jan 2019",
"name": "Elizabeth A. Coker",
"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\nIn this paper, Azuaje et al., utilise transcriptional data from the Cancer Cell Line Encyclopedia to construct gene co-expression networks and identify hubs within them. They assess the biological relevance of the hub genes, and through the use of comparatively simple linear regression-based approaches, can use gene expression of these hubs to predict drug sensitivity in vitro. Despite not initially aiming to predict drug sensitivity, the authors’ models are able to perform comparably to previously published approaches. Azuaje et al., have performed extensive and appropriate validation of their work both in silico and in vitro, including assessing the inter-platform robustness of network hubs, cross-validation of the 47-hub signature, investigating the effect of network size on hub detection, and re-implementation of their own predictive approach to allow comparison with previously published drug sensitivity predictors.\n\nThe paper is well-written and clear, with a thorough and precise account of the approaches used to assess the performance of the predictive model. The authors have produced a thoughtful and detailed account of their work which I enjoyed reading. The online Dr Paso resource is easy to use and has good documentation, although I have only tested it with the example datasets provided. I was unable to access the extended data via Zenodo as described in the manuscript (see comment below).\n\nMy current recommendation is for “Approved with reservations”, as there are a number of points I feel should be addressed prior to approval:\n\n“Furthermore, only four genes in our list of hubs are known drug targets: DKK1 (Irinotecan), MYB, SPARC and TUBB6 (the latter three targeted by paclitaxel).” I disagree with this statement. I have been unable to find any evidence of irinotecan targeting the protein DKK1, apart from an entry in DGIdb that states there is an interaction of type ‘n/a’ between DKKI and irinotecan, based solely upon a paper in Oncotarget that states DDK1 does not affect sensitivity to irinotecan in two cell lines. Equally, MYB and SPARC are not targets of paclitaxel, although they are shown in the DGIdb database as interacting with paclitaxel based on little or no evidence. I recommend the authors remove this sentence from the manuscript as it is based on a misinterpretation of DGIdb results and as such is misleading.\n\nI was able to access the code for the Dr Paso tool via Zenodo, but not the extended data associated with this study. Please ensure this is added and the appropriate link included.\n\nFigure 1: Please update this figure to highlight the filtering stage between collecting transcriptomics data and building co-expression networks. Listing the number of genes present before and after filtering would also be informative.\n\nFigure 2a: Here, the weight of edges is used to represent the expression correlations between nodes, but at present it is very difficult to see this in the examples highlighted. A larger, higher-resolution image of the network and examples is required.\n\nFigure 2b: I am not familiar with the concepts of semantic space and so do not know how to interpret this plot. Consider adding extra discussion in the main text or presenting this data in another way.\n\nFigure 2d: This figure is currently very difficult to read and could be enlarged if 2c is concurrently decreased in size.\n\nFigure 3a: The density plot requires further annotation in relation to the clusters described in the manuscript, for example by arrows or circling the appropriate regions of the plot.\n\nFigure 4a: This figure is currently very difficult to read and could be enlarged.\n\nIn conclusion this is an interesting paper illustrating how co-expression hubs can be used to predict drug responses in vitro with reasonable accuracy. The authors have clearly put a great deal of time and thought into this project and should be pleased with the resulting paper.\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? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "4442",
"date": "22 Feb 2019",
"name": "Francisco Azuaje",
"role": "Author Response",
"response": "Response to Reviewer #1: Elizabeth A. Coker “In this paper, Azuaje et al., utilise transcriptional data from the Cancer Cell Line Encyclopedia to construct gene co-expression networks and identify hubs within them. They assess the biological relevance of the hub genes, and through the use of comparatively simple linear regression-based approaches, can use gene expression of these hubs to predict drug sensitivity in vitro. Despite not initially aiming to predict drug sensitivity, the authors’ models are able to perform comparably to previously published approaches. Azuaje et al., have performed extensive and appropriate validation of their work both in silico and in vitro, including assessing the inter-platform robustness of network hubs, cross-validation of the 47-hub signature, investigating the effect of network size on hub detection, and re-implementation of their own predictive approach to allow comparison with previously published drug sensitivity predictors. The paper is well-written and clear, with a thorough and precise account of the approaches used to assess the performance of the predictive model. The authors have produced a thoughtful and detailed account of their work which I enjoyed reading. The online Dr Paso resource is easy to use and has good documentation, although I have only tested it with the example datasets provided. I was unable to access the extended data via Zenodo as described in the manuscript (see comment below). My current recommendation is for “Approved with reservations”, as there are a number of points I feel should be addressed prior to approval:” Response: We thank the reviewer for her interest in our article and helpful feedback. “Furthermore, only four genes in our list of hubs are known drug targets: DKK1 (Irinotecan), MYB, SPARC and TUBB6 (the latter three targeted by paclitaxel).” I disagree with this statement. I have been unable to find any evidence of irinotecan targeting the protein DKK1, apart from an entry in DGIdb that states there is an interaction of type ‘n/a’ between DKKI and irinotecan, based solely upon a paper in Oncotarget that states DDK1 does not affect sensitivity to irinotecan in two cell lines. Equally, MYB and SPARC are not targets of paclitaxel, although they are shown in the DGIdb database as interacting with paclitaxel based on little or no evidence. I recommend the authors remove this sentence from the manuscript as it is based on a misinterpretation of DGIdb results and as such is misleading.” Response: We agree that this sentence is not accurate. As requested, we have adapted it as follows: We did not find validated evidence that our hub list contains known drug targets. Using DGIdb, we found potential associations between 4 hubs and 2 drugs: DKK1 (with Irinotecan), MYB, SPARC and TUBB6 (the latter three with paclitaxel). However, these associations cannot be interpreted as drug-target interactions and require further investigation.“I was able to access the code for the Dr Paso tool via Zenodo, but not the extended data associated with this study. Please ensure this is added and the appropriate link included.” Response: We confirm that the extended datasets are available in Zenodo, and they can be accessed via the web link included in reference 23. “Figure 1: Please update this figure to highlight the filtering stage between collecting transcriptomics data and building co-expression networks. Listing the number of genes present before and after filtering would also be informative.” Response: Change made, as requested.“Figure 2a: Here, the weight of edges is used to represent the expression correlations between nodes, but at present it is very difficult to see this in the examples highlighted. A larger, higher-resolution image of the network and examples is required.” Response: The figure has been modified to improve its clarity. The edges are represented with a white-to-grey gradient: the darker the edge, the higher the correlation.Although the edges connected to the non-hub node are expected to be more difficult to visualize because they are weaker, the modified figure now shows a better contrast between the hub and non-hub examples in terms of their corresponding edges. The figure caption has also been adapted to improve clarity: “(A) Snapshot of a (fully connected) weighted gene correlation network from untreated cell lines. Nodes and edges representing genes and their correlations respectively. Network hubs and non-hubs are colored in green and black respectively. Nodes are connected by edges, which are depicted in a white-to-grey gradient (the darker the edge, the higher the correlation). A zoom-in view of examples of hub and non-hub nodes reveals that the hub node has more edges with higher weights compared to the non-hub node.” “Figure 2b: I am not familiar with the concepts of semantic space and so do not know how to interpret this plot. Consider adding extra discussion in the main text or presenting this data in another way.” Response: We have expanded the caption of the figure with additional text to facilitate interpretation, as follows: “(B) Graphical summary of (non-redundant) Gene Ontology terms statistically over-represented in the list of 47 hub genes. Significant Biological Processes terms, represented as bubbles, are projected onto a scatterplot using REVIGO (33). Terms sharing common ancestors in the Gene Ontology database are close together; leading to a cluster of GO terms characterizing highly related biological annotations. To facilitate visualization, only a small selection of terms are labeled on the figure. Color and size indicates the term’s level of statistical enrichment in our list of hubs and frequency in the GO database respectively.” “Figure 2d: This figure is currently very difficult to read and could be enlarged if 2c is concurrently decreased in size.”Response: Changes made, as recommended. “Figure 3a: The density plot requires further annotation in relation to the clusters described in the manuscript, for example by arrows or circling the appropriate regions of the plot.”Response: Changes made, as requested. “Figure 4a: This figure is currently very difficult to read and could be enlarged.”Response: Figure has been enlarged, as requested. “In conclusion this is an interesting paper illustrating how co-expression hubs can be used to predict drug responses in vitro with reasonable accuracy. The authors have clearly put a great deal of time and thought into this project and should be pleased with the resulting paper.”"
}
]
},
{
"id": "42997",
"date": "15 Feb 2019",
"name": "Therese Commes",
"expertise": [
"Reviewer Expertise Transcriptomics",
"cancer",
"bioinformatics",
"RNA-seq"
],
"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 an interesting work for predicting cancer cells drug responses based on \"the gene expression profile of 47 genes representing significant hubs in pan cancer transcriptomes\" from untreated cells lines. The approach is original, as they did not select a specific set of genes from drug sensitivity experiments, but proposed \"an unbiased discovery approach, motivated by the hypothesis that co-expression network hubs encode useful information for investigating drug response in vitro\". Next, they investigated co-expression network hubs and drug responses and validated their approach using independant data sets including cell line-drug data. The study is well conceived and executed. The approaches used are suitable, and the description of work is adequately detailed. Data are clearly presented, and for the most part conclusions are reasonable. The method appears interesting, I think the paper would be a nice contribution that will be well-cited. Despite the difficulty of comparison, prediction performance of the model was compared to existing published methods and shows comparable results. Moreover the manuscript is well written. The authors need to clarify on the following points:\nFigure1: It will be better to specify the type of data (microarrays?) and source (CCLE?) in input for \"transcriptomic data from untreated cell lines\".\n\nMethods, page 4 :\n\nThe reason why removal of genes is needed is not clear:\n\"Although lasso or elastic net regularizations are also suitable approaches, they would have required the estimation of additional learning parameters and the removal of genes, which were deemed biologically interesting before model training.......\"\nMethods, page 5 (line 24) : The authors should specify what they mean by \"the original data\" which kind of files they used (raw data, gene counts ...). (line 28: RNS-seq instead of RNA-seq).\nResults, page 9: Could the authors specify if drugs with nearest clusters or CI values belong to a same drug \"class\" or not? Is there a relationship between performance prediction and drug classes?\nResults, page 9: This sentence is not clear: \"As the GDSC dataset shares \"different\" cell lines in common with CCLE...\"\nResults, page 12: \"However elastic net model required 613 genes as input features\": Does this set include the 47-hub model genes?\nDiscussion, page 14: Prediction performance and cancer types is discussed, what about \"hematopoietic and lymphoid tissues\"? This cancer type seems to strongly differ from other types in their drug responses and gene expression (see Figures 2 and 4). Could the authors comment on these data?\nDiscussion, page 16: \"Nevertheless future work could benefit from the incorporation of other \"omics\" data .. CCLE RNA-seq dataset could allow the analysis of transcript-level ( gene isoform)\". However RNA-seq technology has a larger potential than extracting transcript isoform and allows to extract genomic (mutation, indels, gene fusion, ...) and transcriptomic events (gene expression, splice variant, non-coding RNA), could the authors enlarge their comment about this potential and their prediction method? Also what about epigenetics data to predict drug sensitivity?\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": "4443",
"date": "22 Feb 2019",
"name": "Francisco Azuaje",
"role": "Author Response",
"response": "Response to Reviewer# 2: Therese Commes “The authors present an interesting work for predicting cancer cells drug responses based on \"the gene expression profile of 47 genes representing significant hubs in pan cancer transcriptomes\" from untreated cells lines. The approach is original, as they did not select a specific set of genes from drug sensitivity experiments, but proposed \"an unbiased discovery approach, motivated by the hypothesis that co-expression networkhubs encode useful information for investigating drug responsein vitro\". Next, they investigated co-expression network hubs and drug responses and validated their approach using independant data sets including cell line-drug data.The study is well conceived and executed. The approaches used are suitable, and the description of work is adequately detailed. Data are clearly presented, and for the most part conclusions are reasonable. The method appears interesting, I think the paper would be a nice contribution that will be well-cited. Despite the difficulty of comparison, prediction performance of the model was compared to existing published methods and shows comparable results. Moreover the manuscript is well written.The authors need to clarify on the following points:” Response: We thank the reviewer for her interest in our article and helpful feedback. \"Figure1:It will be better to specify the type of data (microarrays?) and source (CCLE?) in input for \"transcriptomic data from untreated cell lines\".\" Response: Figure 1 has been modified with the inclusion of the data type and the source as follows: “(CCLE microarray data)”. “Methods, page 4:The reason why removal of genes is needed is not clear: \"Although lasso or elastic net regularizations are also suitable approaches, they would have required the estimation of additional learning parameters and the removal of genes, which were deemed biologically interesting before model training.......\"” Response: We agree that, in general, the removal of genes is not required. However, note that we compared our 47-gene model with a Lasso model that consists of a reduced set of input genes. This gene selection process is embedded into the Lassso algorithm, which aims at selecting a subset of covariates with a good prediction ability. To do so, the algorithm constraints the sum of the absolute values of the coefficients. During this process, some coefficients will be set to zero and therefore are removed from the model. The elastic net, which can be seen as a compromise between Lasso and Ridge regression, also incorporates feature selection. “Methods, page 5 (line 24):The authors should specify what they mean by \"the original data\" which kind of files they used (raw data, gene counts ...). (line 28: RNS-seq instead of RNA-seq).” Response: In the main text, we have corrected “original data” with “RPKM data”. We have also corrected “RNS-Seq” with “RNA-Seq”. “Results, page 9:Could the authors specify if drugs with nearest clusters or CI values belong to a same drug \"class\" or not? Is there a relationship between performance prediction and drug classes?” Response: To address the reviewer’s request, the following sentence has been added in “Result” section (page 9): Interestingly, we observe that drugs belonging to the same drug class tend to cluster together according to their predicted (and observed) drug response values. For example, samples treated with cytotoxic drugs (e.g., Irinotecan and Topotecan) and kinase inhibitors (e.g., AZD6244 and RAF265) are closely located on the observed vs. predicted sensitivity plot (Supplementary Figure 5).“Results, page 9:This sentence is not clear: \"As the GDSC dataset shares \"different\" cell lines in common with CCLE...\"”Response: The sentence has been corrected: “As the GDSC dataset shares cell lines in common with CCLE...\"” “Results, page 12:\"However elastic net model required 613 genes as input features\": Does this set include the 47-hub model genes?” Response: There are no genes in common between these models. This is now clarified in the manuscript. Also, we note that there was a typo in this sentence: It should be 614 genes. We have corrected this typo in the main manuscript and have changed the caption of the Supplementary Figure 13. We have also changed the URL address that contains all supplementary files (reference 23 in the main manuscript). We include the following in the main manuscript: “However, the elastic net model required 614 genes as input features to achieve this performance (with no genes in common with the 47 hub genes).” “Discussion, page 14:Prediction performance and cancer types is discussed, what about \"hematopoietic and lymphoid tissues\"? This cancer type seems to strongly differ from other types in their drug responses and gene expression (see Figures 2 and 4). Could the authors comment on these data?” Response: To address the reviewer’s request, the following sentence has been added: Importantly, although the gene expression profiles of hematopoietic and lymphoid samples differ from all other samples, our 47-hub model was able to predict their responses with a relatively good accuracy (CI = 0.75). “Discussion, page 16:\"Nevertheless future work could benefit from the incorporation of other \"omics\" data. CCLE RNA-seq dataset could allow the analysis of transcript-level (gene isoform)\". However, RNA-seq technology has a larger potential than extracting transcript isoform and allows to extract genomic (mutation, indels, gene fusion, ...) and transcriptomic events (gene expression, splice variant, non-coding RNA), could the authors enlarge their comment about this potential and their prediction method? Also what about epigenetics data to predict drug sensitivity?” Response: To address the reviewer’s comment, the following text has been added: Furthermore, there are other opportunities to be investigated such as the analysis of genomic alterations, non-coding RNAs and epigenetic markers, which may enhance or complement existing models for predicting drug sensitivity."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1906
|
https://f1000research.com/articles/7-1740/v1
|
02 Nov 18
|
{
"type": "Research Article",
"title": "False signals induced by single-cell imputation",
"authors": [
"Tallulah S. Andrews",
"Martin Hemberg",
"Tallulah S. Andrews"
],
"abstract": "Background: Single-cell RNASeq is a powerful tool for measuring gene expression at the resolution of individual cells. A significant challenge in the analysis of this data is the large amount of zero values, representing either missing data or no expression. Several imputation approaches have been proposed to deal with this issue, but since these methods generally rely on structure inherent to the dataset under consideration they may not provide any additional information. Methods: We evaluated the risk of generating false positive or irreproducible results when imputing data with five different methods. We applied each method to a variety of simulated datasets as well as to permuted real single-cell RNASeq datasets and consider the number of false positive gene-gene correlations and differentially expressed genes. Using matched 10X Chromium and Smartseq2 data from the Tabula Muris database we examined the reproducibility of markers before and after imputation. Results: The extent of false-positive signals introduced by imputation varied considerably by method. Data smoothing based methods, MAGIC and knn-smooth, generated a very high number of false-positives in both real and simulated data. Model-based imputation methods typically generated fewer false-positives but this varied greatly depending on how well datasets conformed to the underlying model. Furthermore, only SAVER exhibited reproducibility comparable to unimputed data across matched data. Conclusions: Imputation of single-cell RNASeq data introduces circularity that can generate false-positive results. Thus, statistical tests applied to imputed data should be treated with care. Additional filtering by effect size can reduce but not fully eliminate these effects. Of the methods we considered, SAVER was the least likely to generate false or irreproducible results, thus should be favoured over alternatives if imputation is necessary.",
"keywords": [
"Gene expression",
"single-cell",
"RNA-seq",
"Imputation",
"Type 1 errors",
"Reproducibility"
],
"content": "Introduction\n\nSingle-cell RNASeq (scRNASeq) is a powerful technique for assaying the whole transcriptome at the resolution of individual cells. Although experimental protocols have evolved rapidly, there is still no strong consensus on how to best analyse the data. An important challenge to analysing scRNASeq data is the high frequency of zero values, often referred to as dropouts, and the overall high levels of noise due to the low amounts of input RNA obtained from individual cells. Recently there have been four methods published (Gong et al., 2018; Huang et al., 2018; Li & Li, 2018; van Dijk et al., 2018) which attempt to address these challenges though imputation, with several more under development (Deng et al., 2018; Mongia et al., 2018; Moussa & Mandoiu, 2018; Wagner et al., 2017).\n\nImputation is a common approach when dealing with sparse genomics data. A notable example has been the improvements to GWAS sensitivity and resolution when using haplotype information to impute unobserved SNPs (Visscher et al., 2017). Unlike scRNASeq data, this imputation employs an external reference dataset, often the 1000 Genomes project, to infer the missing values (Chou et al., 2016). Such a reference does not yet exist for scRNASeq data, and thus imputation methods can only use information internal to the dataset to be imputed. As a result there is a degree of circularity introduced into the dataset following imputation which could result in false positive results. Zero values in scRNASeq may arise due to low experimental sensitivity, e.g. sequencing sampling noise, technical dropouts during library preparation, or because biologically the gene is not expressed in the particular cell. Thus, one challenge when imputing expression values is to distinguish true zeros from missing values.\n\nMany imputation methods, such as SAVER (Huang et al., 2018), DrImpute (Gong et al., 2018) and scImpute (Li & Li, 2018), use models of the expected gene expression distribution to distinguish true biological zeros from zeros originating from technical noise. Because these gene expression distributions assume homogenous cell populations, they first identify clusters of similar cells to which an appropriate mixture model is fitted. Values falling above a given probability threshold to originate from technical effects are subsequently imputed. For example, scImpute models log-normalized expression values as a mixture of gamma-distributed dropouts and normally-distributed true observations. Alternatively some scRNASeq imputation methods perform data smoothing. In contrast to imputation, which only attempt to infer values of missing data, smoothing reduces noise present in observed values by using information from neighbouring data points. Both MAGIC (van Dijk et al., 2018) and knn-smooth (Wagner et al., 2017) perform data smoothing for single-cell data using each cell’s k nearest neighbours either through the application of diffusion models or weighted sums respectively.\n\nPrevious benchmarking of these imputation methods was based on positive controls, i.e. the ability to recover true signals within noisy data (Zhang & Zhang, 2018); the potential for false signals to be introduced into a dataset by these imputation methods was not considered, and it was concluded that most imputation methods provide a small improvement. We consider negative controls to evaluate the risks of introducing false positive when using imputation for single-cell datasets. Testing of the four published imputation methods, MAGIC, SAVER, scImpute, and DrImpute and one currently unpublished method, knn smooth, revealed that all methods can introduce false positive signals into data. While some methods, performed well on simulated data, permuting real scRNASeq data revealed high variability in performance on different datasets. We show that statistical tests applied to imputed data should be treated with care, and that results found in imputed data may not be reproducible.\n\n\nMethods\n\nFive different single-cell RNASeq imputation methods were tested: SAVER (Huang et al., 2018), DrImpute (Gong et al., 2018), scImpute (Li & Li, 2018), MAGIC (van Dijk et al., 2018) and knn-smooth (Wagner et al., 2017). Unless specified otherwise these were run with default parameters (Table 1). Each method was applied to either the raw-counts or log2 counts per million normalized data, as calculated scater (McCarthy et al., 2017), as appropriate.\n\nAs an initial test of imputation methods and to understand the effect of various method-specific parameters on imputation we simulated data from a negative binomial model. Expression matrices containing 1000 cells, equally spread across two cell-types, and 500 genes, with mean expression ranging from 10-3-104, were simulated. Half of the genes were differentially expressed (DE) by an order of magnitude between the two cell-types, half were drawn independently. Ten such expression matrices were independently simulated. Each imputation method was run on each replicate with a range of parameter values (Table 1). Significant gene-gene correlations were identified using Spearman correlation with a conservative Bonferroni multiple testing correction to avoid distributional assumptions on the imputed values. Correlations involving not DE genes or in the incorrect direction were considered false positives.\n\nSplatter (Zappia et al., 2017) was used to generate 60 simulated single-cell RNASeq count matrices using different combinations of parameters (Table 2). Each simulated dataset contained 1,000 cells split into 2–10 groups and 1,000–5,000 genes of which 1–30% were differentially expressed across the groups. For simplicity all groups were equally sized and were equally different from one another. Half the simulations assumed discrete differentiated groups, whereas the other half used the continuous differentiation path model. We also considered the effect of four different amounts of added dropouts plus the no-added dropout model.\n\n*Randomly selected for each possible combination of the other four parameters.\n\nAccuracy of each imputation method was evaluated by testing for differentially expressed (DE) genes between the groups used to simulate the data. To avoid issues of different imputation methods resulting in data best approximated by different probability distribution, we employed the non-parametric Kruskal-Wallis test (Kruskal & Wallis, 1952) with a 5% FDR to identify significant DE genes. Since this test is relatively low-power it is likely to underestimate the number of DE genes compared to alternatives. When filtering DE genes by effect size, in addition to significance, we used the maximum log2-fold-change across all pairs of clusters.\n\nSix 10X Chromium and 12 Smartseq2 datasets were chosen from the Tabula Muris (The Tabula Muris Consortium et al., 2017) consortium data such that: i) there were at least two cell types containing >5% of the total cells and ii) there were between 500–5,000 cells after filtering (Table S1). Each dataset was preprocessed to remove cell-types accounting for <5% of total cells, and any cells not assigned to a named cell-type. Genes were filtered to remove those detected in fewer than 5% of cells.\n\nWe selected the two most similar cell-types in each dataset using the Euclidean distance between their mean expression profiles. Differential expression of each gene between these cell-types was evaluated using a Mann-Whitney-U test on the log2-normalized counts. Genes with a raw p-value > 0.2 were then permuted across the selected cell-types to eliminate any residual biological signals. Permuted raw counts were obtained by de-logging and de-normalizing the permuted log2-normalized expression to avoid library-size confounders.\n\nEach imputation method was applied to the full dataset after permutation using default parameters (Table 1). False-positives introduced by each imputation was assessed by applying the Mann-Whitney-U test to test for differential expression between the two chosen cell-types. A Bonferroni multiple-testing correction was applied to ensure a consistent level of expected total false positives of less than 1.\n\nWe utilized the six tissues for which there exists matching Smart-seq2 and 10X Chromium data from the Tabula Muris (The Tabula Muris Consortium et al., 2017) to evaluate the reproducibility of imputation results. These datasets were filtered as described above, and any cell-types not present in both pairs of the matching datasets were excluded. Each imputation method was applied to the datasets without any permutation.\n\nMarker genes were identified in each imputed dataset using a Mann-Whitney-U test to compare each cell-type against all others, and effect size was calculated as the area under the ROC curve for predicting each cell-type from the others (Kiselev et al., 2017). Genes were assigned to the cell-type for which they had the highest AUC. Significant marker genes were defined for each imputed dataset using a 5% FDR and an AUC over a particular threshold. Reproducibility was evaluated by determining the number of genes that were significant markers in both of a matching pair of datasets and were markers of the same cell-type.\n\n\nResults\n\nWe tested three published imputation methods, SAVER (Huang et al., 2018), scImpute (Li & Li, 2018) and DrImpute (Gong et al., 2018), as well as two data-smoothing methods MAGIC (van Dijk et al., 2018) and knn-smooth (Wagner et al., 2017). We applied each method with the default parameter values (Table 1) to data simulated from a simple negative binomial, since technical noise in scRNASeq data has been observed to follow a negative binomial distribution (Grün et al., 2014). Half the simulated genes were differentially expressed (DE), thus highly correlated with each other, the rest were drawn completely independently. These simulations did not include different library-sizes, batch effects, or zero-inflation to eliminate all possible sources of false-signals that imputation method might mistake for true biology. Thus, these simulations represent the simplest most straightforward case with no technical confounders. Only SAVER strengthened the correlations between lowly expressed DE genes without generating false positive gene-gene correlations between independently drawn genes (Figure 1A). Since SAVER models expression data using a negative binomial, it is expected to perform well on this simulated data. MAGIC generated very strong false positive correlations (r > 0.75) at all expression levels, whereas DrImpute, which only imputes zero values, created false positive correlations mostly among lowly expressed genes. Knn-smooth and scImpute produced a few false-positive correlations among moderately-expressed genes using default parameters.\n\n(A) Gene-gene correlations before and after imputation using suggested parameter values: SAVER (all genes), MAGIC (k=12, t=3), knn (k=50), scImpute (threshold=0.5), DrImpute (remaining zeros=0). Coloured bars indicate genes highly expressed (red) or lowly expressed (blue) in one cell population vs the other, or genes not differentially expressed between the populations (grey). Genes are ordered left to right by DE direction then by expression level (high to low). (B) False positive gene-gene correlations as imputation parameters are changed. Dashed lines are 95% CIs based on 10 replicates. See Figure S1 for true positive rate of gene-gene correlations across the same parameters.\n\nChoice of parameter values has a large influence imputation results (Figure 1B). Four of the imputation methods required the user to set at least one parameter a priori, only SAVER did not. We varied the thresholds scImpute and DrImpute use to determine which zeros to impute. For scImpute some of the lower and moderate expression values were imputed even at a very strict probability threshold (p > 0.8), but changing the threshold had little effect on the imputation. As expected for DrImpute, imputing a greater proportion of zeros generated more false positives. Knn-smooth and MAGIC both perform data smoothing using a k-nearest-neighbour graphs between cells. Increasing the number of nearest-neighbours (k) produces smoother data and more false-positive correlations (Figure 1B). MAGIC provides a default value for k but no indication of how this parameter should be adjusted for different sized datasets, whereas knn smooth provided no default value but a rough suggestion to scale the value depending on the total number of cells. MAGIC also utilizes a second parameter, time (t), for the diffusion process acting on the graph which by default is algorithmically estimated for the dataset. Longer diffusion times produce smoother data and more false positives.\n\nThese simple simulations contained only two cell-types and no technical confounders such as library-size or inflated dropout rates that are observed in some scRNASeq datasets. For a more comprehensive evaluation of imputation methods we simulated data using Splatter (Zappia et al., 2017). We simulated data with 1,000 cells split into 2–10 groups and 1,000–5,000 genes of which 1–30% were differentially expressed across the groups. We considered four different levels of zero inflation and no zero inflation (Table 2). Each simulated dataset was imputed with each method using the default parameters (Table 1). To score each imputation we considered the accuracy of identifying differentially expressed genes between the groups using the non-parametric Kruskal-Wallis test (Kruskal & Wallis, 1952).\n\nNone of the imputation methods significantly outperformed the others or the unimputed data based on the sensitivity and specificity. While both knn-smooth and MAGIC have increased sensitivity they have very low specificity, whereas SAVER and scImpute are very similar to the un-imputed data with high specificity but relatively low sensitivity (Figure 2A & B). DrImpute was in between the two extremes with somewhat higher sensitivity and lower specificity than SAVER and scImpute. Both scImpute and DrImpute are designed specifically to only impute excess zeros but neither showed a clear improvement over the raw counts when the simulations contained various levels of zero inflation (Figure 2C). However, all methods except SAVER readily introduced false-positive signals, as demonstrated by a drop in specificity, when 30% of genes were DE (Figure 2D). We hypothesize that slight biases due to library-size normalization in the presence of strong biological differences, may be amplified by the imputation methods since we also observe a significant but smaller drop in specificity for the normalized but un-imputed data. Biases due to counts-per-million library-size normalization in the presence of strong DE are a known issue from bulk RNASeq analysis (Bullard et al., 2010).\n\n(A & B) Different imputation methods choose a different trade-off between sensitivity and specificity. (C) Zero inflation decreases sensitivity of DE which most imputation methods fail to correct. (D) Strong true signals (high proportion of genes DE) decreases specificity particularly for data-smoothing methods.\n\nIt’s possible that the bulk of false-positives generated by imputation methods result from small biases or sampling noise being amplified to reach statistical significance. If this is true, then filtering DE genes by magnitude in addition to significance should restore the specificity of such tests on imputed data. We observed this to be the case when an additional threshold was set based on the Xth percentile highest log2 fold-change across the whole dataset (Figure 3). However, sensitivity also declined as the fold-change threshold was made more stringent. This suggests the fundamental trade-off between sensitivity and specificity cannot be overcome by imputation.\n\nSensitivity (green) and specificity (blue) of each imputation method applied to the splatter-simulated data, when restricting to only the top X% of genes by fold-change. Dashed lines indicate 95% CI.\n\nSplatter is a widely used simulation framework for scRNASeq but may not fully capture the complexities of real scRNASeq data. To test the performance of each imputation method on real scRNASeq data we selected 12 tissues from the Tabula Muris database (The Tabula Muris Consortium et al., 2017) and applied the imputation methods to the Smartseq2 and 10X data separately. Since the ground truth is not known for these data, we selected two cell-types from each dataset and permuted the expression of those genes that were not differentially expressed between them (p > 0.2) to generate a set of genes that we could confidently consider as being not differentially expressed (Methods). Using these as ground truth we could estimate the number of false positive differentially expressed genes introduced by each imputation method. Strikingly, we observed a very high variability between datasets which appears to be unrelated to the experimental platform (Figure4 A & B). MAGIC, scImpute and knn-smooth consistently produced large numbers of false positives (40–80%). Whereas, DrImpute and SAVER were extremely variable producing few to no false positives in some datasets and over 90% false positives in others.\n\n(A) SmartSeq2 datasets, (B) 10X Chromium datasets. Non-differentially expressed genes were permuted prior to imputation. (C) Reproducibility of marker genes before and after imputation. Number of genes that were significant in both 10X and SmartSeq2 data (AUC > 0.75, q.value < 0.05) and in brackets the proportion that were markers of the same cell-type in both datasets.\n\nTo complement the false positives in the permuted data, we considered whether imputed signals in the original datasets were reproducible in both 10X and Smart-seq2 data. We identified marker genes using a Mann-Whitney-U test, comparing one cell-type to the others in that tissue. Markers were filtered by significance (5% FDR) and magnitude (AUC > 0.75). Each marker was assigned to the cell-type for which it had the highest AUC. Reproducibility was measured as the number of markers that were significant in both datasets and the proportion of those that were markers for the same cell-type. All of the imputation methods increased the number of significant markers (Figure 4C). However, many of these were assigned to contradictory cell-types. Without imputation, 95% of genes that were significant markers in both datasets were highly expressed in the same cell-type. After imputation, this dropped to only 80–90%, except when SAVER was used for the imputation, suggesting that many of the imputed markers are incorrect. Decreasing the magnitude threshold of the markers leads to even more contradictory results in the imputed datasets (Figure 5). While unimputed data retained >90% concordance in cell-type assignments of significant markers regardless of the AUC threshold, this fell to 60–80% in imputed data when a low AUC threshold is used.\n\nMarkers were identified in 10X Chromium and Smartseq2 data for mouse muscle. The number of markers (bars, left axis) and proportion reproducible across both datasets (line, right axis) are plotted. Only significant markers (5% FDR) exceeding the AUC threshold were considered.\n\nThe imputation methods produced different distortions of the gene expression values (Figure 6). When applied to the permuted pancreas data, SAVER made only slight adjustments to the gene expression values. scImpute compressed the gene expression distributions into a more gaussian shape. DrImpute shifted zero values up into the higher mode of the distribution if present. In contrast, MAGIC and knn-smooth tended to generate bimodal expression distributions. The tendency towards bimodality could be problematic for downstream analysis since many methods, e.g. PCA and differential expression, assume either negative binomial or gaussian distributions. Many of these genes were differentially expressed after imputation, despite being permuted previously. Interestingly, the direction of differential expression was not always consistent across imputation methods, for instance Zfp606 was more highly expressed in PP cells than A cells after imputation using MAGIC but the inverse was true after imputing with knn-smooth.\n\nUnimputed indicates the permuted normalized log-transformed expression. Red = PP cell, Blue = A cell. * = p < 0.05, ** = significant after Bonferroni correction.\n\n\nDiscussion\n\nWe have shown that imputation for scRNASeq data may introduce false-positive results when no signal is present. On simulated data all the methods except SAVER generated some degree of false positives (Figure 1 & Figure 2). We find a fundamental trade-off between sensitivity and specificity which imputation cannot overcome (Figure 2 & Figure 3). On permuted real data, imputation results were more variable (Figure 4), and even SAVER generated large numbers of false positives in some datasets. Imputation also reduced the reproducibility of marker genes, unless strict magnitude thresholds were imposed (Figure 4 & Figure 5). In addition to false-positives, distortions in expression distributions (Figure 6) may cause imputed data to violate assumptions of some statistical tests.\n\nWe found a trade-off between sensitivity and specificity across methods (Figure 2). MAGIC and knn-smooth which are data-smoothing methods, as such they adjust all expression values not just zeros. Since they impose larger alterations on the data, these methods generate many more false positives than methods which only impute zero values. However, they also have a greater sensitivity. In contrast, model-based methods which only impute low expression values, generated fewer false positives but had minimal improvements to sensitivity.\n\nThis trade-off between sensitivity and specificity also emerges if one employs an effect size threshold to reduce false-positives generated by imputation (Figure 3 & Figure 5). While using a strict effect size threshold can recover a reasonable specificity for the data-smoothing methods, doing so eliminates the improvements to sensitivity. Likewise, adding an effect size threshold can preserve reproducibility of imputation results but doing so largely eliminates the benefit in terms of number of markers identified.\n\nThese trade-offs reflect the fundamental limitation of single-cell RNASeq imputation, namely that it can only use the information present in the original data. While imputation in other fields often uses external references or relationships for the imputation, scRNASeq imputation only draws on structure within the dataset itself. Hence no new information is gained, making it analogous to simply lowering the significance threshold of any statistical test applied to the data (Fawcett, 2006). However, imputation based on external reference datasets may become possible as various cell-type atlases are completed; however these will be limited to those species and tissues that have been systematically catalogued (Han et al., 2018; Rozenblatt-Rosen et al., 2017; The Tabula Muris Consortium et al., 2017; Zeisel et al., 2018). Alternatively, models could be developed to use gene-gene correlations derived from large external databases of expression data (Obayashi et al., 2008), while more generalizable such methods may not capture cell-type specific relationships.\n\nOf the methods we tested, SAVER was the least likely to generate false-positives, but it’s performance depended on how well data conformed to the negative binomial model it is based on. If imputation is used, combining SAVER with an effect size threshold is the best option to avoid irreproducible results. Alternatively, verifying the reproducibility of results across multiple datasets or multiple imputation methods can eliminate some false positives. However, our results highlight that statistical tests applied to imputed data should be treated with care. Although a previous benchmarking study showed good results for positive controls, our study highlights the importance of considering negative controls when evaluating imputation methods.\n\n\nData and software availability\n\nTabula Muris data\n\nSmartseq2 https://doi.org/10.6084/m9.figshare.5715040.v1 (Consortium, The Tabula Muris, 2017a).\n\n10X Chromium https://doi.org/10.6084/m9.figshare.5715040.v1 (Consortium, The Tabula Muris, 2017b).\n\nR packages\n\nMAGIC: Rmagic (v0.1.0) https://github.com/KrishnaswamyLab/MAGIC\n\nDrImpute: DrImpute (v1.0) https://github.com/ikwak2/DrImpute\n\nscImpute: scImpute(v0.0.8) https://github.com/Vivianstats/scImpute\n\nSAVER: SAVER(v1.0.0) https://github.com/mohuangx/SAVER\n\nKnn-smooth: knn_smooth.R (Version 2) https://github.com/yanailab/knn-smoothing\n\nScater: scater(v1.6.3) : https://www.bioconductor.org/packages/release/bioc/html/scater.html\n\nSplatter: splatter(v1.2.2) : https://bioconductor.org/packages/release/bioc/html/splatter.html\n\nPermute: permute(v0.9-4) : https://cran.r-project.org/web/packages/permute/index.html",
"appendix": "Grant information\n\nFunding was provided by the Wellcome Trust Sanger Institute Core Funding.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1: File containing Table S1 (Tabula Muris permuted datasets) and Figure S1 (Sensitivity single-cell imputation methods).\n\nClick here to access the data\n\n\nReferences\n\nBullard JH, Purdom E, Hansen KD, et al.: Evaluation of statistical methods for normalization and differential expression in mRNA-Seq experiments. BMC Bioinformatics. 2010; 11: 94. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChou WC, Zheng HF, Cheng CH, et al.: A combined reference panel from the 1000 Genomes and UK10K projects improved rare variant imputation in European and Chinese samples. Sci Rep. 2016; 6: 39313. PubMed Abstract | Publisher Full Text | Free Full Text\n\nConsortium, The Tabula Muris: Single-cell RNA-seq data from Smart-seq2 sequencing of FACS sorted cells. figshare. Fileset. 2017a.\n\nConsortium, The Tabula Muris: Single-cell RNA-seq data from microfluidic emulsion. figshare. Fileset. 2017b.\n\nDeng Y, Bao F, Dai Q, et al.: Massive single-cell RNA-seq analysis and imputation via deep learning. bioRxiv. 2018. Publisher Full Text\n\nFawcett T: An introduction to ROC analysis. Pattern Recognit Lett. 2006; 27(8): 861–874. Publisher Full Text\n\nGong W, Kwak IY, Pota P, et al.: DrImpute: imputing dropout events in single cell RNA sequencing data. BMC Bioinformatics. 2018; 19(1): 220. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrün D, Kester L, van Oudenaarden A: Validation of noise models for single-cell transcriptomics. Nat Methods. 2014; 11(6): 637–640. PubMed Abstract | Publisher Full Text\n\nHan X, Wang R, Zhou Y, et al.: Mapping the Mouse Cell Atlas by Microwell-Seq. Cell. 2018; 173(5): 1307. PubMed Abstract | Publisher Full Text\n\nHuang M, Wang J, Torre E, et al.: SAVER: gene expression recovery for single-cell RNA sequencing. Nat Methods. 2018; 15(7): 539–542. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKiselev VY, Kirschner K, Schaub MT, et al.: SC3: consensus clustering of single-cell RNA-seq data. Nat Methods. 2017; 14(5): 483–486. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKruskal WH, Wallis WA: Use of Ranks in One-Criterion Variance Analysis. J Am Stat Assoc. 1952; 47(260): 583–621. Publisher Full Text\n\nLi WV, Li JJ: An accurate and robust imputation method scImpute for single-cell RNA-seq data. Nat Commun. 2018; 9(1): 997. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcCarthy DJ, Campbell KR, Lun AT, et al.: Scater: pre-processing, quality control, normalization and visualization of single-cell RNA-seq data in R. Bioinformatics. 2017; 33(8): 1179–1186. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMongia A, Sengupta D, Majumdar A: McImpute: Matrix completion based imputation for single cell RNA-seq data. bioRxiv. 2018. Publisher Full Text\n\nMoussa M, Mandoiu I: Locality Sensitive Imputation for Single-Cell RNA-Seq Data. bioRxiv. 2018. Publisher Full Text\n\nObayashi T, Hayashi S, Shibaoka M, et al.: COXPRESdb: a database of coexpressed gene networks in mammals. Nucleic Acids Res. 2008; 36(Database issue): D77–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRozenblatt-Rosen O, Stubbington MJT, Regev A, et al.: The Human Cell Atlas: from vision to reality. Nature. 2017; 550(7677): 451–453. PubMed Abstract | Publisher Full Text\n\nThe Tabula Muris Consortium, Quake SR, Wyss-Coray T, et al.: Transcriptomic characterization of 20 organs and tissues from mouse at single cell resolution creates a Tabula Muris. bioRxiv. 2017. Publisher Full Text\n\nvan Dijk D, Sharma R, Nainys J, et al.: Recovering Gene Interactions from Single-Cell Data Using Data Diffusion. Cell. 2018; 174(3): 716–729.e27. PubMed Abstract | Publisher Full Text\n\nVisscher PM, Wray NR, Zhang Q, et al.: 10 years of GWAS discovery: biology, function, and translation. Am J Hum Genet. 2017; 101(1): 5–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWagner F, Yan Y, Yanai I: K-nearest neighbor smoothing for high-throughput single-cell RNA-Seq data. bioRxiv. 2017. Publisher Full Text\n\nZappia L, Phipson B, Oshlack A: Splatter: simulation of single-cell RNA sequencing data. Genome Biol. 2017; 18(1): 174. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZeisel A, Hochgerner H, Lönnerberg P, et al.: Molecular Architecture of the Mouse Nervous System. Cell. 2018; 174(4): 999–1014.e22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang L, Zhang S: Comparison of computational methods for imputing single-cell RNA-sequencing data. IEEE/ACM Trans Comput Biol Bioinform. 2018. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "40895",
"date": "29 Nov 2018",
"name": "Simone Tiberi",
"expertise": [
"Reviewer Expertise Statistics",
"Bioinformatics",
"Transcriptomics",
"(single cell) RNA-seq",
"Biostatistics",
"Systems 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 article investigates how imputation methods of 0 counts in single-cell RNA-seq (scRNA-seq) can introduce false signals, and hence false positives in downstream analyses. The authors explain how scRNA-seq data can present an excess of 0 counts, i.e. dropouts, due to technical artefacts, and introduce a few recent methods that can be used to impute these values. Andrews and Hemberg focus on a sub-set of 5 imputation methods and investigate, in three scenarios, if these methods introduce false signals between genes:\nFirst, data are simulated from a simple negative binomial (NB) model: most imputation methods introduce false signals in the data by increasing the correlation between independent genes.\n\nSecondly, the authors study the effect of imputation methods on downstream differential gene expression (DGE) analyses on 60 scRNA-seq datasets simulated via Splatter (with varying degrees of dropouts and DGE). They find that, compared to the original un-imputed data, albeit some imputation methods result in higher Sensitivity (i.e. true positive rate), all of them have lower the Specificity (i.e. true negative rate).\n\nThirdly, they consider several real scRNA-seq datasets, where counts are permuted to obtain approximately uncorrelated genes, and investigate how imputation methods affect the ability to identify marker genes. The authors find that, compared to the un-imputed data, imputation tools distort expression patterns and increase the number of identified marker genes, although some of these are likely to be false detections.\nThe article treats a relevant problem and provides a comprehensive benchmarking of imputation methods. Overall, the manuscript is clear and its scientific quality is adequate. Below, I suggest several corrections (and identify a few typos) that hopefully will contribute to improving the quality and clarity of the work.\n\nMajor Comments:\nIn some cases it is unclear to me why you take certain decisions: I feel you should motivate more your choices (see Minor comments for specific examples).\n\nPlease provide source code to reproduce all the analysis you present (including obtaining the simulated and permuted data).\n\nThere is some redundancy in the description of the data: you first describe in detail how you obtained the simulated data and the permuted real data in the Methods section, and then you repeat it again (although with fewer sentences) in the Results section. I would avoid or shorten the second description in the Results section.\n\nAlthough the paper aims at investigating on false signals introduced by imputation methods, I feel too much emphasis has been given to false positive results as opposed to jointly considering false and true positive results. Indeed, the paper shows that imputation methods result in increased FPs/Specificity, particularly when the original data are not affected by dropouts, but it only marginally focuses on the increase in TPs/Sensitivity.\nMore informally, I think you should try to show both sides of the coin and avoid (over-)interpreting FP results alone. In this regard, to get a joint picture of Sensitivity and Specificity, I think you should provide (at least for the Splatter simulation) ROC and FDR curves (eventually, also as Supplementary figures). Since you perform 60 simulations from Splatter, you might consider global ROC and FDR plots based on the results from all simulations.\n\nI think that the limitations of the study should be explained more clearly:\n5.1) In the permuted real data analysis, all imputation methods find many more marker genes than the un-imputed data, but the authors mostly focus on the fact that the percentage of “reliable” identifications decreases. I think that: 1) importance should be given also to the fact that many more “reliable” marker genes are identified (also referring to the comment above about FPs and TPs) and: 2) it is essential to explicitly acknowledge that the true state of marker genes is unknown. Importantly, in Figure 4 A) and B) please add the FPR obtained on the un-imputed data to provide a baseline comparison.\n5.2) In the NB simulation you don’t simulate any dropouts, which represents the worst case scenario for imputation methods. In this context, I would expect all imputation methods to worsen downstream results, because there are no dropouts to impute at all. I think you should mention this more explicitly.\n\nIn Splatter simulations you “considered the effect of four different amounts of added dropouts”. How mild or extreme were these dropout levels compared to real scRNA-seq data? I would expect imputation methods to improve the quality of the data as the number of dropouts increases. Did you try to consider more “extreme” dropout rates?\n\nIn Figures 2C and 2D you provide Sensitivity boxplots stratified by dropout rates and Specificity boxplots stratified by DE. Sensitivity and Specificity should always be examined jointly: for both stratification cases, please provide both Sensitivity and Specificity plots (eventually, also as Supplementary figures).\n\nI suggest another round of polish to improve writing and clarity in some parts of the paper. In particular: adding few commas would facilitate the reading in long sentences; past and present tenses are sometimes mixed; some sentences seem a bit out of place and could be better integrated in the flow; I found the last two paragraphs of the Results section a bit hard to follow.\n\nYou refer a few times to the fact that you “find a fundamental trade-off between sensitivity and specificity which imputation cannot overcome”: reading the paper it seems that imputation methods might be responsible for this. But this trade-off is due to the nature of Sensitivity and Specificity; indeed, Sensitivity and Specificity are positively correlated by construction: as one moves the significance threshold, both will increase or decrease. Clearly an ideal method will have Sensitivity 0 and Specificity 1. I think you should remove or edit the sentences referring to this trade-off (particularly in the Discussion) to clarify that imputation methods are not the cause of this trade-off.\n\nIn the Discussion you say that “While imputation in other fields often uses external references or relationships for the imputation, scRNASeq imputation only draws on structure within the dataset itself.”. Actually, “canonical” imputation methods do not require an external reference and only use the available data. While having an additional reference can increase the information at disposal and hence, potentially, improve the accuracy of imputation tools, I don’t think this is the main reason why they result in increased false signals. Besides, there are other issues with using an external reference; e.g. if the reference is not “similar” to the data-set under study, particularly concerning their dropouts. I think you could clarify that using an external reference is one of the possible ways to improve imputation methods, but keeping in mind that imputation (in general) can also work without a reference.\n\nMinor Comments:\n1) General:\nThroughout the text, you use both “Smart-seq2” and “Smartseq2“; I suggest you use only one, for consistency.\n2) Abstract:\n“since these methods generally rely on structure inherent to the dataset under consideration they may not provide any additional information.” You clarify this point later in the text but, when I read the abstract, it was not clear to me what you were referring to. Maybe you could try to be more explicit here or remove the sentence.\n3) Introduction:\nYou cite 4 imputation methods as “under development“ but you only test one. I think you’d motivate this choice.\n\nTypo: “though imputation” -> “through imputation”.\n\nGWAS not defined before.\n\nTypo: “imputation, which only attempt to infer” -> “imputation, which only attempts to infer”.\n4) Methods:\nFig S1: “aka” -> “i.e.” (I would use something more elegant than aka).\n\nTypo: “as calculated scater” -> “as calculated by scater” ?\n\n“ranging from 10^3-10^4” -> “ranging 10^3-10^4” or “ranging from 10^3 to 10^4”.\n\nTypo: “different probability distribution” -> “different probability distributions”.\n\n“When filtering DE genes by effect size, in addition to significance”. This sentence is quite vague, please be more specific.\n\n“Six 10X Chromium and 12 Smartseq2 datasets”. You use words (Six) and digits (12) in the same sentence to refer to numbers: I’d choose one for consistency.\n\nYou use two distinct types of DGE tests for the simulated data (Splatter) and the permuted real data. Please motivate your choice.\n\nTypo (?): “for which there exists matching Smart-seq2 and 10X Chromium” -> “for which there exists matching for Smart-seq2 and 10X Chromium” ?\n5) Results:\n“MAGIC provides … whereas knn smooth provided …”. Present and past tenses are mixed here: I suggest you replace “provided” with “provides” to keep consistency with the rest of the manuscript.\n\nIn the NB simulation, provide more details on the implementation of the correlation test: how did you test correlations? What significance level was used to define a significant correlation in Fig 1B and S1? 0.05?\n\nIn Figure 1B and S1, I guess that “Raw” refers to the original (un-imputed) data; did I understand correctly? It was not obvious to me at a first glance, please make it explicit (in the text or in the Figure caption).\n\nFig 2: typo (?): “Different imputation methods choose a different trade-off ...“; I didn’t understand the use of “choose” in the sentence: is this a typo? If not, can you re-write the sentence in a clearer way?\n\nFig 2: “genes DE” -> “DE genes”.\n\nIn the permutation real data analysis, please clarify the concept of filtering genes: do you refer to independent filtering of genes (based on their estimated FC)?\n\nTypo: “the bulk of false-positives ... result” -> “the bulk of false-positives … results”.\n\n“It’s possible” -> “It is possible”.\n\n“Xth percentile” -> “X-th percentile”.\n\n“Xth percentile highest log2 fold-change“ -> “highest log2 fold-change X-th percentile”.\n\nFig 4 (A) caption: “SmartSeq2 datasets,” -> “SmartSeq2 datasets.” (a comma separates two Figure descriptions instead of a full stop).\n\nFig 4 (C) caption: “the proportion that were markers” -> “the proportions that were markers”.\n\nI would change “many of the imputed markers are incorrect” to “some of the imputed markers are incorrect”. “some” seems more appropriate than “many”, considering that 80-90% of them are estimated to be true marker genes.\n\nThe second last paragraph of Results sounded a bit contorted to me: I would rephrase it in a clearer way.\n\n“The imputation methods produced different distortions of the gene expression values (Figure 6).” Can you better integrate this sentence in the flow? It seems a bit out of place.\n\n“PCA and differential expression” -> “PCA and most differential expression tools/methods”. Tools/methods is missing. I would also add “most” because not all DE methods require NB or Gaussian distributions (e.g. non parametric methods).\n\nTo facilitate a visual comparison, in Figure 5 I would adjust the left y-axis (Genes #) to have the same limits in all examples.\n\nFig 6 caption: “significant after Bonferroni correction”; please add the significance level (I assume 0.05).\n6) Discussion:\nIn the second paragraph you first use “these methods generate” and then “MAGIC generated” mixing present and past tenses; I’d use “generate” in both cases.\n\nThe subject is missing in this sentence: “MAGIC and knn-smooth which are data-smoothing methods, as such they adjust all expression values not just zeros.” -> I would write something like: “MAGIC and knn-smooth are data-smoothing methods, as such they adjust all expression values not just zeros.” Or alternatively, “MAGIC and knn-smooth, which are data-smoothing methods, adjust all expression values not just zeros.”\n\n“it’s performance” -> “its performance”.\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? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "4430",
"date": "19 Feb 2019",
"name": "Tallulah Andrews",
"role": "Author Response",
"response": "Thank you for the helpful suggestions, we have made all suggested Minor corrections and have addressed the Major corrections here and in the revised version of the text:1. We have revised the text and tried to provide motivations for the key decisions.2. The github repo accompanying the study now contains scripts that can be run to reproduce the results reported here.3. We have followed the Reviewer’s suggestion and shortened the descriptions in the Results section.4. We have added the true positive rate to Figure 1B, have added Figure S4 showing the increase in reproducible markers in Tabula Muris datasets, and modified the text to put greater emphasis of the increase in sensitivity provided by imputation (last paragraph of page 6, p9 paragraph 2, p14 paragraph 1) to clarify that sensitivity is increased by using imputation at the cost of specificity, however as the ROC plots show (Figure 2 E), to address the reviewer’s concern below, this increase in sensitivity could be achieved by simply lowing the significance threshold applied to the statistical test and result in fewer false positives than using an imputation method.The reviewer raises a good point and we have calculated and included the ROC for the simulated data in Figure 2 E.5.1. We have updated the text to highlight the advantage of having a larger number of markers from imputed data (Figure S4, p14 paragraph 1). We have added the FPRs for the un-imputed data (counts) to both Fig 4A and B, as expected there were almost none since we used the conservative Bonferroni multiple testing correction. 5.2. We have highlighted the lack of dropouts in the NB simulations in the text, and explicitly mentioned the desired behaviour for both model-based imputation and data-smoothing in this context (Methods: Negative Binomial Simulations).6. We have adjusted the dropout parameters tested to be more similar to those observed in real single-cell RNA-seq data (Figure S1 A) and added the average proportion of zeros in the entire expression matrix for each value to Table 2 to help the readers understand what the different scenarios correspond to. At the highest level of added dropouts 94% of the matrix was composed on zeros and all the methods other than MAGIC and knn-smooth had sensitivity < 0.2, and the resilience of data-smoothing to high dropout rates has been noted in the text (Results: p9, paragraph 1).7. We have followed the Reviewer’s suggestion and now include both Sensitivity and Specificity plots stratified by dropout rate and proportion of DE genes in Figure 2.8. We have tried to improve the clarity of the text with a specific focus on paragraphs highlighted by the Reviewer.9. The Reviewer raises an important point regarding the fundamental relationship between sensitivity and specificity. One of the central aims of our paper was to highlight this particular trade-off and that the effect of most imputation methods is simply to shift the balance between these quantities. Our goal was to say that this is indeed a relationship that is caused by how these quantities are constructed and that imputation methods simply favour one side of the trade off or the other not create it. We have edited the text to better clarify this (Discussion: paragraph 1).10. The Reviewer raises a good point, we have edited the text appropriately (Discussion: paragraph 3)."
}
]
},
{
"id": "40894",
"date": "03 Dec 2018",
"name": "Jean Fan",
"expertise": [
"Reviewer Expertise single-cell methods development",
"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\nOverview:\nAnalysis of single-cell RNA-seq data is often complicated by large amounts of zeros, of some which represent true lack of expression, while others are reflective of poor capture efficiency or other technical limitations. Several methods have been developed to impute the zeros and recover the true gene expression values. Here, Andrews and Hemberg compare the performance of 5 of these single-cell imputation methods using both simulated data and artificially permuted single-cell RNA-seq data. They evaluate the extent to which these methods introduce false differential expression. A number of clarifications are needed to improve the understandability of the manuscript. Performance benchmarks using additional datasets are also needed to ensure that observed performance differences between methods are not biased by how well the datasets conform to underlying distributions assumed by each method.\n\nMajor comments:\nThe authors conclude that SAVER is the least likely to generate false positives and should be favored over the other 4 imputation methods. However, the scImpute manuscript compared its performance with SAVER to draw conflicting conclusions. I am concerned that the conclusion of which method is better is being biased by the way the benchmark data has been simulated in both cases. Here, the authors simulate data using a negative binomial distribution and find that SAVER had the lowest false positive rate. However, as the authors note, this may be expected, since SAVER models expression data using a negative binomial model. In this manner, the simulation results appear rather circular: the method that uses the same model as the simulated data performs best. In contrast, in the scImpute paper, the authors simulate data using a normal distribution with drop-outs introduced using a Bernoulli distribution and find that SAVER imputation does not alter the data by much or improve downstream clustering whereas scImpute recapitulates the complete data. Please discuss this discrepancy.\n\nThere are genes that are not detected in most single-cells due to poor capture efficiency but we know must be expressed, albeit at low levels, based on bulk RNA-seq, FISH, RT-qPCR, or other approaches for measuring gene expression. As a result, most previous methods have assessed performance by comparing imputed values from single-cell RNA-seq against these bulk RNA-seq, FISH, or RT-qPCR datasets, typically focusing, as the authors note, on the imputation method's ability to recover true signals. How often does imputation introduce a significantly differentially expressed gene in single-cell data that we know should not be differentially expressed based on bulk RNA-seq, FISH, RT-qPCR, or etc? Bulk RNA-seq and single-cell RNA-seq datasets exist for both ESC and DEC cells, which were used for benchmarking in the scImpute paper. Both sorted and unsorted PBMCs are also widely available in both bulk and single-cell RNA-seq form. A number of cell lines have also been sequenced by both bulk and single-cell RNA-seq. In general, the manuscript would greatly benefit from the inclusion of additional benchmarks based on at least one of these datasets. Including additional datasets will also help mitigate the concern that SAVER's superior performance over the other methods is simply the result of both the simulated and the Tabula Muris dataset conforming to the negative binomial model.\n\nThe authors find that many randomly permuted genes were differentially expressed after imputation and furthermore, the direction of the differential expression after imputation was different for different imputation methods. How frequently do these different imputation methods lead to these different directions of differential expression and therefore conflicting biological interpretations? Is Zfp606 the only gene that exhibits this issue suggesting this is a rare event? Or do conflicts arise frequently?\n\nThe authors identify marker genes prior to imputation and note that 95% of marker genes are significant markers in both SmartSeq2 and 10X datasets for the same tissues. They use this comparison between SmartSeq2 and 10X datasets to quantify reproducibility. After imputation, only 80% or so of marker genes were significant in both datasets i.e. decreased reproducibility. Is this decreased reproducibility just due to significance thresholds being reached in one dataset but not the other? Are the -log10(p-values) from the Mann-Whitney-U tests correlated before and after imputation? How do the -log10(p-values) from the Mann-Whitney-U tests correlate between SmartSeq2 and 10X? Before and after imputation?\n\nMinor comments:\nThe terms \"false positive\", \"false signal\", and \"false positive signal\" are used throughout the early components of the manuscript, including the abstract, before it is defined in the \"Permuted Tabula Muris datasets\" section. I initially interpreted \"false positive signal\" loosely to mean genes that are not supposed to be expressed but become non-zero after imputation. However, the definition that the authors are using appears more stringent in that not only does a gene become non-zero after imputation but it becomes significantly differentially expressed. I appreciate this more stringent definition since it more directly impacts biological interpretation. Please define \"false positive signal\" earlier or use a more specific term like \"false differential expression\" to minimize confusion due to terminology.\n\nThe terms \"irreproducible results\", \"reproducibility\", etc. are used throughout the early components of the manuscript, including the abstract before it is defined in the \"Reproducibility of markers\" section. I initially interpreted \"reproducibility\" to mean whether I would get the same results from running the same imputation algorithm multiple times. Please define these terms earlier or use a more specific term to minimize confusion due to terminology.\n\nThe authors note that many imputed markers were assigned to \"contradictory cell-types\" (page 8). Please clarify what this means. What fraction of identified markers does this affect? Does this tend to affect one cell-type i.e. are the markers consistently mixed up between two cell-types?\n\nPlease clarify which methods were run on raw counts and which were run on log2 CPM in Table 1. Was a pseudocount used in the log transformation?\n\nThe authors state that \"scRNASeq imputation only draws on structure within the dataset itself\" but this statement should be limited to the scope of the 5 methods currently tested. scRNAseq imputation methods in the future may draw on external datasets.\n\nFigure 1A is very telling. Could a similar figure be included for the Tabula Muris datasets to visualize the effects of imputation?\n\nReaders would greatly benefit from a discussion on when imputation should be used, if at all, given this observed propensity to introduce false differential expression.\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": [
{
"c_id": "4428",
"date": "19 Feb 2019",
"name": "Tallulah Andrews",
"role": "Author Response",
"response": "Thank you for the helpful suggestions, we have made all suggested Minor corrections and have addressed the Major corrections here and in the revised version of the text:1. Our results broadly agree with the results presented in the scImpute paper, in that SAVER makes modest adjustments to the data, and MAGIC introduces many false signals, whereas scImpute falls in between. However, the scImpute paper focuses on the ability of the method to amplify true signals, such as the tightness of clusters, and the strength/detection of true differential expression, within the data. Whereas, our analysis focused on the tendency to introduce false-positives. Thus it provides complementary rather than contradictory information. Since scImpute uses a zero-inflated normal distribution to approximate log-transformed normalized counts it is expected that it would outperform other methods when that model is used for the simulations as in the scImpute paper. However, RNA-seq data is fundamentally a discrete non-negative process, thus violating the assumptions of the normal distribution. It has been established that read counts from scRNA-seq and bulk RNA-seq (or indeed other -seq protocols) are well described by some variant of the negative binomial distribution e.g. (Grün et al. 2014; Robinson and Smyth 2007), which is why that is the model used here for the simulations. We have added Figure S1 to the supplementary material showing the Splatter simulations (zero-inflated negative binomial) are a good match for real scRNA-seq data. However, it should be noted that we find that 10X data was best simulated as a pure negative binomial, whereas Smartseq2 was best simulated with a zero-inflated negative binomial as has been remarked upon previously (see: http://www.nxn.se/valent/2017/11/16/droplet-scrna-seq-is-not-zero-inflated). In addition, when comparing the fits of the zero-inflated negative binomial distribution and zero-inflated normal distribution to the Tabula Muris raw counts and log-normalized counts respectively we found the negative binomial fits most genes better than the normal distribution (Table S1). Thus, we believe the negative binomial based simulations used here are more relevant to real single-cell RNA-seq data than the simulations used in the scImpute paper.2. While we agree bulk RNA-seq intuitively seems like a good ‘ground truth’ for scRNA-seq it is difficult to use it to evaluate imputation since in general simply summing scRNA-seq data is the closest approximation to bulk RNA-seq by the nature of the experiments. The use of bulk RNA-seq as a ground truth assumes that the assayed cell-populations are in truth completely homogeneous. If the “pure” cell populations are a result of sorting this is almost certainly not correct because there is always a fraction of contaminating cells which will result in a bias towards greater smoothing. Although cell populations obtained by growing cells in culture are more likely to be homogenous, they are a poor model for scRNA-seq data obtained from complex tissue samples. There are also reasons to believe that bulk RNA-seq is not a gold standard for identifying truly differentially expressed genes. Bulk RNA-seq is generally limited by its low power due to a small number of samples and the homogenizing effect of bulk samples. Thus, genes that are simply not-detected as differentially expressed using bulk RNA-seq may in truth be differentially expressed just in a small subset of cells or with a low fold-change. Moreover, even though there are many common steps in the experimental protocols for generating bulk and scRNA-seq, it is likely that there will be effects that are specific to each method. For example: with respect to GC content biases or gene-length biases, bulk RNA-seq may not be more correct than scRNA-seq. There is no reason to believe reproducibility across bulk and scRNA-seq is a more reliable method of benchmarking than reproducibility across different scRNA-seq datasets which we have performed using the Tabula Muris data. We attempted to use two datasets (Kolodziejczyk et al. 2015; Tung et al. 2017) for which matching bulk data was available but the results were inconsistent which is not surprising considering the variability we saw with the Tabula Muris datasets.3. We have added Figure S6 which shows that the proportion of markers with conflicting directions across all the Tabula Muris datasets varies from 5% to 35%. We considered the full imputed Tabula Muris dataset since most genes should have some real differential expression, and thus be more likely to be consistent across imputation methods than the permuted genes, which contain no true signal.4. We apologize that this analysis was not explained clearly. The 95% and 80% are not related to differences in power or significance thresholds, they refer to the percent of markers that were most highly expressed in the same cell-type given that the gene was a significant marker in both datasets. We have clarified this in the text (Results: page 13-14). We appreciate the suggestion for comparing p-values directly and have added a supplementary figure S5 that displays these correlations, further reinforcing our original conclusions that imputation results in poorer reproducibility."
}
]
},
{
"id": "40875",
"date": "06 Dec 2018",
"name": "Charlotte Soneson",
"expertise": [
"Reviewer Expertise Bioinformatics",
"(single-cell) RNA-seq",
"Benchmarking"
],
"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\nAndrews and Hemberg present an interesting evaluation of imputation and smoothing methods for scRNA-seq, focusing on false positive signals. Five recent imputation/smoothing methods are compared based on whether they:\n\nIntroduce false correlations between genes in a Negative Binomial simulation without dropouts. Accurately identify differentially expressed genes in simulated data with different degrees of dropout. Induce false positives in differential expression analysis of permuted real scRNA-seq data. Lead to reproducible sets of differentially expressed genes in data sets generated with different platforms.\nThe paper treats a relevant subject and is generally well written and easy to follow. Below are suggestions for clarifications and a few additions, which I feel would strengthen the paper and provide additional guidance for the reader in determining which, if any, method to use.\nMajor comments:\nAs the authors note, the evaluated methods are based on different distributional assumptions. Since the goal of the imputation is to retrieve the \"true underlying signal\", performance is likely to be strongly affected by the distribution of the data used for evaluation. In the evaluation of falsely induced correlations (a), it would thus be informative to consider different plausible distributions (not only the Negative Binomial), and compare the performance of the methods. In order to avoid making distributional assumptions, perhaps an appropriate bulk RNA-seq data set could also be useful at this stage.\n\nIt would be useful to explicitly spell out the underlying models used by each of the methods, as well as the type of input that they were provided with (raw counts or log-transformed normalized values) and the scale of the output (count or log-count scale) in Table 1. I was also wondering whether correlations in (a) were always calculated on the count scale, or whether they were calculated on the log-scale for some methods. It might be useful to also show the correlations with unimputed log-transformed data in Figure 1A, since not all cells have exactly the same library size/size factor.\n\nDepending on the type of protocol used for the library preparation, scRNA-seq data could have different distributional properties. Since the authors include both SmartSeq2 and 10x data, it would be interesting to see a discussion of the relative merits of the different methods related to the platform used to generate the data. In particular, I was wondering what type of data that the Splatter simulations most resemble, and whether simulations similar to different types of scRNA-seq data could be generated. It would be helpful to see a comparison of the main characteristics of the simulated data and those of real scRNA-seq data, to know to what extent the conclusions drawn from the simulations can be expected to be generalizable to real data sets.\n\nNo attempt is made at explaining the large differences between the Tabula Muris tissues in terms of the number of false positives in the permuted data. Are there any apparent differences between the data sets that might (at least partly) explain this? I think it would also be useful to include the results from unimputed data in Figure 4A-B.\n\nGiven that there are already several imputation/smoothing methods available that were not explicitly evaluated in this study, and that it is likely that this number will increase quickly, it would be very useful if the evaluation would be easily extendable. As a minimum, it would be useful to make the code available, preferably structured in a modular way so that new methods can be easily substituted. Depending on the time and effort required to generate and process the data sets, these could also be made available.\nMinor comments:\nIt is not immediately clear what the numbers in the \"Dropouts (midpoint)\" column in Table 2 represent.\n\nI think it would be worth briefly mentioning Figure S1 in the text, rather than just referring to it in the caption of Figure 1, without discussing its content further.\n\nFor the reproducibility evaluation, only the number of significant genes shared between SmartSeq2 and 10x are reported. How many genes were found to be significant in one data set only?\n\nThe panels in Figure 5 would be easier to compare if the y-axes were the same.\n\nThere are a few typos and inconsistencies (e.g., knn-smooth/knn smooth, raw-counts/raw counts, Smart-seq2/Smartseq2, cell-types/cell types) throughout the text.\n\nIt is not always clear how the statistical tests were applied. For the count-scale data, were the values somehow normalized between cells before the tests were applied? Also, for the log-normalization of the data, what pseudo-count was used, and how were the size factors calculated?\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? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "4429",
"date": "19 Feb 2019",
"name": "Tallulah Andrews",
"role": "Author Response",
"response": "Thank you for the helpful suggestions, we have made all suggested Minor corrections and have addressed the Major corrections here and in the revised version of the text:1. It has been established that read counts from scRNA-seq and bulk RNA-seq (or indeed other -seq protocols) are well described by some variant of the negative binomial distribution e.g. (Grün et al. 2014; Robinson and Smyth 2007), which is why that is the model used here for the simulations. We have added Figure S1 to the supplementary material showing the Splatter simulations are a good match for real scRNA-seq data. However, it should be noted that we find that 10X data was best simulated as a pure negative binomial, whereas Smart-seq2 was best simulated with a zero-inflated negative binomial as has been remarked upon previously (see: http://www.nxn.se/valent/2017/11/16/droplet-scrna-seq-is-not-zero-inflated). In addition, when comparing the fits of the zero-inflated negative binomial distribution and zero-inflated normal distribution to the Tabula Muris raw counts and log-normalized counts respectively we found the negative binomial fits the vast majority of genes better than the normal distribution (Table S1). Thus, we believe the negative binomial to be the most sensible distribution for simulating scRNA-seq data.While we agree bulk RNA-seq intuitively seems like a good ‘ground truth’ for scRNA-seq it is difficult to use it to evaluate imputation since in general simply summing scRNA-seq data is the closest approximation to bulk RNA-seq by the nature of the experiments. The use of bulk RNA-seq as a ground truth assumes that the assayed cell-populations are in truth completely homogeneous. If the “pure” cell populations are a result of sorting this is almost certainly not correct because there is always a fraction of contaminating cells which will result in a bias towards greater smoothing. Although cell populations obtained by growing cells in culture are more likely to be homogenous, they are a poor model for scRNA-seq data obtained from complex tissue samples. There are also reasons to believe that bulk RNA-seq is not a gold standard for identifying truly differentially expressed genes. Bulk RNA-seq is generally limited by its low power due to a small number of samples and the homogenizing effect of bulk samples. Thus, genes that are simply not-detected as differentially expressed using bulk RNA-seq may in truth be differentially expressed just in a small subset of cells or with a low fold-change. Moreover, even though there are many common steps in the experimental protocols for generating bulk and scRNA-seq, it is likely that there will be effects that are specific to each method. For example: with respect to GC content biases or gene-length biases, bulk RNA-seq may not be more correct than scRNA-seq. There is no reason to believe reproducibility across bulk and scRNA-seq is a more reliable method of benchmarking than reproducibility across different scRNA-seq datasets which we have performed using the Tabula Muris data.2. We thank the Reviewer for this suggestion. We have added information about the input, output and underlying model to Table 1 and we have also clarified in the Methods how the correlations were calculated.We have also added the unimputed log-transformed data to Figure 1A.3. We have added Figure S1 comparing the general properties of the real Smart-seq2 and 10X datasets with the Splatter simulations. Generally they are a good match, though the 10X data more closely resemble data simulated with few/no added dropouts, whereas the Smart-seq2 data more closely resembles data with relatively high numbers of added dropouts. In another recent publication from the group (Westoby et al. 2018, Genome Biology), we carried out extensive simulations for comparing isoform quantification methods. We concluded that the splatter simulations did a very good job at resembling the Smart-seq2 data, but the comparisons to Drop-seq data were more tenuous (the discussions on the Drop-seq data were removed from the final version but can be found in the Biorxiv version). 10x data closely resembles Drop-seq data, so those conclusions are likely to hold.4. This was also requested by another reviewer and we have included the unimputed data in Fig 4A and B. We considered the diversity of cell-types, average sequencing depth, number of detected genes, and number of cells, and the goodness of fit of genes to a zero-inflated negative binomial distribution (in table S1) as possible explanation for the variability between datasets but none of them were particularly associated with number of false positives by different methods. However manual inspection of the effect of imputation on the Tabula Muris data (Figure S4) suggests the variable performance across datasets is related to biases in correcting for library size, which would be a combination of differences in cell-size and degree of difference (DE) between cell-types.5. We have made the scripts in a modular structure for the comparison available on github. Thus, it should be straightforward to add methods and re-run the study."
}
]
},
{
"id": "40239",
"date": "06 Dec 2018",
"name": "Stephanie C Hicks",
"expertise": [
"Reviewer Expertise statistics",
"scRNA-seq",
"genomics",
"data 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\nThe authors Andrews and Hemberg provided an insightful analysis assessing whether or not false positives (or capturing false signals) are introduced by imputation methods into scRNA-seq data. Previous papers have only assessed true positives (or positive controls or ability to recover true signal). The authors considered both model-based (SAVER, DrImpute, scImpute) and smoothing-based (knn-smooth, MAGIC) imputation approaches where the former infers only the missing values and the latter smooths all the data (nonzero and zeros).\nI have a few suggestions and questions that I believe would help the manuscript:\n\nIn the simulations (negative binomial and/or Splatter), my understanding is that the authors did not consider a simulation with batch effects (linear or non-linear, global effect or just a portion of the genes), and only considered dropouts in the Splatter simulation. An example with batch effects might be more realistic for scRNA-seq data from real biological experiments because batch effects have been shown to introduce false signals in data (Leek, 20101). My concern is that the false positive signals reported here would actually be larger or more extreme in real scRNA-seq data.\n\nCould the authors explain the reason for using Bonferroni instead of Benjamini-Hochberg (BH) in correcting for multiple testing? I believe that BH is more commonly used in the context of high-throughput computational biology and genomics. Was it an intentional choice to impose a very conservative correction? Also, it would be interesting to use e.g. BH or even a more modern-controlling FDR methods (e.g. IHW from Wolfgang Huber's group). Hopefully this would only improve the ability to detect the true positives (e.g. positive controls), which leads me to my next question.\n\nAs sensitivity and specificity was considered in the Splatter simulations (Figure 2), could the authors show an ROC curve (e.g. averaged across the 60 scRNA-seq count matrices)?\n\nIn the 'Permuted Tabula Muris datasets' section, the authors noted they used Euclidean distance as a form of similarity between two cell types. What about using correlation-based similarity measures instead of Euclidean which has been shown to be highly susceptible to the number of dropouts?\n\nFor the approaches that were applied to the log2 transformed and normalized datasets, did the authors consider normalization methods specific for single-cell (e.g. scnorm or scran)? CPM has been shown to be not appropriate for scRNA-seq data (Vallejos et al., 20172), so I'm wondering if using a more appropriate normalization method improves the results any?\n\nI think one of the biggest concerns is the lack of reproducibility from certain imputation methods (as a side note, Figure 4C was confusing for me and I might suggest the authors consider illustrating this result a different way). This suggests more development is needed to make imputation methods more robust or an external dataset is needed (similar to using haplotype information for GWAS data). Could the authors comment on what they recommend? As this is a good example of a benchmarking paper comparing different imputation methods, I think it would be really useful for the authors to provide a set of recommendations for users.\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": "4427",
"date": "19 Feb 2019",
"name": "Tallulah Andrews",
"role": "Author Response",
"response": "Thank you for the helpful suggestions, we have addressed all comments below and in the updated version of the manuscript: 1. The Reviewer raises a very important point regarding batch effects. We agree that they are likely to make the situation worse for real datasets. However, they are still not well understood and differ greatly between studies in magnitude and genes affected making it difficult to simulate them well. We used Splatter to add small batch effects to all genes in our simulated datasets but this had relatively little effect on the imputation methods’ overall efficiency, however manual inspection of some of them showed that in some cases imputation methods can mistake batch effects for the real underlying structure. We have added this consideration to the Discussion.2. This was a deliberate choice both (a) to be conservative and (b) to reduce the impact of imputation methods distorting the p-value distribution. We have clarified this in the text (Discussion: paragraph 4, Methods: Negative Binomial Simulations). This was specifically used for the Negative Binomial simulation as they did not mimic real single-cell datasets very well since they had many genes with very sharp differences between cell-types, and for testing the false-negatives in the permuted Tabula Muris datasets to avoid biases resulting from how the imputation methods affected genes that were actually differentially expressed in those datasets. For the splatter simulations and reproducibility of marker genes we used the more typical Benjamini-Hochberg/FDR correction since these better reflect real single-cell datasets and we were considering the ability to call true positives not specifically focusing on false positives. This has been clarified in the text (Methods: Splatter Simulations).3. This was requested by another reviewer as well and we have added ROC curves to Figure 2.4. We agree with the reviewer that Euclidean distance is susceptible to the number of dropouts. However, we only used the Euclidean distance only for picking which two cell-types to consider for the permutations thus has very little importance to our analysis, we could just as easily have picked cell-types at random, we only chose the two most similar to increase the number of genes that are not differentially expressed between the cell-types.5. Only one method was designed to be run on already log2 transformed and normalized datasets (DrImpute), while several others (MAGIC, knn) internally apply CPM normalization. Thus, for consistency we used CPM for DrImpute. In addition, SCnorm is slow and scran frequently returns negative size factors unless one manually tunes its parameters for each dataset. Because of the high-throughput nature of our benchmarks we chose not to use these methods.6. We have added recommendations for when and which imputation methods should be used to the Discussion (paragraph 5-6)."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1740
|
https://f1000research.com/articles/8-254/v1
|
05 Mar 19
|
{
"type": "Research Article",
"title": "Genotyping analysis of the Pax9 Gene in patients with maxillary canine impaction",
"authors": [
"Evy Eida Vitria",
"Iwan Tofani",
"Lindawati Kusdhany",
"Endang Winiati Bachtiar",
"Evy Eida Vitria",
"Iwan Tofani",
"Lindawati Kusdhany"
],
"abstract": "Background: Paired-box gene 9 (PAX9) mutation is potentially associated with impaction in some patient populations. Here, we analyzed the relationship between PAX9 polymorphism and the occurrence of maxillary canine impaction. Methods: Patients with and without maxillary canine impaction were selected based on specific inclusion criteria, and samples of genomic DNA were obtained from a buccal mucosa swab. DNA was amplified by polymerase chain reaction and sequenced for further bioinformatics analysis to identify single nucleotide polymorphism (SNP) genotypes. Genotype and allele counting was performed in both case and control groups prior to conducting statistical analysis. Results: Four SNPs were identified in patients with maxillary canine impaction, with relative confidence determined based on chromatogram-peak assessment. All SNPs were located in exon 3 of PAX9 and in the region sequenced by the primer pair −197Fex3 and +28Rex3. Three of the SNPs (rs375436662, rs12881240, and rs4904210) were reported previously and are annotated in NCBI (dbSNP version 150), whereas another SNP mapped to chromosome 14 has not been reported. Patients with a CC genotype at SNP 3 [odds ratio (OR): 2.61 vs. TT; 1.28 vs. CT] and a CC genotype at SNP 4 [OR: 0.71 vs. GG; 0.79 vs. CG] were more likely to have maxillary canine impaction. Conclusions: These results demonstrated that the presence of SNPs 3 and 4 is associated with increased likelihood of suffering from maxillary canine impaction.",
"keywords": [
"Canine impaction",
"PAX9 gene",
"PCR",
"sequencing DNA",
"SNPs",
"Genotype"
],
"content": "Introduction\n\nMaxillary canine teeth are the second most common targets of impaction after the third molars1. An impacted maxillary canine occurs in 1% to 3% of the general population and is twice as common in females as males2,3. It is commonly presented in clinics by patients often arriving with an aesthetic-related complaint.\n\nMaxillary canine impaction in the palatal position is possibly caused by genetic factors and often accompanied by dental abnormalities in tooth shape, size, number, and structure. Abnormalities, such as agenesis, oligodontia, and peg-shaped teeth, have a genetic link to the presence of impacted teeth and generally manifest in developmental disorders during growth4–8. There is a relationship between malposition of certain teeth, such as palatal canines, and teeth agenesis. Similar to dental agenesis, canine-tooth-position anomalies affect several family members and are considered to be under strong genetic control9–15. Paired-box gene 9 (PAX9) is most commonly involved in affecting the odontogenesis process and thought to determine the localization of tooth seeds16. In this study, we identified an association between PAX9 genotype and the occurrence of maxillary canine impaction.\n\n\nMethods\n\nPatients were recruited from the Dental Hospital Faculty of Dentistry, Universitas Indonesia, and three different junior high schools in South Jakarta, and the study was conducted through clinical observations from May 2018 to August 2018. Those meeting the inclusion criteria (male or female, 10–25-years old, no systemic disease, and no hereditary disease) were either diagnosed with maxillary canine impaction (group I) or diagnosed as without (group II; control). Diagnosis was based on clinical examinations and radiographic interpretations performed by radiologists and orthodontists who were experts in their fields. Comprehensive clinical data were obtained for 132 patients (see Underlying data17). All participants gave their written informed consent to participate in this study, which was approved by the Ethics Committee of the Faculty of Dentistry, Universitas Indonesia (No. 07107105/Ethical Approval/ FKGUI/2018).\n\nGenomic DNA was collected from the buccal mucosa via swabbing and extracted using the Gene Jet whole blood genomic DNA purification kit (Cat. No. K0781; Thermo-Biogen, Karlsruhe, Germany). The area of the buccal mucosa/cheek to be treated was dried with a cotton roll to prevent salivary contamination. Samples from the buccal mucosa were obtained using a cytobrush (#C0104; Medscan; Cooper Surgical, Trumbull, CT, USA) on the bilateral buccal mucosa, with each side swabbed 10 to 15 times. The cytobrush swab was inserted into a screw-capped Eppendorf tube (Cat. No. SPL-60015; Extragene, Taichung City, Taiwan) containing 200 µL of 1% phosphate-buffered saline. Eppendorf tubes were labeled and stored in a freezer at −4°C. DNA concentration was measured using a Qubit Fluorometer 3 (#Q33216; Invitrogen, Carlsbad, CA, USA) at standard fluorescence wavelengths (excitation/emission: ~480/530 nm) with Qubits assay reagent (MP423-#Q32851; Qubit dsDNA HS assay kit; Invitrogen).\n\nWe used four sets of primers spanning the PAX9-coding region (exons 2, 3, and 4) (Table 1) and standard PCR procedures for amplification of genomic DNA. PCR was performed using a T100 thermal cycler (No. #186-1096; Bio-Rad, Hercules, CA, USA) in a total volume of 25 µL containing 20 µL master mix (MyTag, 12.5 µL; Primer F, 0.5 µL; Primer R, 0.5 µL; and nuclease-free water, 6.5 µL) and 20 ng DNA. Samples were initially heated to 95°C for 10 min, followed by 30 cycles of denaturation at 95°C for 2 min, primer annealing at the optimal annealing temperature for the PAX9 primer for 1.5 min (optimal annealing temperature for PAX9 primers were as follows: F1-R1, 60°C; F2-R2, 64°C; F3-R3, 60°C; F4-R4, 56°C to 64°C), and extension/elongation at 72°C for 2 min then a final extension at 72°C for 15 min.\n\nThe PCR products were analyzed by electrophoresis using a 2% agarose gel (UltraPure Agarose; Cat. No. #16500500; Thermo Fisher Scientific, Waltham, MA, USA). To make the 2% agarose gel, 2 g agarose was added to 100 mL Tris base-acetate-EDTA (TAE) buffer, followed by the addition of 1 µL of GelRed nucleic acid gel stain (Cat. No. #41003-1; Biotium, Fremont, CA, USA). The wells were loaded with 5 µL of sample, with a DNA standard added to one well ; Cat. No. #SM0241; Thermo Fisher Scientific), and the gel was run at 100 V for 30 min.. The results of gel electrophoresis were visualized by using UV-transiluminator Gel DocTM 2000 (Cat. No. #170-8101; Bio-Rad, Hercules, CA,, USA).\n\nExtracted and adapted from Vastardis et al. (1996) and Nieminen et al. (2001). (+) indicates a sequence of DNA that is in front of the codon where transition begins and (−) the sequence of DNA that is behind the codonF, forward; R, reverse; ex, exon; numeration, nucleotide position in the sequence.\n\nDNA from the PCR products was purified using an QIAquick PCR purification kit (Cat. No 28106; Qiagen, Hilden, Germany), and purified DNA was sequenced by First-Base Laboratories (Selangor, Malaysia). Sequencing data were edited using BioEdit software (v.7.0.9; Ibis Therapeutics, Carlsbad, CA, USA) and verified using NCBI BLAST (GenBank accession No. NG_0133557.1:5001-25240).\n\nTo detect sites containing potential single nucleotide polymorphisms (SNPs), sequencing data were converted to FASTQ format, and chromatogram-peak of each patient DNA sequence was performed as can be seen in the representative data from one sample (Figure 1). Heterozygous SNPs were determined by visually identifying sites containing two overlapping peaks in both the forward and reverse sequences.\n\nFor each SNP identified, genotypes across all samples were obtained, annotated, and cross-checked against the NCBI SNP reference database (dbSNP v.150). Both genotype and allele counting were performed before conducting statistical analysis for both the case and control groups.\n\nStatistical analysis of the data was performed using SPSS (v.20.0; SPSS, IBM Corp., Armonk, NY, USA). Analysis was initiated with a quality check of the variables. Analysis of genotype data was conducted using a chi-squared test to compare SNP frequency between patients with maxillary canine impaction and controls. A P < 0.05 was considered statistically significant.\n\n\nResults\n\nFour SNPs were identified, with all of these located in exon 3 of PAX9 sequenced using primer pair 3 (−197Fex3 and +28Rex3). Sequencing data for this region were available for 121 of 132 samples, and no SNPs were identified in regions sequenced using the other primer pairs.\n\nTable 2 summarizes the number of SNPs identified. Of the four identified, three were reported previously in exon 3 of PAX9 and are annotated in NCBI dbSNP (v.150; rs375436662, rs12881240, and rs4904210). The fourth SNP in exon 3 maps to chromosome 14, position 36,666,530, and has not been previously reported.\n\nTable 3 lists the SNPs identified in this study along with their annotation/reference identification and the location of the nucleotide substitutions. Nucleotide changes included 640A>G in SNP 1, 700C>T in SNP 2, 717C>T and 718G>C in SNP 3.\n\nGenotype assessment was performed for all of the identified SNPs, as well as controls, focusing on counts of homozygous (wild-type) and heterozygous and homozygous (mutant) alleles (Table 4–Table 7).\n\nFor SNPs 3 and 4, all genotype and allele variations were observed in both the case and control samples, whereas for SNPs 1 and 2, the homozygous and heterozygous mutants were present in only one of the cases (Table 6 and Table 7).\n\nWe then analyzed associations between SNPs 3 and 4, which were present in both case and control samples, and maxillary canine impaction. We verified Hardy–Weinberg equilibrium (HWE) for both SNPs in case and control samples (Table 8), with the TT genotype in SNP 3 (rs12881240) showing a higher degree of association with maxillary canine impaction than those with the CC genotype [odds ratio (OR): 2.61; 95% Confidence interval (CI): 0.29–23.61] or the CT genotype (OR: 1.28; 95% CI: 0.57–2.89), although this results was not statistically significant (P > 0.05). This result indicated a similar frequency of recessive and dominant carriers of SNP 3 between case and control samples. Furthermore, no statistical difference in allele frequency was observed between the two groups.\n\nChi-square test p<0,05 ( Sig 2- Tailed ), HWE: Hardy–Weinberg equilibrium\n\nIndividuals with the GG genotype in SNP 4 (rs4904210) were less likely to have maxillary canine impaction than those with the CC genotype (OR: 0.71; 95% CI: 0.23–2.16] and the CG genotype (OR: 0.79; 95% CI: 0.31–2.00] (Table 8). Additionally, individuals were less likely to have the G allele than the C allele (OR: 0.84; 95% CI: 0.49–1.45]. However, similar to SNP 3, these results were not statistically significant (P > 0.05).\n\nThe use of patient genotype information for clinical assessment represents a possible diagnostic strategy for predicting disease likelihood. Patient characteristics and genotype information were available from 121 of 132 samples, with Table 9 summarizing the clinical information associated with both the case and the control groups.\n\nChi-square test p<0,05 ( Sig 2- Tailed )\n\nStratification of cases with SNP 3 (rs12881240) showed that 21 of 53 males and 24 of 68 females in the case group harbored the CT genotype, whereas only four males and two females harbored the TT genotype, with the difference between males and females in this group not statistically significant (P = 0.3996). Additionally, the OR according to gender was 0.76 (95% CI: 0.36–1.62) for individuals harboring the CT genotype of SNP 3 (rs12881240) relative to the CC genotype, whereas it was 0.33 (95% CI: 0.06–1.94) for individuals harboring the TT genotype relative to the CC genotype.\n\nFor SNP 4 (rs4904210), 26 of 53 males and 36 of 68 females in the case group had a CG genotype, whereas 13 males and 13 females had the GG genotype. Additionally, the OR according to gender was 1.02 (95% CI: 0.43–2.40) for individuals harboring the CG genotype of SNP 4 (rs4904210) relative to the CC genotype, whereas it was 0.74 (95% CI: 0.26–2.07) for individuals harboring the GG genotype relative to the CC genotype.\n\nAlthough the frequency of genotype variation was higher in females than in males, the difference was not statistically significant (P = 0.7725). Consequently, these findings suggested that gender was not a major influence on the occurrence of maxillary canine impaction.\n\n\nDiscussion\n\nThe etiology of canine impaction might be multifactorial and involve external factors, such as environmental input (e.g., trauma), local factors (e.g., lack of space, prolonged retention of primary teeth, trauma to permanent tooth seeds, rotation of permanent seed teeth, and the presence of pathological lesions, such as dentigerous cysts or odontoma), genetic factors, and systemic disease18–24. Dentistry is increasingly making use of genetic information, which plays an important role in addressing clinical problems, especially with regard to dental abnormalities or anomalies, including impaction of the maxillary canines. Previous studies show that tooth development is regulated by >200 genes23. PAX9 encodes a transcription factor and is among the most frequently identified genes affecting the odontogenic process and involved in the occurrence of dental anomalies, such as agenesis teeth, congenital missing teeth, and variabilities in tooth size and position. The identification of genetic risk factors associated with canine impaction has recently become the subject of intensive research.\n\nIn this study, DNA sequencing of 121 of 132 patient and control samples identified four SNPs located in a similar region of PAX9 exon 3 (Table 2). SNPs play a role in determining disease characteristics, including etiology and the incidence and risk of disease development. Subsequent analysis revealed that all of the identified SNPs would result in missense mutations (Table 3). These findings suggest that SNPs might be efficacious for determining dental anomalies, specifically the impaction of maxillary canines.\n\nAlthough we found no statistically significant association between PAX9 genotype and maxillary canine impaction (Table 8), there were variations between patients with and without this condition according to the presence of SNP 3 and SNP 4, which carried a greater risk for maxillary canine impaction. A previous study reported that the genes involved in the impaction or displacement of canines into the palate are also responsible for controlling the growth and eruption of teeth25, and Klein et al.26 showed that dental anomalies (size, shape, and position of teeth and agenesis of teeth or supernumerary teeth) were determined by a set of genes involved in tooth development. The results of the present study suggested a potential role for PAX9 in tooth growth and development.\n\nIn summary, our findings showed no statistically significant association between SNP genotype and gender and demonstrated that although the frequency of impaction-related genotype variation in women was higher than that in men, the differences were not statistically significant. These results suggested that gender-associated variations in genetic profile do not contribute to the incidence of maxillary canine impaction.\n\n\nConsent\n\nWritten informed consent for publication of patient details was obtained.\n\n\nData availability\n\nABI files and chromatograms that support the findings of this study are available on reasonable request from the corresponding author [author initials] and by submitting the applicable request form (‘Form Padia’ to request access for genetic resources (available as part of the OSF deposit)). The data are not publicly available due to them containing information that could compromise research participant privacy.\n\nOpen Science Framework: Genotyping Analysis Pax9 Gene In Patients With Maxillary Canine Impaction. https://doi.org/10.17605/OSF.IO/B37CJ17.\n\nThis project contains the following underlying data:\n\nELECTROPHORESIS_ RESULTS- PAX9 GENE.docx_05-02-2018 (Gel images)\n\nRAW DATA _F1000 Vitria E 05-02-2019L.xlsx (Patient data)\n\nOpen Science Framework: Genotyping Analysis Pax9 Gene In Patients With Maxillary Canine Impaction. https://doi.org/10.17605/OSF.IO/B37CJ17\n\nThis project contains the following extended data:\n\nPADIA form-Pax0 gene_Vitria E-05-02-2019.docx (Data access form)\n\nData are available under CC0 1.0 Universal Public Domain Dedication",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nWe are grateful for the editage for their copyediting services. We also thank you to Dr. Miesje Karmiati who helped establish the clinical diagnosis of canine tooth impaction cases in this study. Lastly, we thank Astridevitanti (Vivi) who helped work on PCR in Oral Biology Laboratory, Faculty of Dentistry UI.\n\n\nReferences\n\nPower SM, Short MB: An investigation into the response of palatally displaced canines to the removal of deciduous canines and an assessment of factors contributing to favourable eruption. Br J Orthod. 1993; 20(3): 215–22. PubMed Abstract | Publisher Full Text\n\nDachi SF, Howell FV: A survey of 3,874 routine full-mouth radiographs. I. A study of retained roots and teeth. Oral Surg Oral Med Oral Pathol. 1961; 14: 1165–9. PubMed Abstract | Publisher Full Text\n\nThilander B, Myrberg N: The prevalence of malocclusion in Swedish schoolchildren. Scand J Dent Res. 1973; 81(1): 12–20. PubMed Abstract | Publisher Full Text\n\nWang Y, Wu H, Wu J, et al.: Identification and functional analysis of two novel PAX9 mutations. Cells Tissues Organs. 2009; 189(1–4): 80–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDas P, Stockton DW, Bauer C, et al.: Haploinsufficiency of PAX9 is associated with autosomal dominant hypodontia. Hum Genet. 2002; 110(4): 371–6. PubMed Abstract | Publisher Full Text\n\nKapadia H, Frazier-Bowers S, Ogawa T, et al.: Molecular characterization of a novel PAX9 missense mutation causing posterior tooth agenesis. Eur J Hum Genet. 2006; 14(4): 403–9. PubMed Abstract | Publisher Full Text\n\nPereira TV, Salzano FM, Mostowska A, et al.: Natural selection and molecular evolution in primate PAX9 gene, a major determinant of tooth development. Proc Natl Acad Sci U S A. 2006; 103(15): 5676–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKapadia H, Mues G, D'Souza R: Genes affecting tooth morphogenesis. Orthod Craniofac Res. 2007; 10(4): 255–44. PubMed Abstract | Publisher Full Text\n\nMercuri LG, O’Neill R: Multiple impacted and supernumerary teeth in sisters. Oral Surg Oral Med Oral Pathol. 1980; 50(3): 293. PubMed Abstract | Publisher Full Text\n\nPeck S, Peck L, Kataja M: The palatally displaced canine as a dental anomaly of genetic origin. Angle Orthod. 1994; 64(4): 249–56. PubMed Abstract\n\nPeck S, Peck L, Kataja M: Prevalence of tooth agenesis and peg-shaped maxillary lateral incisor associated with palatally displaced canine (PDC) anomaly. Am J Orthod Dentofacial Orthop. 1996; 110(4): 441–3. PubMed Abstract | Publisher Full Text\n\nPirinen S, Arte S, Apajalahti S: Palatal displacement of canine is genetic and related to congenital absence of teeth. J Dent Res. 1996; 75(10): 1742–6. PubMed Abstract | Publisher Full Text\n\nPeck S, Peck L: Palatal displacement of canine is genetic and related to congenital absence of teeth. J Dent Res. 1997; 76(3): 728–9. PubMed Abstract | Publisher Full Text\n\nPeck S, Peck L, Hirsh G: Mandibular lateral incisor-canine transposition in monozygotic twins. ASDC J Dent Child. 1997; 64(6): 409–13. PubMed Abstract\n\nPeck S, Peck L, Kataja M: Mandibular lateral incisor-canine transposition, concomitant dental anomalies, and genetic control. Angle Orthod. 1998; 68(5): 455–66. PubMed Abstract\n\nNeubüser A, Koseki H, Balling R: Characterization and developmental expression of PAX9, a paired-box-containing gene related to PAX1. Dev Biol. 1995; 170(2): 701–16. PubMed Abstract | Publisher Full Text\n\nBachtiar EW: Genotyping Analysis Pax9 Gene In Patients With Maxillary Canine Impaction. 2019. http://www.doi.org/10.17605/OSF.IO/B37CJ\n\nLitsas G, Acar A: A review of early displaced maxillary canines: Etiology, diagnosis and interceptive treatment. Open Dent J. 2011; 5: 39–47. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBishara SE: Impacted maxillary canines: A review. Am J Orthod Dentofacial Orthop. 1992; 101(2): 159–71. PubMed Abstract | Publisher Full Text\n\nThilander B, Jakobsson SO: Local factors in impaction of maxillary canines. Acta Odontol Scand. 1968; 26(2): 145–68. PubMed Abstract | Publisher Full Text\n\nSchindel RH, Duffy SL: Maxillary transverse discrepancies and potentially impacted maxillary canines in mixed-dentition patients. Angle Orthod. 2007; 77(3): 430–5. PubMed Abstract | Publisher Full Text\n\nRichardson G, Russell KA: A review of impacted permanent maxillary cuspids--diagnosis and prevention. J Can Dent Assoc. 2000; 66(9): 497–501. PubMed Abstract\n\nBecker A, Sharibi S, Chaushu S: Maxillary tooth size variation in dentitions with palatal canine displacement. Eur J Orthod. 2002; 24(3): 313–8. PubMed Abstract | Publisher Full Text\n\nPeck S, Peck L, Kataja M: The palatally displaced canine as a dental anomaly of genetic origin. Angle Orthod. 1994; 64(4): 249–56. PubMed Abstract\n\nNIH US National Library: What are single nucleotide polymorphisms (SNPs)?2019. Reference Source\n\nKlein ML, Nieminen P, Lammi L, et al.: Novel mutation of the initiation codon of PAX9 causes oligodontia. J Dent Res. 2005; 84(1): 43–7. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "45394",
"date": "22 Mar 2019",
"name": "Heni Susilowati",
"expertise": [
"Reviewer Expertise 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 study has reported new information that is very important to explain the effect of genes mutation in canine impaction. This research is very interesting. The researcher analysed the relationship between mutations and Pax9 gene polymorphisms with cases of canine tooth impaction. This research has been carried out using gene sequencing. The results showed an association between Pax9 gene mutations (SNPs 3 and 4) with the appearance of maxillary canine teeth impaction. The abstract contains concise and clear information about the background, objectives, methods, results and conclusions, so that it is quite easy for the readers. The following are inputs to improve the quality of this manuscript:\nIn the abstract, it is written in the background section: \"Paired-box gene 9 (PAX9) mutation is often associated with impaction in some patient populations.\" This indicates that previous researchers may have conducted similar studies in different populations. It is advisable to write about this in the introduction, to strengthen the reasons for choosing the pax9 gene and also in the discussion, to analyse possible similarities or differences in pax9 polymorphism in different study populations. The results showed that gender has no effect on the incidence of canine impaction, gender-associated genetic variation has no contribution. It would be better if the authors explain the possible causes, or show references from previous research reports that are relevant to the phenomenon. Does this occur only in canine impaction or is it possible for other tooth impaction cases? Please add to the discussion section, factors that are likely to cause the emergence of Pax9 gene polymorphism. Material and methods: the researcher conducted a sequencing analysis, please explain whether the method used is next generation sequencing.\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": "47551",
"date": "26 Apr 2019",
"name": "Fathilah Binti Abdul Razak",
"expertise": [
"Reviewer Expertise Oral 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 attempt to associate paired-box gene 9 (PAX9) mutation to tooth impaction in patients was interesting and the decision to target on the occurrence of maxillary canine impaction was logical and appropriate. The study design was clearly outlined, well described and detailed. Data obtained was adequate, which enabled derivation of meaningful outcomes. The authors were able to demonstrate that the presence of SNPs 3 and 4 is actually associated with the increased likelihood of suffering from maxillary canine impaction. This finding definitely adds valuable information to the current knowledge.\n\nIn general, this is a nicely written manuscript. The flow of ideas and thoughts of the authors throughout the manuscript were effectively presented, focused and of high quality.\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/8-254
|
https://f1000research.com/articles/8-253/v1
|
05 Mar 19
|
{
"type": "Systematic Review",
"title": "Safety and efficacy of Azithromycin in prevention of chronic obstructive pulmonary disease exacerbation: systematic review and meta-analysis",
"authors": [
"Almegdad Ahmed",
"Abubaker Koko",
"Ahmed Abdelsalam",
"Awab Hilali",
"Mohamed Elsheikh",
"Abubaker Koko",
"Ahmed Abdelsalam",
"Awab Hilali",
"Mohamed Elsheikh"
],
"abstract": "Background: Chronic obstructive pulmonary disease (COPD) causes a major burden in terms of deaths and hospitalizations worldwide; it is associated with progressive lung function loss, and frequent exacerbations. Administration of macrolides has been considered beneficial in reducing the frequency of COPD exacerbations. The aim of this study is to assess the safety and efficacy of long-term administration of Azithromycin for patients with chronic obstructive pulmonary disease. Methods: An extensive search was conducted on SCOPUS, and PubMed databases, CENTRAL, and ClinicalTrials.gov clinical trial registers for randomized clinical trials conducted on COPD patients and administered Azithromycin for more than two weeks. The selected studies underwent assessment for the risk of bias. We conducted random-effect model meta-analysis for the frequency of acute exacerbations during follow-up as a primary outcome. Results: Out of 1021 screened records, 3 RCTs (Randomized controlled trials) involving 1264 patients were included in the final analysis. The pooled data of all 3 trials showed that administration of Azithromycin reduced the frequency of acute exacerbation of COPD [risk ratio (RR) = 0.69; 95% CI 0.53, 0.91, p = 0.01]. Subgroup analysis indicated that 500 mg Azithromycin [risk ratio (RR) =0.65; 95% CI 0.53-0.79, p=0.01] was found to be more beneficial than 250 mg Azithromycin [risk ratio (RR) = 0.60; 95% CI 0.27-1.33, p=0.21] in reducing acute exacerbation rate, however due to many limitations the analysis of the dosage was not conclusive. Conclusion: Long-term Azithromycin administration for COPD patients is statistically not associated with increased risk of developing adverse events; in addition, it might be effective in reducing the frequency of acute exacerbations of COPD. However, dosage and duration of Azithromycin administration analysis was not conclusive and thus more RCTs are needed in these areas.",
"keywords": [
"chronic obstructive pulmonary disease",
"acute exacerbation",
"macrolides",
"Azithromycin"
],
"content": "Abbreviations\n\nCOPD: Chronic obstructive pulmonary disease.\n\nFEV 1: Forced expiratory volume in 1 second.\n\nFVC: Forced vital capacity.\n\nRR: Risk ratio.\n\nCI: Confidence interval.\n\nSGRQ: St George Respiratory Questionnaire.\n\nSF-36: Short-form 36 Health Survey.\n\nSF-12: 12-Item Short-Form Health Survey.\n\n\nIntroduction\n\nChronic obstructive pulmonary disease (COPD) is “a disease state characterized by airflow limitation that is not fully reversible”1. It is a progressive life threatening lung disease, with frequent symptoms that include: chronic cough, breathlessness, and sputum production2. It is the fourth leading cause of death worldwide affecting more than 60 million people and it is estimated to become the third by 2030. 3.2 million deaths occur due to COPD, with 90% of them occurring in middle and low income countries3–5. Until now, there is no definitive treatment for COPD and thus effective management of COPD relies on improving health status, reducing the mortality, as well as preventing and treating complications and exacerbations of the disease6.\n\nCOPD exacerbation is defined as “an acute worsening of respiratory symptoms that result in additional therapy”, and classified as mild (responsive to short acting bronchodilators), moderate (requiring antibiotics and/or oral corticosteroids in addition to short acting bronchodilators), or severe (requiring hospitalization)7. Numerous international guidelines have discussed recommendations about the benefits of antimicrobial agents in patients with acute exacerbation of COPD, as they are more prone to develop bacterial infections8–10. Among these antimicrobial agents, macrolides (such as Azithromycin) are especially favored as prophylactic antibiotics in COPD, mostly due to their potentially beneficial anti-inflammatory and immunomodulatory roles, in addition to their known antibacterial effect11.\n\nAzithromycin belongs to macrolides family; it is a semi-synthetic antibiotic derived from another macrolide, erythromycin. Azithromycin has bactericidal and bacteriostatic properties, hence it is used in treating certain bacterial infections12. Azithromycin acts by binding reversibly to the 50S ribosomal subunit of the 70S ribosome of the targeted bacteria, thereby inhibiting the translocation step of protein synthesis stopping bacterial growth13.\n\nSeveral studies have discussed the beneficial role of macrolides for COPD patients14–19, some of them suggested that long-term Azithromycin therapy has been shown to decrease the rate of recurrence and severity of exacerbations of COPD20,21. However, there is little evidence as to whether Azithromycin is safe and efficacious for management of COPD; thus our study aims to evaluate the efficacy and safety of long-term administration of Azithromycin therapy for chronic obstructive pulmonary disease by accumulating the evidence from randomized control trials that studied the efficacy and safety of Azithromycin administration for COPD patients.\n\n\nMethods\n\nWe systematically searched the literature on PubMed and Scopus databases, as well as the Cochrane central register of controlled trials (CENTRAL) and ClinicalTrials.gov register clinical trial registers. We searched using the terms; chronic obstructive pulmonary disease, exacerbation, chronic bronchitis, emphysema, Azithromycin. The electronic search string used for PubMed was as follows: (((chronic obstructive pulmonary disease OR chronic bronchitis) OR emphysema) AND exacerbation) AND Azithromycin. Our search was limited to human studies only. Our last search was on January 24, 2018. Moreover, hand searching was done by screening the reference lists of all included studies (Figure 1).\n\nOur eligibility criteria for studies to be included in our meta-analysis were: (1) Randomized controlled trials that included only stable COPD patients in any stage of the disease. (2) COPD is defined clinically as [forced expiratory volume in 1 second (FEV 1) /forced vital capacity (FVC)] <70%, FEV 1 <80% predicted, and an increase in FEV 1 <12% (or 200 ml) after inhaling bronchodilators, according to Tiffeneau-Pinelli Index22. (3) Studies that used Azithromycin. (4) The drug was administered orally and the therapy lasted more than two weeks. (5) Information about clinical efficacy or the safety of the drug were reported.\n\nFour reviewers, AA, KA, AA, and HA screened the titles and abstracts of the search results independently for potentially eligible studies. After removing duplicates and irrelevant records, they independently reviewed the full-text of potentially eligible studies using the previously mentioned inclusion criteria. Any differences between the reviewers were solved through consensus. There were no disagreements that needed to be resolved by the senior author.\n\nWe measured the risk of bias for all studies that fulfilled our inclusions criteria using RevMan 5.3 software (Cochrane Collaboration, Copenhagen, Denmark) risk of bias assessment tool using: allocation concealment, random sequence generation, blinding of participants and personnel, incomplete outcome data, blinding of outcome assessment, selective reporting as the main parameters for bias. The information that we extracted from the selected studies were study setting information, patients’ characteristics information, treatment information (dosage, therapy strategy, course of therapy, concomitant medication to treat COPD), frequency of exacerbations, and adverse events reports. The reviewers extracted this information independently. Output RevMan file used for analysis is available as Underlying data23.\n\nWe used random-effects models to pool treatment effects and to calculate the risk ratios (RR) with 95% CI for all clinical end-points, which were the frequencies of exacerbations. All types of exacerbations (mild, moderate, and severe) were included in the pooled analysis with no distinction between them, because some of the included studies reported severe exacerbations while others reported non severe exacerbations. To examine the robustness of the effect we performed a sensitivity analysis by removing the trials with the highest weights and computing the overall estimates for the remaining studies. Regarding statistical heterogeneity, we used the I2 statistic on a scale of 0–100% (>50% indicated a statistical between-study inconsistency, and >75% represented a very large degree of heterogeneity). Subgroup analysis were performed for two strata of data, dosage of Azithromycin used (250 mg Azithromycin or 500 mg Azithromycin) and the duration of the administration (3 months or 12 months). Funnel plot method was used to assess the publication bias. P < 0.05 was considered statistically significant. RevMan 5.3 software (Cochrane Collaboration, Copenhagen, Denmark) was used to perform the pooling analyses of the data.\n\n\nResults\n\nInitially 1031 published records were identified (677 from SCOPUS, 199 from PubMed, 15 from ClinicalTrials.gov register, and 140 from CENTRAL) (Figure 1). 4 studies24–27 meet the inclusion criteria. 1348 participants were included in the four trials, the follow-up duration varied among the studies: 3 months25, 6 months27, and 12 months24,26. Of the total number of participants in the included trials, 674 participants were randomly assigned to receive Azithromycin, while 674 were allocated to receive Placebo; in addition both groups continued receiving their concomitant medications (long acting β2 agonists, long acting anticholinergics, inhaled corticosteroids, short acting β2 agonists, oral Prednisolone) (Table 1), Azithromycin dosages used were 250 mg Azithromycin24,25,27 and 500 mg Azithromycin26.\n\nN: number, Yrs: years, COPD: chronic obstructive pulmonary disease, LAMAs: long acting anticholinergics, LABAs: long acting β2 agonists.\n\nThe duration of Azithromycin therapy was for 3 months25,27 and 12 months24,26, the total number of withdrawals among the selected studies was 66. Three studies24,26,27 compared the frequency of acute exacerbation difference that occurred during the follow-up period between the Azithromycin group and the Placebo group using Poisson regression model. One study25 reported acute exacerbations as percentages for both of the study groups. All four studies reported the adverse events that occurred during the follow-up period. In addition to acute exacerbations rate and the adverse events, other outcomes like health related quality of life24–27, bacterial culture, macrolides resistance24,26,27, and pulmonary functions test25–27 were also reported.\n\nThe time to the first acute exacerbation of COPD in days was higher among patients who received Azithromycin compared patients who received the Placebo24–26. Uzun et al. revealed that hospital admission odds ratio did not differ between the two groups26, however, according to the other studies hospital admissions were reported more amongst the Placebo group24,25,27. Furthermore, chest computed tomography symptom score did not show any difference between the two groups27, and after 12 weeks of Azithromycin therapy a significant increase in the Leicester Cough Questionnaire total score was reported in the Azithromycin group compared with the Placebo. Pulmonary function test was conducted in all four studies24–27, with the exception of Albert et al., all indicated that no significant difference in the test results were reported between Azithromycin and Placebo groups.\n\nNasopharyngeal colonization was assessed in three studies24,26,27 , two of them24,26 reported that fewer patients in the Azithromycin group had positive culture during the course of the study (p value < 0.001, p value= 0.044 respectively) while the other27 reported no difference between the Azithromycin and Placebo group. Macrolide-resistant bacteria were detected more in the Azithromycin group of Simpson et al., in which the incidence of resistance was 81% among Azithromycin group, and 41% in the placebo (p value < 0.001)27, however it was detected more among the Placebo group with 24%, in contrast to 6% among Azithromycin group (p value = 0.036) from Uzun et al.26. Nasopharyngeal colonization was not associated with the occurrence of acute exacerbations in either group (p value 0.31)24.\n\nQuality of life was recorded using different tools: St George Respiratory Questionnaire (SGRQ)24–27, Short-form 36 Health Survey (SF-36)24,25, and the 12-Item Short Form Health Survey (SF-12)26. Three studies24–26 reported a decrease in SRGQ total scores in the Azithromycin group compared with the Placebo group, however one study27 reported no decrease in the total score among Azithromycin group. The results of the SF-36 scores varies among studies as one study24 stated that no changes were seen in the scores while the other study25 reported a significant difference in the scores between Azithromycin and Placebo groups. In addition, the mean change differed significantly between the Placebo group and the Azithromycin group in the mental component score of SF-1226. The most commonly reported adverse event was hearing decrement which was reported in 18% (252) of the participants, an audiogram-confirmed hearing decrement occurred more frequently in participants receiving Azithromycin compared to those receiving Placebo (P=0.04), however, audiogram-confirmed hearing decrement was used in only one study24. Table 2 demonstrates the side effects that occurred among the patients.\n\n* = diarrhea, nausea, vomiting and gastric ulcer...etc.\n\n#= QTc prolongation, myocardial infarction, supraventricular tachycardia, heart failure…etc.\n\n&= common cold, dyspnea and cough, Pneumonia...etc.\n\n$= Neoplasm, Tinnitus, Allergic reactions, abnormal lab tests...etc.\n\nThe final analysis of the pooled data of the all three trials showed that administration of Azithromycin reduced the frequency of acute exacerbation of COPD [risk ratio (RR) = 0.69: 95% CI 0.53, 0.91, p = 0.008] (Figure 2). Subgroup analysis of Azithromycin dosage (Table 3) revealed that 500 mg Azithromycin [risk ratio (RR) =0.65; 95% CI 0.53–0.79, p=0.01] is more beneficial than 250 mg Azithromycin [risk ratio (RR) = 0.60; 95% CI 0.27–1.33, p=0.21] in reducing acute exacerbation rate. Moreover, administration of Azithromycin for 12 months [risk ratio (RR) = 0.75; 95% CI 0.59–0.94, p=0.01] was more effective than 3 months. [risk ratio (RR) = 0.36; 95% CI 0.16–0.79, p=0.01]. Additionally, long-term administration of Azithromycin to COPD patients was not associated with increased risk of developing adverse events [risk ratio (RR) = 0.94; 95% CI 0.81, 1.11, p = 0.48] (Figure 3), subgroup analysis of Azithromycin dosage and therapy duration (Table 4) reported no significant difference between Azithromycin and Placebo groups in developing adverse events.\n\nThe size of the square is proportional to the weight of the individual studies. M-H = Mantel-Haenszel method.\n\nThe size of the square is proportional to the weight of the individual studies. M-H = Mantel-Haenszel method.\n\nRegarding the risk of bias assessment, the included studies had low risk of bias. Additionally, sensitivity analysis and publication bias sensitivity analysis on the frequency of acute exacerbation performed by deleting the highest weight trial24 revealed no significant difference on the results (p value changed to 0.03). In addition, sensitivity analysis on the adverse events performed by deleting the highest weight trials24,27 also revealed no significant difference on the results (p value changed to 0.36). Furthermore, regarding publication bias analysis, funnel plots of both acute exacerbation frequency and adverse events were symmetrical. In regard to heterogeneity among the studies, 77% heterogeneity was found among the frequency of acute exacerbation meta-analysis studies, while 33% was found among adverse events meta-analysis studies.\n\n\nDiscussion\n\nLong-term administration of Azithromycin was associated with reduction in the frequency of acute exacerbation among COPD patients, similar findings have been detected among cystic fibrosis patients28, however Yao and colleagues deduced that Erythromycin was more effective than Azithromycin in decreasing acute exacerbation frequency29.\n\nMany studies have investigated the immunomodulatory effects of macrolides, like decreasing the synthesis of pro-inflammatory cytokines in response to viral infections30, decreasing the hypersecretion of pro-inflammatory cytokines and chemokines31, enhancement of phagocytosis function of alveolar macrophages32, and preserving airway epithelial integrity33. Moreover, in addition to the immunomodulatory effects, the decrease in airway bacterial colonization in patients receiving Azithromycin might also be linked to the reduction in the systemic inflammation34 since infections are the most common cause of acute exacerbation7.\n\nThe statistical pooling of the data revealed that 500 mg dose of Azithromycin is superior to 250 mg dose, however this results is not enough to draw a conclusion that it is more beneficial, since there is only one study included in the meta-analysis that used a 500 mg dose with only 92 patients in contrast to 1256 patients that were included in the studies that used a 250 mg dose.\n\nMoreover, In the meta-analysis, no difference was found regarding the development of adverse events between the Azithromycin group and the Placebo group even at 500 mg dose and for 12 months’ therapy duration, in contrast to a previous meta-analysis, which revealed that nonfatal adverse events (gastrointestinal reactions, ototoxicity, rash, and liver injury) were associated with the Placebo group. In addition, hearing decrement occurred more frequently in participants receiving Azithromycin compared to those receiving Placebo, but the improvements in hearing that occurred on repeat testing, suggested that hearing decrements were overestimated in both groups. Gastrointestinal adverse events were the second most common encountered, conversely, they were the most common side effects reported in two previous studies28,29. nevertheless, the usage of serial audiometry was only in one study while the other included studies report no usage of audiometry in assessing hearing decrement among the participants, this might have led to underreporting of adverse events in these studies, which may in turn have led eventually to this result in the meta-analysis. Additionally, Albert et al. and Simpson et al. used a daily dose of 250mg Azithromycin, but Berkhof et al. used a three times a week dose of 250mg Azithromycin, while Uzun et al. used a three times a week dose of 500mg Azithromycin. The meta-analysis did not differentiate between intermittent and daily dosing, which might affect the results of the subgroup analysis of dosage and adverse events.\n\nIn summary, this study revealed that, long-term Azithromycin administration for COPD patients is not statistically associated with increased risk of developing adverse events, yet three studies did not perform serial audiometry which might have led to under reporting of the adverse events. Furthermore, there is a lack of differentiation between intermittent and daily dosing in the adverse events meta-analysis, these limitations might proscribe such a conclusion. In addition, the results of the pooled data suggest that Azithromycin is effective in reducing the frequency of acute exacerbation of COPD, and the use of Azithromycin may be beneficial for COPD patients. Only four studies were included in this review, more studies are required to confirm our findings especially regarding dosage and adverse events.\n\n\nData availability\n\nHarvard Dataverse: Systematic Revman file. https://doi.org/10.7910/DVN/EQMBCH23\n\nHarvard dataverse, PRISMA checklist for ‘Safety and efficacy of Azithromycin in prevention of chronic obstructive pulmonary disease exacerbation: systematic review and meta-analysis’. https://doi.org/10.7910/DVN/3XMNFR35\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 authors declared that no grants were involved in supporting this work\n\n\nReferences\n\nLoscalzo J, Wiener C, Brown C, et al.: Harrison's principles of internal medicine. 19th ed. New York: McGraw-Hill Education; 2015; 1700–1707. Reference Source\n\nChronic obstructive pulmonary disease (COPD): World Health Organization. 2018; [cited 10 May 2018]. Reference Source\n\nWho.int: WHO | COPD: Definition. 2018; [Accessed 8 Mar. 2018]. Reference Source\n\nLozano R, Naghavi M, Foreman K, et al.: Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012; 380(9859): 2095–128. PubMed Abstract | Publisher Full Text\n\nFermont J, Wood A, Müllerova H, et al.: P96 Death related to cardiovascular disease in chronic obstructive pulmonary disease. Multi-morbidity in COPD. 2017; 72(Suppl 3). Publisher Full Text\n\nWorld Health Organization: COPD management. World Health Organization; 2013; [cited 2018May10]. Reference Source\n\nGlobal Initiative for Chronic Obstructive Lung Disease - GOLD: Gold Reports 2018 - Global Initiative for Chronic Obstructive Lung Disease - GOLD. 2018; [Accessed 10 Mar. 2018].infections – full version. Clin Microbiol Infect. 2011; 17:E1–E59. Reference Source\n\nMiravitlles M, Grupo de trabajo de la Asociación Latinoamericana del Tórax (ALAT): Update to the Latin American Thoracic Association (ALAT) Recommendations on Infectious Exacerbation of Chronic Obstructive Pulmonary Disease. Arch Bronconeumol. 2004; 40(7): 315–325. PubMed Abstract | Publisher Full Text\n\nBalter MS, La Forge J, Low DE, et al.: Canadian Guidelines for the Management of Acute Exacerbations of Chronic Bronchitis. Can Respir J. 2003; 10 Suppl B: 3B–32B. PubMed Abstract | Publisher Full Text\n\nCelli BR, MacNee W, ATS/ERS Task Force: Standards for the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J. 2004; 23(6): 932–46. PubMed Abstract | Publisher Full Text\n\nBlasi F, Ewig S, Torres A, et al.: A review of guidelines for antibacterial use in acute exacerbations of chronic bronchitis. Pulm Pharmacol Ther. 2006; 19(5): 361–9. PubMed Abstract | Publisher Full Text\n\nShinkai M, Henke MO, Rubin BK: Macrolide antibiotics as immunomodulatory medications: proposed mechanisms of action. Pharmacol Ther. 2008; 117(3): 393–405. PubMed Abstract | Publisher Full Text\n\nAzithromycin. In: National Center for Biotechnology Information. PubChem Compound Database. Accessed 28 May 2018. Reference Source\n\nBanerjee D, Khair OA, Honeybourne D: The effect of oral clarithromycin on health status and sputum bacteriology in stable COPD. Respir Med. 2005; 99(2): 208–215. PubMed Abstract | Publisher Full Text\n\nSuzuki T, Yanai M, Yamaya M, et al.: Erythromycin and common cold in COPD. Chest. 2001; 120(3): 730–733. PubMed Abstract | Publisher Full Text\n\nSeemungal TA, Wilkinson TM, Hurst JR, et al.: Long-term erythromycin therapy is associated with decreased chronic obstructive pulmonary disease exacerbations. Am J Respir Crit Care Med. 2008; 178(11): 1139–1147. PubMed Abstract | Publisher Full Text\n\nYamaya M, Azuma A, Tanaka H, et al.: Inhibitory effects of macrolide antibiotics on exacerbations and hospitalization in chronic obstructive pulmonary disease in Japan: a retrospective multicenter analysis. J Am Geriatr Soc. 2008; 56(7): 1358–1360. PubMed Abstract | Publisher Full Text\n\nHe ZY, Ou LM, Zhang JQ, et al.: Effect of 6 months of erythromycin treatment on inflammatory cells in induced sputum and exacerbations in chronic obstructive pulmonary disease. Respiration. 2010; 80(6): 445–52. PubMed Abstract | Publisher Full Text\n\nBlasi F, Bonardi D, Aliberti S, et al.: Long-term azithromycin use in patients with chronic obstructive pulmonary disease and tracheostomy. Pulm Pharmacol Ther. 2010; 23(3): 200–207. PubMed Abstract | Publisher Full Text\n\nGómez J, Baños V, Simarro E, et al.: [Prospective, comparative study (1994-1998) of the influence of short-term prophylactic treatment with azithromycin on patients with advanced COPD]. Rev Esp Quimioter. 2000; 13(4): 379–383. PubMed Abstract\n\nTaylor SP, Sellers E, Taylor BT: Azithromycin for the Prevention of COPD Exacerbations: The Good, Bad, and Ugly. Am J Med. 2015; 128(12): 1362.e1–6. PubMed Abstract | Publisher Full Text\n\nNicholson TT, Franciosi A, Landers S, et al.: Assessing potential risks of treatment with long-term azithromycin in COPD patients: long-term oxygen users beware? Ir J Med Sci. 2016; 185(4): 993–7. PubMed Abstract | Publisher Full Text\n\nAhmed A: systematic Revman file. Harvard Dataverse, V1. 2019. https://www.doi.org/10.7910/DVN/EQMBCH\n\nAlbert RK, Connett J, Bailey WC, et al.: Azithromycin for Prevention of Exacerbations of COPD. N Engl J Med. 2012; 366: 1356. Publisher Full Text\n\nBerkhof FF, Doornewaard-ten Hertog NE, Uil SM, et al.: Azithromycin and cough-specific health status in patients with chronic obstructive pulmonary disease and chronic cough: a randomised controlled trial. Respir Res. 2013; 14(1): 125. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUzun S, Djamin RS, Kluytmans JA, et al.: Azithromycin maintenance treatment in patients with frequent exacerbations of chronic obstructive pulmonary disease (COLUMBUS): a randomised, double-blind, placebo-controlled trial. Lancet Respir Med. 2014; 2(5): 361–368. PubMed Abstract | Publisher Full Text\n\nSimpson JL, Powell H, Baines KJ, et al.: The effect of azithromycin in adults with stable neutrophilic COPD: a double blind randomised, placebo controlled trial. PLoS One. 2014; 9(8): e105609. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFlorescu DF, Murphy PJ, Kalil AC: Effects of prolonged use of azithromycin in patients with cystic fibrosis: a meta-analysis. Pulm Pharmacol Ther. 2009; 22(6): 467–472. PubMed Abstract | Publisher Full Text\n\nYao GY, Ma YL, Zhang MQ, et al.: Macrolide Therapy Decreases Chronic Obstructive Pulmonary Disease Exacerbation: A Meta-Analysis. Respiration. 2013; 86(3): 254–260. PubMed Abstract | Publisher Full Text\n\nSuzuki T, Yamaya M, Sekizawa K, et al.: Erythromycin inhibits rhinovirus infection in cultured human tracheal epithelial cells. Am J Respir Crit Care Med. 2002; 165(8): 1113–18. PubMed Abstract | Publisher Full Text\n\nMarjanović N, Bosnar M, Michielin F, et al.: Macrolide antibiotics broadly and distinctively inhibit cytokine and chemokine production by COPD sputum cells in vitro. Pharmacol Res. 2011; 63(5): 389–97. PubMed Abstract | Publisher Full Text\n\nHodge S, Hodge G, Jersmann H, et al.: Azithromycin improves macrophage phagocytic function and expression of mannose receptor in chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2008; 178(2): 139–48. PubMed Abstract | Publisher Full Text\n\nHalldorsson S, Gudjonsson T, Gottfredsson M, et al.: Azithromycin maintains airway epithelial integrity during Pseudomonas aeruginosa infection. Am J Respir Cell Mol Biol. 2010; 42(1): 62–68. PubMed Abstract | Publisher Full Text\n\nMarin A, Garcia-Aymerich J, Sauleda J, et al.: Effect of bronchial colonisation on airway and systemic inflammation in stable COPD. COPD. 2012; 9(2): 121–30. PubMed Abstract | Publisher Full Text\n\nAhmed A: PRISMA. Harvard Dataverse, V1. 2019. https://www.doi.org/10.7910/DVN/3XMNFR"
}
|
[
{
"id": "46534",
"date": "10 Apr 2019",
"name": "Bruce K. Rubin",
"expertise": [
"Reviewer Expertise airway biology",
"biomedical engineering",
"airway inflammation and immunity"
],
"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\nNoting that the macrolide antibiotics have immunomodulatory (and not anti-inflammatory) properties, these authors have performed a limited systematic review and meta analysis of one of these medications, azithromycin, for the prevention of exacerbations of COPD. After a search of SCOPUS, PubMed, and ClinicalTrials.gov they identified three studies for the acute exacerbations frequency analysis and concluded that the use of azithromycin for three to six months reduces the frequency of exacerbations; but it is not clear if this is clinically significant, if the effect is greatest in severe or less severe exacerbations, or if the dosage or dosing frequency makes a difference. The authors refer to this as a “research study”. It is in fact a systematic review which not does not directly evaluate the safety or efficacy of long-term administration of azithromycin as the authors claim in the first paragraph of their Abstract.\n\nThe greatest shortcoming of this study is the authors’ failure to perform a complete literature search of clinical trials studying the use of macrolides in the prevention of COPD exacerbations. In October 2018 a Cochrane review by Herath et al.1 on this exact subject, includes studies of all macrolide antibiotics for the prevention of COPD exacerbations. The Cochrane review is not only much more comprehensive and robust in terms of analysis and conclusions, but Ahmed et al., fail to cite this previously published systematic review. Furthermore, there was an additional complete systematic review on this subject published in the JAMA in 20142 also not cited by these authors; strongly suggesting that their literature review was superficial at best. Therefore, although, the current manuscript is a narrowly focused systematic review; given the published Cochrane review this paper adds little to the literature.\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? Partly\n\nAre the conclusions drawn adequately supported by the results presented in the review? Partly",
"responses": []
},
{
"id": "66395",
"date": "14 Jul 2020",
"name": "Miles Weinberger",
"expertise": [],
"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 submission by Ahmed et al. is an attempt at a systematic review and meta-analysis of azithromycin in prevention of exacerbations of chronic obstructive pulmonary disease (COPD). It should be noted that azithromycin should not be capitalized, nor should placebo. A primary limitation of this review includes only 4 studies that differ in the proportion of males and females. The largest of the studies,1142 subjects, had 58% males, the next largest, 92 subjects, had 43% males. A study with 84 subjects had 75% males. This suggests that the 4 studies may not be looking at the same disease. The heterogeneity of COPD is well established1-3. Moreover, another meta-analysis that included 11 studies with 1910 subjects4 was not included in the current submission.\nThe interest of macrolide antibiotics, particularly azithromycin, was the serendipitous discovery of the curative effect of these antibiotics in diffuse panbronchiolitis5 through a mechanism not related to their antibiotic effect. The wide range of immunomodulator effect of azithromycin has resulted in studies of various obstructive lung disease including cystic fibrosis, an asthma phenotype, post-transplant bronchiolitis in addition to non-pulmonary inflammatory diseases6. However, no other disease has been associated with essentially absolute cure as seen with diffuse panbronchiolitis. As such, a report such as this by Ahmed et al. needs to consider the heterogeneity of COPD and whether the 4 studies are looking at subjects with the same COPD. With the number of subjects predominantly from one study and the failure to include a larger systematic review and meta-analysis further limit the relevance of this report. The authors need to expand their source of data and consider the complexity of COPD and whether azithromycin might best target a specific COPD phenotype.\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? I cannot comment. A qualified statistician is required.\n\nAre the conclusions drawn adequately supported by the results presented in the review? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-253
|
https://f1000research.com/articles/7-1334/v1
|
24 Aug 18
|
{
"type": "Research Article",
"title": "Caucasian and south Asian men show equivalent improvements in surrogate biomarkers of cardiovascular and metabolic health following 6-weeks of supervised resistance training",
"authors": [
"Allan Knox",
"Nicholas Sculthorpe",
"Fergal Grace",
"Nicholas Sculthorpe",
"Fergal Grace"
],
"abstract": "Background: The South Asian population have greater cardiovascular risk than their age-matched Caucasian counterparts, characterized by unfavorable biomarkers. South Asians may also be partially resistant to the pleiotropic benefits of physical activity on cardiovascular health. There is a current absence of studies that compare markers of cardio-metabolic health between Caucasians and South Asians employing resistance exercise. This study set out to compare the response in biomarkers of cardio-metabolic health in Caucasians and South Asians in response to resistance exercise. Methods: Caucasian (n=15, 25.5 ± 4.8 yrs) and South Asian (n=13, 25.4 ± 7.0 yrs) males completed a 6-week progressive resistance exercise protocol. Fasting blood glucose, insulin, and their product insulin resistance (HOMA-IR), triglycerides (TRIGS), low density lipoprotein (LDL), high density lipoprotein (HDL), total cholesterol (TC), vascular endothelial growth factor (VEGF), asymmetric dimythylarginine (ADMA), L-arginine (L-ARG) and C-reactive protein (CRP) were established at baseline and following resistance exercise. Results: There were significant improvements in fasting glucose, TC, LDL, HDL and VEGF in both groups following resistance exercise (p<0.05, for all). No change was observed in insulin, HOMA-IR, TRIGS, ADMA, L-ARG following resistance exercise (p>0.05, in both groups). CRP increased in the South Asian group (p<0.05) but not the Caucasian group (p>0.05) Conclusions: The cardio-metabolic response to resistance exercise is comparable in young Caucasian and South Asian males though inflammatory response to exercise may be prolonged in South Asians.",
"keywords": [
"Resistance",
"Strength",
"Exercise",
"Training",
"South",
"Asian",
"Cardiovascular",
"Metabolic"
],
"content": "Introduction\n\nDiseases that centre on the cardiovascular (CVD) and glycolytic systems are primary contributants to annual global mortality, accounting for approximately 17.7 million and 1.6 million deaths in 2015, respectively (WHO, 2017a; WHO, 2017b). However, there is considerable heterogeneity of CVD prevalence across racial groups, where the immigrant South Asian (SA) community of the United Kingdom have an approximate 50–100% elevation in cardio-metabolic risk compared with the general population (Wild et al., 2007). This translates to an estimated 5.3 year earlier occurrence of CVD in SAs when compared with their Caucasian (CAUC) counterparts (Hughes et al., 1989).\n\nPronounced CVD risk amongst SAs is due in part, to the presence of consistently augmented traditional risk factors. For instance, unfavourable lipid profiles been reported in SAs in comparison to CAUCs (Misra & Khurana, 2011), due to inherent variation of lipid particles. SAs exhibit smaller low density lipoprotein (LDL) particles and dysfunctional high density lipoprotein (HDL) particles compared with CAUCs (Dodani, 2008; Tziomalos et al., 2008). Premature development of Type 2 diabetes (and associated complications) occur approximately 10 years earlier in SAs than CAUCs, and data from large-scale studies such as United Kingdom Asian Diabetes Study (UKADS) reported diabetes related deaths in SAs occurring 7.4 years earlier than CAUCs (Bellary et al., 2010). Comparably higher systemic C-reactive protein (CRP) with regional obesity of the abdomen is another aspect of cardio-metabolic disease that contrives to increase CVD in SAs when compared with CAUCs. The latter has required revisions in the ‘normal’ BMI thresholds for overweight and obese SAs (Gray et al., 2011). These racially specific reclassifications are based on data demonstrating that comparable levels of glycaemia and lipid profiles are seen at BMI levels of 21–26 kg/m2 in SAs compared to 30 kg/m2 in CAUCs (Gray et al., 2011).\n\nPhysical exercise is a well-known prophylactic for the development of CVD. In this respect, the cardio protective effects of aerobic exercise are well documented in comparison to fewer studies of resistance exercise (RES). RES has been shown to promote healthogenic effects by improving lipid profiles (Kelley & Kelley, 2009; Sheikholeslami Vatani et al., 2011), insulin sensitivity (Mann et al., 2014; Schwingshackl et al., 2014), and CRP levels (Donges et al., 2010). However, these studies almost exclusively enrol CAUC participants with the result that there is a general paucity of RES studies and resultant data amongst the SA population.\n\nFew investigations have included SA participants and have used lifestyle modification methodologies which have produced modest results. For instance, a 3-year lifestyle intervention, promoting beneficial dietary and physical activity behaviours in SAs report trivial effects in body mass and no change in either blood pressure or fasting blood glucose (Bhopal et al., 2014). Other investigations that have studied the interaction between physical activity and cardio-metabolic risk in SAs have also found encouraging results. Data from the Health Survey of England show an inverse association between the participation of 30 minutes of moderate intensity exercise per week and cardio-metabolic risk in CAUCs and SAs (Williams et al., 2011). However, findings from the Indian Diabetes Prevention Programme and the ADDITION-Leicester study identified a blunted response in cardio-metabolic biomarkers to physical activity in the SA cohort compared with CAUCs (Yates et al., 2010), which corroborates the findings from other racial comparison studies (Fretts et al., 2009; Steinbrecher et al., 2012). Recently, IIiodromiti and colleagues (2016) have made a valuable addition to the literature by offering that SAs would require ~50% greater participation in physical activity than CAUCs (232 min/week-1 vs 150 min/week-1 in CAUCs) to address the imbalance in healthogenenic benefits of aerobic exercise between these racial groups.\n\nWhile these data may have potentially important consequences for exercise prescription for SAs, the field is limited by the dearth of objective assessment of exercise participation. Our previous work has demonstrated differences in muscular strength adaptation in response to supervised RES in SAs, which may translate to an attenuated cardio-metabolic response (Knox et al., 2017). These findings are in line with two previous studies which compared objectively determined physical activity (Iliodromiti et al., 2016) and aerobic exercise (Hall et al., 2010) which support the potential for a blunted response to physical activity in SAs compared with CAUCs. However, no study has explored this phenomenon using the medium RES. Therefore, the purpose of this study was to assess whether the different response of exercise in the SA population suggested by previous reports is applicable to RES. We will address this aim by establishing the response of traditional cardio-metabolic biomarkers to 6-weeks progressive RES between young, healthy CAUC and SA males. We hypothesised that 1) no difference in any biomarkers of cardio-metabolic health are evident at baseline (PRE) between CAUCs and SAs, 2) no difference in biochemical markers in both groups following RES (POST), 3) both groups do not differ at POST in any biochemical marker of cardio-metabolic health.\n\n\nMethods\n\nThis study was carried out in accordance with the recommendations of the University of the West of Scotland. The protocol was approved by the School of Science and Sport ethics committee (approval: HREC_Sci2013/02/Knox). All subjects gave written informed consent in accordance with the Declaration of Helsinki. In the present study, the main outcome measure was squat performance. A power calculation was performed using G*Power V3 with reference to previously published data regarding squat strength in healthy, untrained, young men (MacDonald et al., 2012). A single-tailed within-group comparison revealed a required sample size of 15 per group (alpha set to 0.05 and power at 0.95). Therefore, the presented data should be considered as hypothesis generating. All participants were recruited by local advertisement and word of mouth. Inclusion criteria were that (i) participants did not engage in any recreational or competitive sports, (ii) were naive to RES prior to study enrolment, (iii) no history of cardio-metabolic disease prior to study participation (iv) or any illness that may provide a contraindication to exercise participation. Race and familial generation of United Kingdom (UK) patriotage was self-reported.\n\nAll anthropometric measurements were performed by AK in the Human Performance Laboratory of the University of the West of Scotland (HPLUWS). Height was determined using a portable stadiometer (Leicester Height Measure, Seca, Birmingham, U.K.). Body mass was measured to the nearest 0.1 kg using commercially available scales (TBF-300, Tanita, Tokyo, Japan), where participants were required to remove footwear and unnecessary clothing before measurement which were obtained in adherence with manufacturers guidelines. Body mass index was calculated using the following formula; BMI = body mass (kg) ÷ height (m)2. Total body fat percentage (%) was calculated using bioelectrical impedance analysis (BIA) using a commercially available analyser (body composition analyser TBF-300, Tanita, Tokyo, Japan). Criterion validity of this method has been suitably comparable (r = 0.952) compared with the gold standard dual energy x-ray absorptiometry (DEXA) (Bozkirli et al., 2007).\n\nWaist circumference (WC) and waist-hip ratio (WHR) was calculated according to the guideline published by the WHO (WHO, 2008) using a commercially available ergonomic circumference measuring tape (Seca 201, Seca, Birmingham). Participants were asked to remove any clothing on their upper body before standing with their arms by their side with their feet positioned close together with their body weight distributed evenly across both feet. For WC, the tape was placed around the approximate midpoint between the lower palpable rib and the top of the iliac crest (WHO, 2008). The participant was then asked to relax and take several deep breaths to account for any abdominal tension caused by the nature of this procedure where criterion measurement was taken at the end of normal expiration to control for any diaphragm movements. Hip circumference (HC) was calculated by measuring the circumference of the widest portion of the buttocks.\n\nSystemic blood pressure data was collected by AK in the HPLUWS. Arterial blood pressure (BP) was measured by an automated BP monitor (M6, Omron, Milton Keynes, U.K.). Participants sat in a quiet room for 10 minutes with their legs uncrossed and with the arm being used for measurement supported approximately level with the heart. A total of 3 readings of systolic blood pressure (SBP) and diastolic blood pressure (DBP) were recorded. The final data was calculated from the mean values of the three reading. From these values, it was possible to calculate mean arterial pressure (MAP) the following formula; MAP = 0.33(SBP-DBP) + DBP.\n\nAll blood sampling and analysis was performed by AK in the HPLUWS. To account for diurnal variation in blood variables, participant blood was sampled at the same time of day and following a 12-hour fast. Briefly, each participant remained in the supine position for at least 30 minutes during sample collection enabling control of plasma volume shifts (Hagan et al., 1978). Venous blood was collected from the antecubital vein via venepuncture. Blood glucose, triglycerides (TRIGS), HDL and total cholesterol (TC) were determined by spectrophotometery (Rx Monza, Randox Laboratories, UK). Samples were mixed with respective reagents and incubated at room temperature for the duration stated by reagent manufacturer (Randox Laboratories, UK) before being inserted into the spectrophotometer and read at the corresponding wavelength. LDL concentrations were established indirectly by the following Friedwalde equation (Friedewald et al., 1972); LDL = TC - HDL - (TRIGS/5). This method was appropriate as all samples did not exceed fasting TRIGS concentrations of 4.52 mmol/L and were free from chylomicrons and hyperlipidaemia. Insulin (Alpco Diagnostics Salem, USA: catalogue number: 80-INSHU-E01.1), vascular endothelial growth factor (VEGF: Invitrogen, Life Technologies, USA: catalogue number: BMS277-2), asymmetric dimythylarginine (ADMA) and L-arginine (Diagnostika GMBH, Hamburg, Germany: catalogue number: EA207/192) concentrations were established using commercially available ELISA kits. CRP was determined by commercially available immunoturbidimetric assay (Rx Monza, Randox laboratories, UK: catalogue number: CP3885). Insulin resistance was calculated from the standard formula: (fasting glucose × fasting insulin)/22.5. A small subsample (n=6 from both groups) was used to establish changes in ADMA and L-arginine for hypothesis generating purposes.\n\nMuscular strength assessments were performed by AK in the HPLUWS. All training sessions were performed in local authority fitness centres under the supervision of AK. Measurement of muscular strength, prescribed training protocol, and associated data has been previously reported (Knox et al., 2017). Briefly, the protocol utilised 5 main compound exercises including back squats, bench press, deadlifts, shoulder press, and lateral pull down and progressed in a linear approach three times per week. Each exercise was prescribed at 3 sets of 10 repetitions with 2 minutes’ rest between sets. The training programme was separated into two sessions; Session A and Session B, with each being performed consecutively throughout the duration of the study. Training days were separated by at least one day but no longer than two days.\n\nData were analysed using SPSS version 24.0 for Windows. Group and time interactions, main effects of time and simple main effects of time were determined by a mixed model ANOVA with repeated measures and Bonferonni correction. Distribution of data was confirmed by Shapiro-Wilks’ test. Homogeneity of variances were assessed using Levenes’ test for homogeneity of variance. Homogeneity of covariance’s was established by Box’s test of equality covariance matrices. Assumptions were unviolated (p>0.05) unless otherwise stated. When Mauchlys’ test of Spericity were violated (p<0.05), a Greenhouse-Geisser estimate was reported. ANOVA are presented as F test [(degrees of freedom, error terms degrees of freedom) F value, p value, partial eta-squared (pɳ2)]. Post-hoc data are presented as differences in mean (M), 95% CI, p value. A chi-squared test for association was conducted between groups and impaired fasting glucose levels at PRE and POST. Data is presented as group mean + standard deviation (SD). An alpha of p<0.05 was used to indicate statistical significance.\n\n\nResults\n\nThirty-eight males (n=19 CAUCs, n=19 SAs) were screened and deemed eligible for participation and consequently enrolled. All participants completed PRE measures and initiated supervised RES. Eight participants did not perform POST measures due to non-compliance of the protocol (n=4 CAUCs, n=4 SAs) and two had to withdraw due to personal circumstances (n=2 SAs). The final analysis consisted of 15 CAUC (25.5 ± 4.8 years) and 13 SA (25.4 ± 7.0 years) participants. Because 2 SAs participants withdrew from the study due to personal reasons (bereavement), we elected not to conduct an intention to treat analysis (See Figure 1).\n\nNo significant group × time interactions were observed following RES for; height, BM, BMI, %BF, FM, FFM, WC, HC, WHR (p > 0.05 for all). No statistically significant main effects of time following RES were identified in any anthropometric measurement (p > 0.05). A simple main effect of time for HC was detected in the CAUC group (f(1, 24) = 4.38, p = 0.047, pɳ2 = 0.15) with a significant decrease from PRE-to-POST (M = 3.62 cm, 95% CI: 0.52 – 7.18, p = 0.047). No significant main or simple effects of group were observed in any anthropometric measure apart from WHR, where there was a significant main effect of group (f(1, 24) = 5.46, p = 0.028, pɳ2 = 0.19). The CAUC group presented with a significantly higher WHR at PRE (M = 0.52, 95% CI: 0.002 – 0.102, p = 0.041) and POST (M = 0.057, 95% CI: 0.006 – 0.107, p = 0.029). (See Table 1).\n\nData are presented as group mean ± SD. a - p<0.05 from baseline. b - p<0.05 between groups at corresponding time point. Abbreviations: BM – body mass (kg), BMI – body mass index (kg/m2), FM – fat mass (kg), FFM – fat free mass (kg), SBP – systolic blood pressure (mmHg), DBP – diastolic blood pressure (mmHg), MAP – mean arterial pressure (mmHg), HOMA-IR – insulin resistance, HDL – high density lipoprotein (mg/dL), LDL – low density lipoprotein (mg/dL), TC – total cholesterol (mg/dL), TRIGS – triglycerides (mg/dL), VEGF – vascular endothelial growth factor (pg/ml), ADMA – asymmetric dimethylarginine (μmol/L), L-ARG – L-arginine (μmol/L), CRP – C-reactive protein (mg/L), CAUC – Caucasian, SA – South Asian, PRE – baseline, POST – following resistance training intervention.\n\nNo group × time interaction (f(1, 25) = 0.327, p = 0.572, pɳ2 = 0.01), main effects of time (f(1, 25) = 0.002, p = 0.96, pɳ2 = 0.01) or main effects of group were observed (f = (1, 25) = 0.01, p = 0.909, pɳ2 = 0.001) in SBP following RES. Neither was there any change within in the CAUC (M = 1.04 mmHg, 95% CI: -4.63 – 6.69, p = 0.710) or SA groups (M = 1.23mmHg, 95% CI: -4.64 – 7.11, p = 0.670) following RES as both groups were similar at PRE (M = 1.49 mmHg, 95% CI: -5.15 – 7.14, p = 0.724) and POST (M = 0.76 mmHg, 95% CI: -5.81 – 7.35, p = 0.812) RES. There was a group × time interaction in DBP following RES (f(1, 25) = 5.95, p = 0.022, pɳ2 = 0.20). No main effects of time (f(1, 25) = 0.83, p = 0.373, pɳ2 = 0.03), or main effects of group (f(1, 25) = 0.04, p = 0.85, pɳ2 = 0.001) were evident in DBP following RES. The CAUC group reduced DBP (M = 5.39 mmHg, 95% CI: 0.69 – 10.08 mmHg, p = 0.026), with no difference in the SA group (M = 2.46 mmHg, 95% CI: -2.23 – 7.16 mmHg, p = 0.290). Both groups were similar at PRE (M = 3.39 mmHg, 95% CI: -2.72 – 9.49 mmHg, p = 0.264) and POST (M = 4.46 mmHg, 95% CI: -2.93 – 11.85 mmHg, p = 0.225).\n\nA trend was seen for a group × time interaction (f(1, 25) = 3.03, p = 0.094, pɳ2 = 0.11) in MAP following RES. No main effects of time (f(1, 25) = 0.14, p = 0.72, pɳ2 = 0.01) or main effects of group (f(1, 25) = 0.17, p = 0.687, pɳ2 = 0.01) was seen in MAP following RES. No change was seen in the CAUC (M = 3.21 mmHg, 95% CI: 1.14 – 7.57 mmHg, p = 0.141) or SA (M = 2.09 mmHg, 95% CI: 1.43 – 6.61 mmHg, p = 0.349) group and both groups were similar at PRE (M = 1.47 mmHg, 95% CI: -5.33 – 8.27 mmHg, p = 0.660) and POST (M = 3.84 mmHg, 95% CI: 2.88 – 10.55 mmHg, p = 0.250).\n\nNo group × time interactions were observed in any variable following RES (p > 0.05 for all comparisons). There was a main effect of time in glucose (f(1, 16) = 10.16, p = 0.006, pɳ2 = 0.39), TC (f(1, 21) = 23.12, p = <0.001, pɳ2 = 0.52), HDL (f(1, 17) = 30.42, p = <0.001, pɳ2 = 0.64), LDL (f(1, 16) = 40.39, p = <0.001, pɳ2 = 0.72) and VEGF (f(1, 14) = 12.74, p = 0.003, pɳ2 = 0.48) following RES. Amongst the CAUC group, there were significant reductions in fasting blood glucose (M = 0.77 mmol/L, 95% CI: 0.03 – 1.51mmol/L, p = 0.041), TC (M = 42.09 mg/dL, 95% CI: 11.26 – 72.92mg/dL, p = 0.010) and LDL (M = 47.84 mg/dL, 95% CI: 24.44 – 71.23mg/dL, p = 0.001), improvements in HDL (M = 10.335 mg/dL, 95% CI: 5.49 – 15.17 mg/dL, p = <0.001) and VEGF (M = 33.51 pg/ml, 95% CI: 6.62 – 60.41mg/dL, p = 0.018) concentrations (see Figure 2). SA participants enjoyed comparable improvements in fasting blood glucose (M = 0.79 mmol/L, 95% CI: 0.06 – 1.53 mmol/L, p = 0.036) TC (M = 60.98 mg/dL, 95% CI: 28.77 – 93.18 mg/dL, p = 0.001) and LDL (M = 57.38 mg/dL, 95% CI: 31.22 – 83.53 mg/dL, p = <0.001) and increases in HDL (M = 9.16 mg/dL, 95% CI: 3.49 – 14.83 mg/dL, p = 0.003), VEGF (M = 29.77 pg/ml, 95% CI: 2.88 – 56.68 pg/ml, p = 0.032) and CRP (M = 3.54 mg/L, 95% CI: 0.07 – 7.01 mg/L, p = 0.046). Figure 3 shows the metabolic response of both groups.\n\nBoth groups presented favourable lipid responses following resistance exercise. a – p<0.05 from baseline. HDL – high density lipoprotein (mg/dL), LDL – low density lipoprotein (mg/dL), TC – total cholesterol (mg/dL).\n\nGlucose concentrations were significantly lower in both groups following resistance exercise. a – p<0.05 from baseline. HOMA-IR – insulin resistance.\n\nNo main effects of group were seen in any biochemical marker (p > 0.05 for all). A trend was observed for a main effect of group on IR (f(1, 15) = 3.91, p = 0.067, pɳ2= 0.207) and CRP (f(1, 15) = 3.22, p = 0.093, pɳ2 = 0.177). Both groups were similar at PRE for glucose (M = 0.03 mmol/L, 95% CI: -0.88 – 0.95 mmol/L, p = 0.943), insulin (M = 1.11 U/ml, 95% CI: -1.01 – 3.23 U/ml, p = 0.281), TC (M = 41.97 mg/dL, 95% CI: -18.38 – 102.32 U/ml, p = 0.163), HDL (M = 2.17 mg/dL, 95% CI: -7.64 – 11.96 mg/dL, p = 0.647), LDL (M = 1.53 mg/dL, 95% CI: -37.92 – 40.97 mg/dL, p = 0.936), TRIGS (M = 8.92 mg/dL, 95% CI: -25.34 – 43.18mg/dL, p = 0.594), VEGF (M = 7.99 pg/ml, 95% CI: -20.69 – 36.68 pg/ml, p = 0.560), L-ARG (M = 8.92 μmol/L, 95% CI: -23.39 – 41.24 μmol/L, p = 0.552), ADMA (M = 0.06 μmol/L, 95% CI: -0.09 – 0.20 μmol/L, p = 0.376) and CRP (M = 0.57 mg/L, 95% CI: -2.69 – 3.83 mg/L, p = 0.717). A trend was seen at PRE for IR between groups (M = 0.51, 95% CI: -0.10-1.12, p = 0.096).\n\nAt POST, both groups were comparable for glucose (M = 0.06 mmol/L, 95% CI: -0.98 – 1.10 mmol/L, p = 0.910), insulin (M = 0.45 μ/ml, 95% CI: -1.79 – 2.69 μ/ml, p = 0.676), TC (M = 23.09 mg/dL, 95% CI: -13.64 – 59.81 mg/dL, p = 0.205), HDL (M = 3.34 mg/dL, 95% CI: -7.36 – 14.05 mg/dL, p = 0.519), LDL (M = 8.01 mg/dL, 95% CI: -11.67 – 27.69 mg/dL, p = 0.401), TRIGS (M = 17.02 mg/dL, 95% CI: -31.62 – 65.66 mg/dL, p = 0.475), VEGF (M = 4.28 pg/ml, 95% CI: -39.76 – 48.28 pg/ml, p = 0.839), L-ARG (M = 4.05 μmol/L, 95% CI: -12.96 – 21.07 μmol/L , p = 0.608), ADMA (M = 0.04 μmol/L, 95% CI: -0.04 – 0.12 μmol/L, p = 0.317) with a trend evident for IR (M = 0.55, 95% CI: -0.10 – 1.22, p = 0.090). The SA group presented higher CRP levels at POST (M = 3.61 mg/L, 95% CI: 0.29 – 6.93 mg/L, p = 0.035, see Figure 4). Figure 5 illustrates the percentage change of biomarkers in both groups.\n\nC-reactive protein concentrations increased in South Asians following resistance exercise, with no change observed within the Caucasian group. a – p<0.05 from baseline, b – p<0.05 between groups.\n\nData are presented as means ± SD. a - p<0.05 within groups. b - p<0.05 between groups. Abbreviations: IR – insulin resistance, HDL – high density lipoprotein (mg/dL), LDL – low density lipoprotein (mg/dL), TC – total cholesterol (mg/dL), TRIGS – triglycerides (mg/dL), VEGF – vascular endothelial growth factor (pg/ml), ADMA – asymmetric dimethylarginine (μmol/L), L-ARG – L-arginine (μmol/L), CRP – C-reactive protein (mg/L).\n\nA significant group × time interaction was evident for lower body strength (f(1,26) = 11.23, p = 0.002, pɳ2 = 0.30). A main effect of time was also observed (f(1,26) = 349.64, p < 0.001, pɳ2 = 0.93). The CAUC (M = 61.33 Kg, 95% CI: 53.54 – 69.13 Kg, p < 0.001) and SA (M = 42.69 Kg, 95% CI: 34.32 – 51.06 Kg, p < 0.001) groups significantly improved lower body strength. Both groups presented similar lower body strength at PRE (M = 4.82 Kg, 95% CI: -8.21 – 17.85 Kg, p = 0.454) however, the CAUC group showed demonstrated higher strength at POST than the SA group (M = 23.46 Kg, 95% CI: 5.28 – 41.65 Kg, p = 0.013).\n\nNo group × time interaction was observed for upper body strength (f(1,26) = 0.52, p = 0.476, pɳ2 = 0.02). A main effect of time was seen (f(1,26) = 112.89, p < 0.001, pɳ2 = 0.81). The CAUC (M = 13.67 Kg, 95% CI: 10.29 – 17.04 Kg, p < 0.001) and SA (M = 11.92 Kg, 95% CI: 8.30 – 15.55 Kg, p < 0.001) both improved upper body strength. Both groups were similar in upper body strength at PRE (M = 2.18 Kg, 95% CI: -10.50 – 14.86 Kg, p = 0.727) and POST (M = 0.44 Kg, 95% CI: -11.01 – 11.88 Kg, p = 0.94).\n\n\nDiscussion\n\nThe main findings of the current study demonstrate comparable responses of biomarkers of cardio-metabolic health between CAUC and SA males in response to short-term progressive RES. These responses were evident despite significant differences in muscular strength adaptation. Both groups significantly improved lipid profiles, fasting glucose and VEGF following RES. Although, no difference was observed in fasting TRIGS, insulin, L-ARG, ADMA, or IR following RES in any group, with no apparent discrepancy between groups. CRP increased in the SA group following RES with no change evident in CAUCs.\n\nPrevious findings have reported that increasing physical activity levels is beneficial for cardio-metabolic risk in SAs (Dhawan & Bray, 1997; Eriksen et al., 2015; Mathews et al., 2007) (Dhawan & Bray, 1997; Eriksen et al., 2015; Mathews et al., 2007). Dhawan & Bray (1997) showed marked differences in physical activity levels between CAUCs and SAs, however increased physical activity was associated with reductions in insulin, BMI, TRIGS, and blood pressure in both groups. Similarly, Mathews et al. (2007) reported improvements in body composition, blood pressure, and lipid profiles following the Khush Dil initiative which involved participation in physical activity and nutritional workshops. Data from the SABRE study reported positive associations between physical activity and markers of cardio-metabolic health in SAs (Eriksen et al., 2015). However, conflicting data suggest a blunted response to physical activity in the SAs cohort.\n\nData from the Indian Diabetes Prevention Programme (Ramachandran et al., 2006) show no change in body mass following diet and physical activity advice in native Indians, although the intervention group did have a significantly lower incidence of diabetes development at year 3 than the controls. The lack of change in physiological markers are in agreement with later data that show no positive associations with increasing levels of physical activity and body composition in SAs (Yates et al., 2010). These data are supported by a more recent intervention which resulted in mixed body composition responses in SAs (Bhopal et al., 2014).\n\nThese studies not only have a key limitation of self-reported measures of physical activity, but also provide difficulty in establishing a dose-response relationship between racial groups to an exercise stimulus. The current study, to our knowledge, is the first to demonstrate a comparable cardio-metabolic response to RES between CAUC and SA males. These data provide evidence that RES is an effective method to improve cardio-metabolic health in the high-risk SA population, with no apparent discrepancy in response when compared with CAUCs. Considering previous investigations have reported a blunted response to aerobic exercise in SAs, the current data provides intriguing evidence that RES interventions may prove superior to aerobic exercise to improve cardio-metabolic health in young SAs. Further research is warranted to determine this theory.\n\nThe limited exercise response suggested by the results from previous associations of physical activity and cardio-metabolic health in SAs has initiated several comparative studies. Hall et al. (2010) demonstrated that SAs have a reduced capacity to oxidise fat during submaximal aerobic exercise in comparison with age and body composition matched CAUCs. These data suggest the SAs may need to participate in longer durations or higher exercise intensities to elicit the same physiological benefit from exercise. More recent evidence showing a blunted response in SAs from the same research group used age-adjusted regression models to determine the association between moderate intensity physical activity and cardio-metabolic risk (Iliodromiti et al., 2016). This study reports that middle-aged SA men and women need to participate in 232 min/week-1 to elicit comparable cardio-metabolic benefits as CAUCs who perform the currently recommended 150 min/week-1. Taken these data together, they suggest a blunted response to exercise in the SA cohort. However, the current investigation demonstrates a comparable cardio-metabolic response to RES between young CAUC and SA males. Further study is warranted to establish if the ability to utilise metabolic substrates during RES is varied between racial groups. The lack of RES studies within the SA population provides difficulty to establish a comparison of existing data which potentially can contribute to our understanding of optimal RES prescription. Nevertheless, the current study demonstrates that progressive RES is effective at improving cardio-metabolic health in young CAUCs and SAs. These data also demonstrate the importance of targeting younger SA individuals to reduce the risk of developing cardio-metabolic diseases that older generations have suffered.\n\nCRP has been regarded as an emerging risk factor for the development of cardiovascular diseases that involve systemic inflammation (Singh et al., 2008). It has been reported that SA males have a 17% higher concentration of CRP than matched CAUC males, with differences seen in individuals as young as 10 years old (Chambers et al., 2001; Cook et al., 2000). The current literature concerning the effect of RES on circulating CRP levels is mixed. RES has been shown to reduce CRP levels in African American males with no difference observed in White Americans (Heffernan et al., 2009). These results may have been a consequence of differential baseline values as the African American group presented higher levels of CRP. Additional work has also reported no change in CRP levels following 10 weeks of RES in individuals with metabolic risk factors (Levinger et al., 2009). The current study is in agreement with previous research as no changes were seen in the CAUC group. The lack of change in CRP levels following RES within the CAUC group may have been a consequence of no statistical change in body composition as there is a correlation between CRP levels and BMI (Vlachopoulos et al., 2015). However, a significant increase was seen in the SA group. This may suggest that adequate recovery from a training bout may take longer in the SA group. Previous studies have reported abnormal heart rate recovery from a graded exercise test to be positively associated with CRP levels, independent of physical fitness and disease status (Jae et al., 2007). Although the current data cannot provide evidence on recovery rates, an abnormal recovery from an exercise bout may be present in the SA population. This requires further exploration.\n\nThe current findings may have important clinical implications. According to current guidelines (American Diabetes Association, 2007), both groups demonstrated impaired fasting glucose at baseline. RES resulted in a significant improvement in both groups to normal fasting glucose levels, suggesting that RES protocols similar to that employed in the current study is sufficient to reduce the risk of type 2 diabetes development. The significant improvement in lipid profiles also demonstrates the efficacy of RES in improving cardio-metabolic health in young CAUC and SA males. Both groups demonstrated significant improvements in TC, LDL and HDL concentrations, which may translate to a significant reduction in cardiovascular risk. This is of importance as our previous work has demonstrated differing rates of adaptation to exercise training between CAUCs and SAs (Knox et al., 2017), which may imply differing rate of adaptation in other physiological markers. However, the current study has demonstrated that this is not evident in biomarkers of cardio-metabolic health.\n\nThis study does not come without several limitations. The small sample size of both racial groups may not provide sufficient statistical power to determine differences in the biomarkers that lacked disparity between groups in response to RES. Nevertheless, the current data would suggest that the biomarkers associated with cardio-metabolic health would respond similarly between CAUCs and SAs in response to RES. The small sample size also makes these data difficult to generalise to whole populations. Additionally, as previous literature using self-reported physical activity levels has suggested a blunted response to exercise in the SA population, a larger sample size that includes participants of similar age to existing literature would be necessary for validation. A larger profile of biomarkers would have also been necessary. There is strong evidence of emerging risk factors such as plasminogen activator inhibitor-1, lipoprotein A, homocysteine, tumor necrosis factor A and several adhesion molecules that are all linked with the development of cardiovascular disease. Interestingly, these compounds seem to be more prevalent in SA individuals when compared to CAUCs (Gupta et al., 2006). Inclusion of these measures would have given greater detail in the efficacy of RES in the prevention of cardiovascular risk between CAUCs and SAs. Future research concerning the SA population and RES should consider these emerging risk factors to develop the limited knowledge of exercise response within this population.\n\nIn conclusion, supervised progressive RES is effective at improving biomarkers of cardio-metabolic health within young CAUC and SA males, with no apparent discrepancy in response to exercise training. SAs significantly increased CRP levels which may indicate a requirement of a longer recovery following exercise. However, this theory has yet to be confirmed. These data provide an evidence base for interventions to target the younger generation of SAs which may assist in the reduction of future cardio-metabolic diseases that has been prevalent in the SA population for decades.\n\n\nData availability\n\nUnderlying data for this study is available from figshare: Dataset 1. Short-Term Resistance Exercise and Cardiometabolic Health in Caucasian and South Asian Males, https://doi.org/10.6084/m9.figshare.6741731.v1 (Knox et al., 2018)\n\nData available under CC0 1.0 Universal licence",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAmerican Diabetes Association: Diagnosis and classification of diabetes mellitus. Diabetes Care. 2007; 30 Suppl 1: S42–47. PubMed Abstract | Publisher Full Text\n\nBellary S, O'Hare JP, Raymond NT, et al.: Premature cardiovascular events and mortality in south Asians with type 2 diabetes in the United Kingdom Asian Diabetes Study - effect of ethnicity on risk. Curr Med Res Opin. 2010; 26(8): 1873–1879. PubMed Abstract | Publisher Full Text\n\nBhopal RS, Douglas A, Wallia S, et al.: Effect of a lifestyle intervention on weight change in south Asian individuals in the UK at high risk of type 2 diabetes: a family-cluster randomised controlled trial. Lancet Diabetes Endocrinol. 2014; 2(3): 218–227. PubMed Abstract | Publisher Full Text\n\nBozkirli E, Ertorer ME, Bakiner O, et al.: The validity of the World Health Organisation's obesity body mass index criteria in a Turkish population: a hospital-based study. Asia Pac J Clin Nutr. 2007; 16(3): 443–447. PubMed Abstract | Publisher Full Text\n\nChambers JC, Eda S, Bassett P, et al.: C-reactive protein, insulin resistance, central obesity, and coronary heart disease risk in Indian Asians from the United Kingdom compared with European whites. Circulation. 2001; 104(2): 145–150. PubMed Abstract\n\nCook DG, Mendall MA, Whincup PH, et al.: C-reactive protein concentration in children: relationship to adiposity and other cardiovascular risk factors. Atherosclerosis. 2000; 149(1): 139–150. PubMed Abstract | Publisher Full Text\n\nDhawan J, Bray CL: Asian Indians, coronary artery disease, and physical exercise. Heart. 1997; 78(6): 550–554. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDodani S: Excess coronary artery disease risk in South Asian immigrants: can dysfunctional high-density lipoprotein explain increased risk? Vasc Health Risk Manag. 2008; 4(5): 953–961. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDonges CE, Duffield R, Drinkwater EJ: Effects of resistance or aerobic exercise training on interleukin-6, C-reactive protein, and body composition. Med Sci Sports Exerc. 2010; 42(2): 304–313. PubMed Abstract | Publisher Full Text\n\nEriksen A, Tillin T, O'Connor L, et al.: The impact of health behaviours on incident cardiovascular disease in Europeans and South Asians--a prospective analysis in the UK SABRE study. PLoS One. 2015; 10(3): e0117364. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFretts AM, Howard BV, Kriska AM, et al.: Physical activity and incident diabetes in American Indians: the Strong Heart Study. Am J Epidemiol. 2009; 170(5): 632–639. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFriedewald WT, Levy RI, Fredrickson DS: Estimation of the Concentration of Low-Density Lipoprotein Cholesterol in Plasma, Without Use of the Preparative Ultracentrifuge. Clin Chem. 1972; 18(6): 499–502. PubMed Abstract\n\nGray LJ, Yates T, Davies MJ, et al.: Defining obesity cut-off points for migrant South Asians. PLoS One. 2011; 6(10): e26464. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGupta M, Singh N, Verma S: South Asians and cardiovascular risk: what clinicians should know. Circulation. 2006; 113(25): e924–929. PubMed Abstract | Publisher Full Text\n\nHagan RD, Diaz FJ, Horvath SM: Plasma volume changes with movement to supine and standing positions. J Appl Physiol Respir Environ Exerc Physiol. 1978; 45(3): 414–417. PubMed Abstract | Publisher Full Text\n\nHall LM, Moran CN, Milne GR, et al.: Fat oxidation, fitness and skeletal muscle expression of oxidative/lipid metabolism genes in South Asians: implications for insulin resistance? PLoS One. 2010; 5(12): e14197. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHeffernan KS, Jae SY, Vieira VJ, et al.: C-reactive protein and cardiac vagal activity following resistance exercise training in young African-American and white men. Am J Physiol Regul Integr Comp Physiol. 2009; 296(4): R1098–1105. PubMed Abstract | Publisher Full Text\n\nHughes LO, Raval U, Raftery EB: First myocardial infarctions in Asian and white men. BMJ. 1989; 298(6684): 1345–1350. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIliodromiti S, Ghouri N, Celis-Morales CA, et al.: Should Physical Activity Recommendations for South Asian Adults Be Ethnicity-Specific? Evidence from a Cross-Sectional Study of South Asian and White European Men and Women. PLoS One. 2016; 11(8): e0160024. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJae SY, Ahn ES, Heffernan KS, et al.: Relation of heart rate recovery after exercise to C-reactive protein and white blood cell count. Am J Cardiol. 2007; 99(5): 707–710. PubMed Abstract | Publisher Full Text\n\nKelley GA, Kelley KS: Impact of progressive resistance training on lipids and lipoproteins in adults: a meta-analysis of randomized controlled trials. Prev Med. 2009; 48(1): 9–19. PubMed Abstract | Publisher Full Text\n\nKnox A, Sculthorpe N, Baker JS, et al.: Strength adaptation to squat exercise is different between Caucasian and South Asian novice exercisers. Res Sports Med. 2017; 25(3): 373–383. PubMed Abstract | Publisher Full Text\n\nKnox A, Sculthorpe N, Grace F: Short-Term Resistance Exercise and Cardiometabolic Health in Caucasian and South Asian Males. figshare. Dataset. 2018. http://www.doi.org/10.6084/m9.figshare.6741731.v1\n\nLevinger I, Goodman C, Peake J, et al.: Inflammation, hepatic enzymes and resistance training in individuals with metabolic risk factors. Diabet Med. 2009; 26(3): 220–227. PubMed Abstract | Publisher Full Text\n\nMacDonald CJ, Lamont HS, Garner JC: A comparison of the effects of 6 weeks of traditional resistance training, plyometric training, and complex training on measures of strength and anthropometrics. J Strength Cond Res. 2012; 26(2): 422–431. PubMed Abstract | Publisher Full Text\n\nMann S, Beedie C, Balducci S, et al.: Changes in insulin sensitivity in response to different modalities of exercise: a review of the evidence. Diabetes Metab Res Rev. 2014; 30(4): 257–268. PubMed Abstract | Publisher Full Text\n\nMathews G, Alexander J, Rahemtulla T, et al.: Impact of a cardiovascular risk control project for South Asians (Khush Dil) on motivation, behaviour, obesity, blood pressure and lipids. J Public Health (Oxf). 2007; 29(4): 388–397. PubMed Abstract | Publisher Full Text\n\nMisra A, Khurana L: Obesity-related non-communicable diseases: South Asians vs White Caucasians. Int J Obes (Lond). 2011; 35(2): 167–187. PubMed Abstract | Publisher Full Text\n\nRamachandran A, Snehalatha C, Mary S, et al.: The Indian Diabetes Prevention Programme shows that lifestyle modification and metformin prevent type 2 diabetes in Asian Indian subjects with impaired glucose tolerance (IDPP-1). Diabetologia. 2006; 49(2): 289–297. PubMed Abstract | Publisher Full Text\n\nSchwingshackl L, Missbach B, Dias S, et al.: Impact of different training modalities on glycaemic control and blood lipids in patients with type 2 diabetes: a systematic review and network meta-analysis. Diabetologia. 2014; 57(9): 1789–1797. PubMed Abstract | Publisher Full Text\n\nSheikholeslami Vatani D, Ahmadi S, Ahmadi Dehrashid K, et al.: Changes in cardiovascular risk factors and inflammatory markers of young, healthy, men after six weeks of moderate or high intensity resistance training. J Sports Med Phys Fitness. 2011; 51(4): 695–700. PubMed Abstract\n\nSingh U, Dasu MR, Yancey PG, et al.: Human C-reactive protein promotes oxidized low density lipoprotein uptake and matrix metalloproteinase-9 release in Wistar rats. J Lipid Res. 2008; 49(5): 1015–1023. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSteinbrecher A, Erber E, Grandinetti A, et al.: Physical activity and risk of type 2 diabetes among Native Hawaiians, Japanese Americans, and Caucasians: the Multiethnic Cohort. J Phys Act Health. 2012; 9(5): 634–641. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTziomalos K, Weerasinghe CN, Mikhailidis DP, et al.: Vascular risk factors in South Asians. Int J Cardiol. 2008; 128(1): 5–16. PubMed Abstract | Publisher Full Text\n\nVlachopoulos C, Xaplanteris P, Aboyans V, et al.: The role of vascular biomarkers for primary and secondary prevention. A position paper from the European Society of Cardiology Working Group on peripheral circulation: Endorsed by the Association for Research into Arterial Structure and Physiology (ARTERY) Society. Atherosclerosis. 2015; 241(2): 507–532. PubMed Abstract | Publisher Full Text\n\nWHO: Waist Circumferance and wasit-hip ratio: Report of a WHO expert consultation. (8–11 December 2008 ed.). Geneva: World Health Organisation. 2008. Reference Source\n\nWHO: Cardiovascular Diseases (CVD). 2017a. Reference Source\n\nWHO: Diabetes. 2017b. Reference Source\n\nWild SH, Fischbacher C, Brock A, et al.: Mortality from all causes and circulatory disease by country of birth in England and Wales 2001–2003. J Public Health (Oxf). 2007; 29(2): 191–198. PubMed Abstract | Publisher Full Text\n\nWilliams ED, Stamatakis E, Chandola T, et al.: Physical activity behaviour and coronary heart disease mortality among South Asian people in the UK: an observational longitudinal study. Heart. 2011; 97(8): 655–659. PubMed Abstract | Publisher Full Text\n\nYates T, Davies MJ, Gray LJ, et al.: Levels of physical activity and relationship with markers of diabetes and cardiovascular disease risk in 5474 white European and South Asian adults screened for type 2 diabetes. Prev Med. 2010; 51(3–4): 290–294. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "38063",
"date": "21 Sep 2018",
"name": "Eliza Prodel",
"expertise": [
"Reviewer Expertise Exercise physiology",
"cardiology",
"integrative 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\nIn the paper entitled “Caucasian and south Asian men show equivalent improvements in surrogate biomarkers of cardiovascular and metabolic health following 6-weeks of supervised resistance training” you had the aim to investigate the cardio-metabolic response to resistance exercise training in South Asian and Caucasian population. The work is well written, and presents very interesting data, however a have a few issues to clarify.\n\nMajor The whole introduction is based on the fact that SA’s are cardio-metabolic different from CAUCC’s, so why you hypothesized no differences would be found in the results?\n\nWhy only male participants were enrolled in the study? You would have more complete results and an avenue to discuss if women were tested.\n\nData of muscular strength gain could be more clear in a table or graph.\n\nIt is not mandatory, however I feel that the authors could explore more about cardiovascular data and their relationship with the metabolic data. For example, changing in rest and recovery heart rate and heart rate variability with the RES training.\n\nMethods Participants section:\nPerformance measurements (upper and lower body) should be in a specific section, and would be appropriate to provide more detailed information. “A single-tailed within-group comparison revealed a required sample size of 15 per group (alpha set to 0.05 and power at 0.95). Therefore, the presented data should be considered as hypothesis generating.” Should be in statistical analysis section.\n\nMinor Page 5: “Neither was there any change within in the CAUC”. Small writing mistake. Page 8: “however, the CAUC group showed demonstrated higher strength at POST than the SA group”. Small writing mistake.\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?\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? Yes",
"responses": [
{
"c_id": "4448",
"date": "05 Mar 2019",
"name": "Allan Knox",
"role": "Author Response",
"response": "The authors thank you for taking the time to review our manuscript. We have considered all comments and provided our responses below; Comment: The whole introduction is based on the fact that SA’s are cardio-metabolic different from CAUCC’s, so why you hypothesized no differences would be found in the results? Response: We have clarified our hypothesis to reflect the background information that is reported in the introduction of the manuscript. Comment: Why only male participants were enrolled in the study? You would have more complete results and an avenue to discuss if women were tested. Response: There is paucity of experimental exercise research in south Asian males and none in south Asian females. We sought to focus on the male population to control for the possibility of an effect of sex in responses to resistance exercise training. We agree that more data is required on exercise responses in south Asian females, something that we wish to explore in a more comprehensive sample in future experimental work. Comment: Data of muscular strength gain could be more clear in a table or graph. Response: We thank you and appreciate your comment regarding the request for a graph illustrating muscular strength gains. We feel it is unfeasible to provide additional graphs for strength gains due to the number of muscle groups trained. All data regarding strength gains is presented in tables. In addition, these tables are more informatory presented as means and standard deviations due to the differences in strength response. Comment: It is not mandatory, however I feel that the authors could explore more about cardiovascular data and their relationship with the metabolic data. For example, changing in rest and recovery heart rate and heart rate variability with the RES training. Response: We fully appreciate this comment and are considering exploring these data in more detail. Comment: Performance measurements (upper and lower body) should be in a specific section, and would be appropriate to provide more detailed information. Response: We are constrained by the word count available to provide additional information regarding the performance measures. We have referenced previous research that has full details of all procedures. Comment: “A single-tailed within-group comparison revealed a required sample size of 15 per group (alpha set to 0.05 and power at 0.95). Therefore, the presented data should be considered as hypothesis generating.” Should be in statistical analysis section. Response: This has now been moved to the statistical analysis section. Comment: Page 5: “Neither was there any change within in the CAUC”. Small writing mistake Response: Amended in manuscript. Page 8: “however, the CAUC group showed demonstrated higher strength at POST than the SA group”. Small writing mistake. Response: Amended in manuscript."
}
]
},
{
"id": "41730",
"date": "11 Dec 2018",
"name": "Majid Ghayour-Mobarhan",
"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 manuscript by Knox et al., tries to compare the response in biomarkers of cardio-metabolic health in Caucasians and South Asians in response to resistance exercise. The study appears to have been done in a careful and competent manner. I have few comments that require further clarification.\nMethod part: Why were samples selected from male participants in the study? Please explain more in this regard. While more details of method will provide to allow replication by others.\nResults and discussion parts: Also please include other anthropometric indices as CVD risk factors such as the waist circumference-to-hip circumference ratio, ect in the text, if it is possible because it has been proposed to assess central obesity and as measures for cardio-metabolic disorders and events. Therefore, insert and explain the associations them in the result and discussion sections.\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?\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/7-1334
|
https://f1000research.com/articles/7-1818/v1
|
20 Nov 18
|
{
"type": "Correspondence",
"title": "N-terminal pro-brain natriuretic peptide: a potential follow-up biomarker of mandibular advancement device efficacy on cardiac function in obstructive sleep apnea",
"authors": [
"Denis Monneret"
],
"abstract": "Interrelationships between obstructive sleep apnea (OSA) and cardiovascular diseases are now recognized, but some underlying pathophysiological mechanisms remain controversial. Circulating cardiac biomarkers are diagnostic tools that can help understand them, in particular the N-terminal pro-brain natriuretic peptide (NT-proBNP), a marker of myocardial stretch, and a potential indicator of subclinical cardiac stress in OSA. Continuous positive airway pressure (CPAP), the first-line treatment of moderate to severe OSA, may be considered as uncomfortable, resulting in poor adherence, and reduced effectiveness. In this case, mandibular advancement devices (MAD) are an effective alternative therapy, more comfortable, and generally well accepted, with higher compliance. To date, few studies have compared the cardiovascular effects of CPAP and MAD. From recent literature reviews, it emerges that both therapies are effective in blood pressure reduction. However, the effects of MAD on other cardiovascular outcomes are conflicting, in particular as regards to its impact on circulating cardiac biomarkers. In a recent ancillary study from a randomized controlled trial, Recoquillon et al concluded that two months of MAD treatment had no effect on NT-proBNP plasma levels in patients with severe OSA. The present discussion analyses this result from a biological, statistical, and analytical standpoint, in light of results from other studies evaluating natriuretic peptides in MAD-treated OSA, with the aim to support further longitudinal studies designed with a high methodological quality.",
"keywords": [
"N-terminal pro-brain natriuretic peptide",
"mandibular advancement device",
"obstructive sleep apnea",
"cardiac biomarker"
],
"content": "\n\nMandibular advancement devices (MADs) are an effective alternative to continuous positive airway pressure (CPAP) in the treatment of obstructive sleep apnea (OSA). OSA is associated with increased cardiovascular morbidity and mortality, and an increasing number of studies highlight the efficacy of MADs in terms of both sleep apnea, and cardiac outcomes1–3. Unlike CPAP-related studies, few studies to date have focused on cardiac biomarkers under MAD therapy in OSA.\n\nIn a recent randomized controlled trial, Recoquillon et al. evaluated the effect of two months of MAD treatment on N-Terminal Pro-Brain Natriuretic Peptide (NT-proBNP) plasma levels in patients with severe OSA4. Compared to a sham device, the high treatment adherence (6.6 hours/night) significantly reduced the mean apnea-hypopnea index (AHI), and the oxygen desaturation index. Nevertheless, according to their model, the authors stated that MADs had no effect on NT-proBNP levels, nor on other inflammatory and metabolic biomarkers. To our knowledge, to date only two studies have investigated the natriuretic peptides in such contexts5,6. Given their scarcity, any type of study looking at relevant cardiac biomarkers of MAD efficacy must be encouraged, and designed in as detailed and robust a manner as possible. In this way, some issues have to be discussed regarding the evaluation of NT-proBNP from Recoquillon et al.\n\n1) Biological standpoint.\n\nAfter two months of MAD use, the NT-proBNP plasma concentrations decreased from 296.8 to 252.5 pg/mL (–14.9%) in treated patients, whereas they decreased from 189.8 to 184.3 pg/mL (–2.9%) in patients with the sham device, resulting in a mean adjusted intergroup difference of 12.0 pg/mL (–40.9 to 64.9, 95%CI; P =0.65). The question arises as to whether a NT-proBNP decrease of about 15% after treatment is biologically significant. Indeed, according to the specifications of the desirable biological variation database7, this decrease should be considered as significant according to the within-subject biological variation (CVi 10%), but not significant according to the between-subject biological variation (CVg 16%). In any case, this decrease should be considered as analytically significant since it exceeds the allowable limit of total error, which combines the analytical imprecision and the inter-method inaccuracy, fixed at 13% for NT-proBNP. Moreover, one could argue that a longer treatment period, even one extra month, could be sufficient to significantly lower its circulating level. In support of this assumption, Hoekema et al. showed a significant decrease in NT-proBNP (‒58%, P =0.035) in ten patients with moderate to severe OSA treated by MAD (adherence 6.8 hours/night, 6.9 nights/week) after a period of 69 to 82 days5. For these ten patients, baseline (52 pg/mL, interquartile range (IQR): 13‒105), and follow-up NT-proBNP values (22 pg/mL, IQR: 15‒33) were within or close to the reference intervals established according to the method8, and were thus in accordance with exclusion criteria discarding patients with a history of cardiovascular disease (CVD). Unlike Hoekema et al., NT-proBNP values from Recoquillon et al. reached 500 to 700 pg/mL, i.e. much higher than the normal values announced by the manufacturer (median 47 pg/mL, min–max: 3.9–155 pg/mL9). This is somewhat in contradiction with the exclusion criteria supposed to discard patients with a history of CVD, including heart failure4. Another study showed a significant decrease of plasma BNP levels (‒24%) after 6 months of MAD therapy in patients with stable, mild to moderate congestive heart failure (CHF), and OSA6. Although less stable in plasma than NT-proBNP10, BNP, the other biomarker of CHF, is still widely and routinely assayed on analyzers in hospital laboratories, and thus remains of potential interest for the follow-up of cardiac function under MAD treatment.\n\n2) Statistical standpoint.\n\nIn the supplemental statistical section, Recoquillon et al. mentioned that variables with non-continuous distributions are described as median (IQR)4. However, NT-proBNP results were expressed as mean (standard deviation (SD)), and reached 296.8 (401.6) pg/mL. Such a wide SD suggests a strong skewness of distribution. Median (IQR) expression would therefore have been more appropriate, and a graph detailing the scatter dot plots for both groups, with connecting lines before and after treatment, would have been required. Moreover, a linear regression analysis was used for the adjustment of baseline values and potential covariates: age, gender, body mass index, and baseline AHI. Nevertheless, these covariates were used for the adjustment of all biomarkers, but no statistical proof was provided as regards to their degree of correlation with NT-proBNP specifically. Furthermore, given the limited number of patients (n ±55), if NT-proBNP results were not normally distributed (as seems to be the case), nonparametric ANCOVA or robust regression methods would probably have been more appropriate11–14.\n\n3) Analytical standpoint.\n\nRecoquillon et al. assayed plasma NT-proBNP using a multiplex electrochemiluminescent immunoassay on a MESO QuickPlex® SQ120 analyzer (MSD, Rockville, USA). Using this technology for assaying this cardiac biomarker is somewhat unusual. Indeed, as reminded by the manufacturer, this method is for research use only, but not for use in diagnostic or therapeutic procedures. It involves three incubation steps, interspersed with three wash sequences, requiring at least five hours of preparation for one 96-well plate. To our knowledge, no studies based on this assay have been published up to now, not even the eight references cited in the MSD technical sheet of the human NT-proBNP assay kit9. Given the long and tedious assay protocol, which is impractical in hospital laboratory routine, and given the absence of hindsight about its analytical performance, the authors should rather have used a most widespread, reliable, and rapid automated method, like the electrochemiluminescent immunoassay method on Roche analyzers (Roche Diagnostics, Mannheim, Germany)15. In this way, an interesting perspective is the ongoing MOSAIC study, whose main objective is to assess the impact of three months of MAD therapy on AHI in Asian patients with heart failure and OSA16. Indeed, one of the planned secondary objectives is the evaluation of cardiac remodeling and of cardiac biomarkers, including NT-proBNP. Meanwhile, the present standpoints remind and emphasize the need for close collaborations between sleep specialists and laboratory practitioners to strengthen the methodological quality and robustness of studies involving biomarkers.\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\nAcknowledgements\n\nThe author is grateful to Vincent Fitzpatrick for his English proof-reading.\n\n\nReferences\n\nSharples LD, Clutterbuck-James AL, Glover MJ, et al.: Meta-analysis of randomised controlled trials of oral mandibular advancement devices and continuous positive airway pressure for obstructive sleep apnoea-hypopnoea. Sleep Med Rev. 2016; 27: 108–24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBratton DJ, Gaisl T, Schlatzer C, et al.: Comparison of the effects of continuous positive airway pressure and mandibular advancement devices on sleepiness in patients with obstructive sleep apnoea: a network meta-analysis. Lancet Respir Med. 2015; 3(11): 869–78. PubMed Abstract | Publisher Full Text\n\nBratton DJ, Gaisl T, Wons AM, et al.: CPAP vs Mandibular Advancement Devices and Blood Pressure in Patients With Obstructive Sleep Apnea: A Systematic Review and Meta-analysis. JAMA. 2015; 314(21): 2280–93. PubMed Abstract | Publisher Full Text\n\nRecoquillon S, Pépin JL, Vielle B, et al.: Effect of mandibular advancement therapy on inflammatory and metabolic biomarkers in patients with severe obstructive sleep apnoea: a randomised controlled trial. Thorax. 2018; pii: thoraxjnl-2018-212609. PubMed Abstract | Publisher Full Text\n\nHoekema A, Voors AA, Wijkstra PJ, et al.: Effects of oral appliances and CPAP on the left ventricle and natriuretic peptides. Int J Cardiol. 2008; 128(2): 232–9. PubMed Abstract | Publisher Full Text\n\nEskafi M, Cline C, Nilner M, et al.: Treatment of sleep apnea in congestive heart failure with a dental device: the effect on brain natriuretic peptide and quality of life. Sleep Breath. 2006; 10(2): 90–7. PubMed Abstract | Publisher Full Text\n\nRicós C, Alvarez V, Cava F, et al.: Current databases on biological variation: pros, cons and progress. Scand J Clin Lab Invest. 1999; 59(7): 491–500, Updated database available on http://www.westgard.com/biodatabase1.htm (accessed in November 3rd, 2018). PubMed Abstract | Publisher Full Text\n\nGalasko GI, Lahiri A, Barnes SC, et al.: What is the normal range for N-terminal pro-brain natriuretic peptide? How well does this normal range screen for cardiovascular disease? Eur Heart J. 2005; 26(21): 2269–76. PubMed Abstract | Publisher Full Text\n\nTechnical sheet of Human NT-proBNP Assay Kit, Multi-Array® Assay, Meso Scale Discovery®. (accessed in November 3rd, 2018). Reference Source\n\nYeo KT, Wu AH, Apple FS, et al.: Multicenter evaluation of the Roche NT-proBNP assay and comparison to the Biosite Triage BNP assay. Clin Chim Acta. 2003; 338(1–2): 107–15. PubMed Abstract | Publisher Full Text\n\nCarlsson MO, Zou KH, Yu CR, et al.: A comparison of nonparametric and parametric methods to adjust for baseline measures. Contemp Clin Trials. 2014; 37(2): 225–33. PubMed Abstract | Publisher Full Text\n\nAustin PC, Manca A, Zwarenstein M, et al.: A substantial and confusing variation exists in handling of baseline covariates in randomized controlled trials: a review of trials published in leading medical journals. J Clin Epidemiol. 2010; 63(2): 142–53. PubMed Abstract | Publisher Full Text\n\nPocock SJ, Assmann SE, Enos LE, et al.: Subgroup analysis, covariate adjustment and baseline comparisons in clinical trial reporting: current practice and problems. Stat Med. 2002; 21(19): 2917–30. PubMed Abstract | Publisher Full Text\n\nWang SJ, Hung HM: Adaptive covariate adjustment in clinical trials. J Biopharm Stat. 2005; 15(4): 605–11. PubMed Abstract | Publisher Full Text\n\nCollinson PO, Barnes SC, Gaze DC, et al.: Analytical performance of the N terminal pro B type natriuretic peptide (NT-proBNP) assay on the Elecsys 1010 and 2010 analysers. Eur J Heart Fail. 2004; 6(3): 365–8. PubMed Abstract | Publisher Full Text\n\nMandibular Advancement Device for Treatment of Obstructive Sleep Apnea and Its Impact on Cardiac Remodeling (MOSAIC). ClinicalTrials.gov Identifier: NCT02948894, (accessed in November 3rd, 2018). Reference Source"
}
|
[
{
"id": "42053",
"date": "17 Jan 2019",
"name": "Gen-Min Lin",
"expertise": [
"Reviewer Expertise Sleep medicine and cardiovascular 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\nIn general, the comments were written well, as the cited paper is a randomised control trial design and the author made a critical appraisal. Since many readers might not have time to review the original paper, it might be helpful to write a summary before the comments. Otherwise no further comments were made; and the report is suitable to be accepted in current version.\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": []
},
{
"id": "43027",
"date": "31 Jan 2019",
"name": "Micha Maeder",
"expertise": [
"Reviewer Expertise Heart failure",
"valve disease",
"coronary disease",
"OSA"
],
"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 is a relatively extensive comment on a short original manuscript (brief communication) published in Thorax. The key finding of the paper by Recoquillon et al.1 was a lack of effect of effective OSA treatment by mandibular advancement therapy on NT-proBNP. This finding is in the centre of the comment by Denis Monneret.\nFrom my point of view the result was far away from being statistically significant, and the observed (non-significant) difference may have been the result of chance. Therefore, I cannot fully follow the argument listed under “biological standpoint”. The key problem however is the fact that it is currently unknown what BNP and NT-proBNP reflect in patients with OSA. The literature on this issue has been summarized in detail by Maeder MT et al. in Clin Chim Acta (2016)2, and the overall summary is that the data is conflicting. No study has in detail evaluated whether NT-proBNP in OSA reflects cardiac dysfunction or whether NT-proBNP is mainly under the influence of obesity in OSA. Variable effects of CPAP on BNP and NT-proBNP in uncontrolled studies have been observed. In the present manuscript I would really like the author to mention these aspects.\nWith regards to the statistical standpoint I fully agree that BNP and NT-proBNP always display a skewed distribution, and that this should have been taken into account. When using non-parametric tests, the result is even more unlikely to be statistically significant.\nWith regards to the analytical standpoint I also agree: the use of an established easy to use NT-proBNP assay would have made it easier to check the plausibility of the data and to compare it with the existing literature.\n\nIs the rationale for commenting on the previous publication clearly described? Yes\n\nAre any opinions stated well-argued, clear and cogent? Partly\n\nAre arguments sufficiently supported by evidence from the published literature or by new data and results? Partly\n\nIs the conclusion balanced and justified on the basis of the presented arguments? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1818
|
https://f1000research.com/articles/8-251/v1
|
04 Mar 19
|
{
"type": "Opinion Article",
"title": "Advancing the international data science workforce through shared training and education",
"authors": [
"John Van Horn",
"Sumiko Abe",
"José Luis Ambite",
"Teresa K. Attwood",
"Niall Beard",
"Louisa Bellis",
"Avnish Bhattrai",
"Alex Bui",
"Gully Burns",
"Lily Fierro",
"Jonathan Gordon",
"Jeffrey Grethe",
"Jeana Kamdar",
"Xiaoyu Lei",
"Kristina Lerman",
"Annette McGrath",
"Nicola Mulder",
"Caroline O'Driscoll",
"Crystal Stewart",
"Sonika Tyagi",
"Sumiko Abe",
"José Luis Ambite",
"Teresa K. Attwood",
"Niall Beard",
"Louisa Bellis",
"Avnish Bhattrai",
"Alex Bui",
"Gully Burns",
"Lily Fierro",
"Jonathan Gordon",
"Jeffrey Grethe",
"Jeana Kamdar",
"Xiaoyu Lei",
"Kristina Lerman",
"Annette McGrath",
"Nicola Mulder",
"Caroline O'Driscoll",
"Crystal Stewart",
"Sonika Tyagi"
],
"abstract": "The increasing richness and diversity of biomedical data types creates major organizational and analytical impediments to rapid translational impact in the context of training and education. As biomedical data-sets increase in size, variety and complexity, they challenge conventional methods for sharing, managing and analyzing those data. In May 2017, we convened a two-day meeting between the BD2K Training Coordinating Center (TCC), ELIXIR Training/TeSS, GOBLET, H3ABioNet, EMBL-ABR, bioCADDIE and the CSIRO, in Huntington Beach, California, to compare and contrast our respective activities, and how these might be leveraged for wider impact on an international scale. Discussions focused on the role of i) training for biomedical data science; ii) the need to promote core competencies, and the ii) development of career paths. These led to specific conversations about i) the values of standardizing and sharing data science training resources; ii) challenges in encouraging adoption of training material standards; iii) strategies and best practices for the personalization and customization of learning experiences; iv) processes of identifying stakeholders and determining how they should be accommodated; and v) discussions of joint partnerships to lead the world on data science training in ways that benefit all stakeholders. Generally, international cooperation was viewed as essential for accommodating the widest possible participation in the modern bioscience enterprise, providing skills in a truly “FAIR” manner, addressing the importance of data science understanding worldwide. Several recommendations for the exchange of educational frameworks are made, along with potential sources for support, and plans for further cooperative efforts are presented.",
"keywords": [
"biomedicine",
"data science",
"training",
"education",
"personalized",
"standards",
"collaboration",
"global partnerships"
],
"content": "1. Introduction\n\nThe ability to reap the wealth of information contained in large-scale biomedical data holds the promise to further our understanding of human health and disease. Biomedical big data come from many sources, from massive stand-alone data-sets generated by large collaborations to small data-sets produced by individual investigators. The value of all these data can be amplified through aggregation and integration. However, the diversity of such data types creates major organizational and analytical impediments to rapid translational impact. As biomedical data-sets increase in size, variety and complexity, they challenge conventional methods for sharing, managing and analyzing those data. Furthermore, researchers’ abilities to capitalize on approaches based on biomedical data science are limited by poor data accessibility and interoperability, the lack of appropriate tools, and insufficient training (Brazas et al., 2017).\n\nIn response to the opportunities and challenges presented by the era of “big data” in biological and biomedical research, the National Institutes of Health (NIH) launched the Big Data to Knowledge (BD2K) initiative in 2015 (Van Horn, 2016). Earlier, in 2006, the European Strategy Forum for Research Infrastructures (ESFRI) published its first European Roadmap for Research Infrastructures. As part of this initiative, a 5-year pan-European preparatory-phase project was launched in 2007 to create a European Life-Science Infrastructure for Biological Information (ELIXIR); ELIXIR entered its operational phase in 2013, tasked to construct a world-class, globally-positioned European infrastructure for the management and integration of life-science information. Recognizing the need for training to support the development and use of the bioinformatics tools and databases disseminated by such infrastructures, the Global Organisation for Bioinformatics Learning, Education and Training (GOBLET) was established in 2012, to cultivate the global bioinformatics trainer community, set standards, and provide high-quality training resources; as part of its mission to professionalize training, GOBLET worked closely with ELIXIR to develop a joint training strategy in 2015 (Attwood et al., 2015). The H3Africa initiative, funded by the National Institutes of Health and Wellcome Trust, was established in 2012, when the first projects were launched to investigate the genetic basis of diseases in Africa. H3ABioNet was established as a pan-African bioinformatics network for H3Africa to build capacity for genomics research on the continent. Similarly, a range of programs in Australia, such as EMBL-ABR (Schneider et al., 2017) and BPA/CSIRO training network, were also established. Collectively, these programs seek to nurture the training elements of the digital research enterprise within biomedicine, to facilitate “discovery science”, to support the identification of new knowledge, and to capitalize on community engagement.\n\nIn May 2017, during a two-day meeting between the BD2K Training Coordinating Center (TCC), ELIXIR Training/TeSS, GOBLET, H3ABioNet, EMBL-ABR, bioCADDIE and the CSIRO, convened in Huntington Beach, California, to describe our respective activities and how these might be leveraged for wider impact on a global scale. Subsequent interactions between participants at international events associated with each group have reinforced a focus on the following questions:\n\nWhat is the role of training for biomedical data science, promoting core competencies and the development of career paths?\n\nIs there value in standardizing and sharing data science training resources?\n\nWhat challenges exist in encouraging adoption of training material standards?\n\nHow should resources be personalized or customized to the individual, and what ways would work best?\n\nWho are the stakeholders in this enterprise, and how should they be accommodated?\n\nHow can we jointly partner to lead the world on data science training in ways that are seen as positive from the point of view of the stakeholders?\n\nIn what follows, we summarize the themes of our discussions, each of our individual efforts, and driving questions important for our interactions. We propose several key processes to advance international best practices in training standards and programs.\n\n\n2. Training in biomedical data science\n\nEducation and training are fundamental components of many major data science initiatives worldwide. For example, the BD2K has funded training activities that include a variety of grant mechanisms: data science educational courses and Open-Educational-Resource (OER) grants, data science training programs, young researcher career development awards, and training components of major BD2K research centers. This arguably represents the largest commitment ever undertaken by the NIH toward data science education. Their intention to achieve a new, forward-thinking goal of broader data science for biomedicine will require: 1) bringing together individuals involved in training and career-development programs to network and share experiences; 2) engaging individuals developing OERs in meaningful dialogue to ensure that the resources they develop will be freely available and easy to access by the broader biomedical community; and 3) developing an online educational resource index for the creation of personalized learning paths through linked educational resources.\n\nWith this in mind, the BD2K TCC has been established (Bui et al., 2017) to pursue the following: 1) coordinate activities across the BD2K Training Consortium to enable the exchange of ideas and best practices in data science training, both within BD2K and in the broader biomedical research community; 2) facilitate the discovery, access and citation of educational resources through the development of a living educational resource discovery index (ERuDIte) (Ambite et al., 2017; Van Horn et al., 2018); 3) personalize the discovery of biomedical data science educational resources; 4) facilitate outreach and engagement with the data science training community to identify and hold relevant workshops; 5) support biomedical research training collaborations through short-term rotations into biomedical data science labs; and 6) evaluate and summarize these supported activities. In addition to sponsoring and coordinating meetings, workshops and rotations in biomedical data science, the TCC seeks to develop technology and a user interface for the BD2K’s Training Consortium to communicate activity, and, for the broader biomedical community, to provide personalized access to educational resources.\n\nELIXIR unites Europe’s leading life-science organizations in managing and safeguarding the increasing volumes of data being generated by publicly-funded research. It coordinates, integrates and sustains bioinformatics resources across its member states, and enables users in academia and industry to access services that are vital for their research. To inform researchers about associated bioinformatics training courses and materials, and where in Europe (and the rest of the world) they can be found, ELIXIR has developed a Training eSupport System (TeSS; Beard et al., 2016; Larcombe et al., 2017). ELIXIR also develops and deploys training (both face-to-face and via e-learning approaches) to researchers, developers, infrastructure operators, and to trainers themselves (Morgan et al., 2017); the training focuses on topics identified as knowledge or skill gaps, whether in specific ELIXIR Use Cases or within the general ELIXIR community. This training program aims to both transform scientists into effective users of ELIXIR’s data, tools, standards and compute infrastructure, and to increase the number of capable data science trainers.\n\nGOBLET (Attwood et al., 2015) was established as a Dutch Foundation during ELIXIR’s preparatory phase. It was created as an umbrella organization (incorporating international and national networks and societies, research and academic groups, etc.) to unite, inspire and equip bioinformatics trainers worldwide – ultimately, to serve as a professional body for bioinformatics training. To this end, GOBLET aims to provide a global, sustainable support infrastructure for the international community of bioinformatics1 trainers (which includes teachers in high-schools) and trainees. Part of the Foundation’s work has involved developing a portal (Corpas et al., 2015) for sharing training materials and publications (including guidelines and best practice documents (Via et al., 2013)), and resources to help train trainers. The portal is now one of the core content providers for ELIXIR’s TeSS. Alongside ELIXIR and other organizations, GOBLET also monitors bioinformatics skill gaps and training needs, and plays a role in advocacy for the introduction of bioinformatics training into core life-science curricula (Attwood et al., 2017; Brazas et al., 2017). GOBLET has a particular interest in developing standards, both to help drive up the quality of training materials and courses, and to facilitate their discovery. In this latter context, GOBLET works closely with ELIXIR and Bioschemas to harmonize metadata standards and avoid duplication of effort.\n\nThe vision of H3Africa has been to create and support a pan-continental network of laboratories that will be equipped to apply leading-edge research to the study of the complex interplay between environmental and genetic factors that determine disease susceptibility and drug responses in African populations. Data generated from their efforts should inform strategies to tackle health inequity and ultimately lead to health benefits in Africa. To achieve this, the following issues are being addressed: 1) ensuring access to, and education on, relevant genomic technologies for African scientists; 2) facilitating integration between genomic and clinical studies; 3) facilitating training at all levels, and, in particular, training research leaders; and 4) establishing necessary research infrastructure and providing training in its use. H3ABioNet, the bioinformatics network for H3Africa, has led an extensive and diverse training program to build skills in bioinformatics and genomic data analysis. H3ABioNet uses a variety of training approaches, including face-to-face and distance-based courses to teach a variety of audiences.\n\nEMBL Australia Bioinformatics Resource (EMBL-ABR) is a developing national research infrastructure, providing bioinformatics resources and support to life science and biomedical researchers in Australia. As a collaboration with the European Bioinformatics Institute (EMBL-EBI), it was established to maximize Australia’s bioinformatics capability. This close partnership is made possible by Australia’s associate membership of EMBL.\n\nEMBL-ABR aims to: 1) increase Australia’s capacity to collect, integrate, analyze, share and archive the large heterogeneous data-sets that are now part of modern life science research; 2) contribute to the development of, and provide training in, data, tools and platforms to enable Australia’s life science researchers to undertake research in the age of big data; 3) showcase Australian research and data-sets at an international level; 4) enable engagement in international programs that create, deploy and develop best practice approaches to data management, software tools and methods, computational platforms and bioinformatics services. EMBL-ABR is a truly national resource that consists of a Hub and eleven nodes, which are organized around six key areas: Data, Tools, Compute, Standards, Training and Platforms, mapped to their respective expertise in terms of bioscience domains. EMBL-ABR is a member of GOBLET and has contributed to the development of Bioschemas, and works closely with the EBI and other ELIXIR Nodes to exchange and share resources (e.g., using the TeSS portal and Bio.tools utilities). EMBL-ABR uses these international connections to 1) provide Hub workshops on crucial topics not covered elsewhere in national bioinformatics training, 2) bring expertise to Australia through visits by international experts and improved access to training resources, and 3) help coordinate and disseminate node end-user training activities.\n\nThe bioCADDIE team has worked to develop a Data Discovery Index (DDI) prototype which, like the TCC’s ERuDIte, indexes primary research data that are stored elsewhere. The DDI seeks to play an important role in promoting data integration through the adoption of content standards and alignment to common data elements and high-level schema. bioCADDIE intends to provide the means to test the utility of these standards against its DDI, thus serving as an incubator for spurring the types of quality metrics that are currently being developed around article metrics, including citation analysis and other metrics of resource utilization.\n\nFinally, the Commonwealth Scientific and Industrial Research Organization (CSIRO) is the federal government agency for scientific research in Australia. Its chief role is to benefit the community by improving the economic and social performance of industry. Through a diverse set of flagship programs, the CSIRO works with leading organizations around the world, and, with CSIRO Publishing, issues journals with the latest research by leading scientists. Data science education and training play a central role in many CSIRO activities and, since 2012, has forged a very successful training partnership with BioPlatforms Australia (BPA) to deliver national bioinformatics training in the life sciences.\n\nBy joining forces in creating and adopting training metadata standards, these distinct initiatives would help reach critical mass to expedite standards adoption, and would avoid the risk of having “yet another, but not quite compatible” standard. One such beneficiary of these resources, and the standards that underpin them, is H3Africa, who are ideally positioned to reach out to a broad spectrum of international scientists requiring advanced training in the most recent scientific methods and approaches.\n\nWe seek to adopt a collaborative approach to enhance mechanisms for creating a tight coupling between life-science/biomedical training resources and their underlying tools and data-sets, for scientific audiences throughout the US, Europe, Australia and Africa. This work will permit the deployment and use of data science methods in biomedical applications to keep up with the overwhelming pace of methodological development in data science.\n\nBuilding and sharing virtual social environments for collaboration enables massive-scale analysis of primary bioscience data. Exploration of the means for interoperability between such systems will be an important point of discussions. A collaborative approach will enhance training workflow support and promote development of a “training workflow bazaar” – a social environment for personalized access to publicly available biomedical and bioinformatics educational resources.\n\nFinally, we seek to utilize a synergistic approach across these efforts for defining the basic collaborative building blocks and using them in our respective training platforms. This will allow us to reduce duplication of efforts, to produce compatible, harmonized standards, and to prevent loss of time in reinventing code, thus alleviating the risk of having to build bridges later on and make (likely imperfect) mappings between incompatible infrastructures.\n\n\n3. International interactions\n\nBringing together experts from training consortia across four continents (North America, Europe, Australia and Africa) to discuss, share and plan for enhanced collaboration, the exchange of knowledge, and the development of further joint activities, has many advantages. These alliances may result in a set of guiding principles and shared standards that will form a solid basis for interoperability and exchange of large-scale data science training across these major international efforts. Consequently, such interactions of the TCC with these international training experts will be of direct benefit to the overall BD2K community, as well as investigators seeking a broad basis for training in large-scale biomedicine. These will be of particular importance to training and education for the BD2K community, as well as the larger biomedical data science community.\n\n\n4. Assessing the FAIRness of ELIXIR TeSS and BD2K TCC ERuDIte\n\nTo facilitate and encourage the free exchange of learning resources between training organizations, we assessed whether ELIXIR’s and the TCC’s respective international training web portals complied with the FAIR principles. For data to be classed as being ‘FAIR’, it must comply with the guiding principles of being: Findable, Accessible, Interoperable and Reusable. We specify our aligned efforts with each principle (as noted on The FAIR Data principles) below. In the sections that follow we assess TeSS, the web portal of ELIXIR, and ERuDIte, the index that powers the TCC web portal.\n\nTo be Findable:\n\nF1. (meta)data are assigned a globally unique and eternally persistent identifier.\n\nBoth TeSS and ERuDIte assign URIs to resource pages, as well as collect and store external links and references.\n\nF2. data are described with rich metadata.\n\nTeSS and ERuDIte use Schema.org schemas for core resource metadata (which are converging). As TeSS uses the EDAM ontology and ERuDIte uses the DSEO ontology to assign concept/tags to learning resources, it was necessary to begin to map the ontology terms to each other.\n\nF3. (meta)data are registered or indexed in a searchable resource.\n\nTeSS and ERuDIte provide search engines for accessing resource metadata.\n\nF4. metadata specify the data identifier.\n\nTeSS and ERuDIte link to the original learning resource URI.\n\nTo be Accessible:\n\nA1. (meta)data are retrievable by their identifier using a standardized communications protocol.\n\nTeSS and ERuDIte use HTTP/S URLs.\n\nA1.1. the protocol is open, free and universally implementable.\n\nTeSS and ERuDIte use HTTP/S protocols.\n\nA1.2. the protocol allows for an authentication and authorization procedure, where necessary.\n\nTeSS and ERuDIte use HTTP/S protocols.\n\nA2. metadata are accessible, even when the data are no longer available.\n\nTeSS and ERuDIte registries are independent of original resources.\n\nTo be Interoperable:\n\nI1. (meta)data use a formal, accessible, shared and broadly applicable language for knowledge representation.\n\nTeSS and ERuDIte use Schema.org classes and properties to represent resources. ERuDIte also uses the SKOS ontology to represent the DSEO’s concepts (see for example).\n\nI2. (meta)data use vocabularies that follow FAIR principles.\n\nSchema.org is FAIR, and DSEO and EDAM are also FAIR, both being indexed on bioportal.org, the primary repository of biomedical ontologies, and both available on GitHub.\n\nI3. (meta)data include qualified references to other (meta)data.\n\nBoth TeSS and ERuDIte strive to make data more linked (e.g., link to URIs of people, organizations) in order to contribute to create a web of linked learning resource data (see for example) (Ambite et al., 2019).\n\nTo be Reusable:\n\nR1. meta(data) have a plurality of accurate and relevant attributes.\n\nSignificant work has been completed to ensure that classes and properties appropriately describe learning resources in ERuDIte’s and TeSS’s domains. We have established an alignment between ERuDIte’s and TeSS’s schemas, and both have a strong bias to map into Schema.org\n\nR1.1. (meta)data are released with a clear and accessible data-usage licence.\n\nThe TCC provides resource metadata under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International Licence.\n\nR1.2. (meta)data are associated with their provenance.\n\nERuDIte and TeSS always cite resource providers.\n\nR1.3. (meta)data meet domain-relevant community standards.\n\nBy using Schema.org and emerging EDAM/DSEO ontologies to describe the resources, the standards defined by the linked data/semantic web community have been met.\n\n\n5. ELIXIR and TCC’s shared challenges and approaches to enriching experiences on learning web portals\n\nIn the following sections, we describe areas of development for TeSS and the TCC Portal to support online learning communities.\n\nTraining portals, such as ELIXIR TeSS and the BD2K TCC Portal, powered by ERuDIte, aim to provide access to high-quality training resources that are available in a wide variety of online sources. Identifying such resources is a challenging problem, given that they are widely distributed and continuously evolving. We are following a three-pronged approach to resource identification:\n\nA. Expert domain knowledge – Both portals are being developed to supplement existing training networks. These communities include experienced trainers who are familiar with projects, other trainers and resources in a wide variety of disciplines. We leverage this expertise to identify and evaluate new high-quality resources.\n\nB. Provider-supplied resources –As the user-bases and reputations of the portals grow within their target communities, trainers wishing to raise the exposure of their work can notify a portal that they want to be included; or, if the option is available, register their resources through a self-service interface. To encourage trainers to register their materials with the portals, the portals should ensure there are clear benefits to the trainers. One suggestion is to display usage metrics, such as the number of searches and clicks, and the amount of positive feedback a resource receives. Trainers can then use this information not only to demonstrate their impact to funders and show leadership within their research communities, but also to keep track of usage and popularity of the material.\n\nC. Open-web search – We build upon the capabilities of existing search engines, such as Google, to cast a wide net for relevant training materials, which may be subsequently filtered to ensure resource quality allowing for refined searches. For example, the BD2K TCC executes about 100 queries for specific data science concepts on YouTube to identify relevant videos. These searches produce tens of thousands of videos, which are filtered by applying machine-learning classifiers trained to select those of high quality. A complementary approach is to use Google’s structured data API to identify web resources containing specific Schema.org classes and properties indicative of learning resources. The metadata of such resources can be obtained and added to the index.2\n\nHaving rich, structured metadata annotations benefits multiple audiences that interact with resource aggregators, such as ELIXIR’s TeSS and the BD2K TCC’s ERuDIte. First, with richer metadata, learners have more information to use when evaluating whether or not a resource suits their needs. Secondly, search engines can more easily find and index resources, thereby increasing discoverability by learners, resource creation organizations, and any resource aggregator. Finally, providing structured metadata facilitates data interchange between resource indexes. However, to maximize these benefits, common metadata standards are desirable.\n\nIn this workshop, we focused on aligning the metadata standards used by TeSS and ERuDIte. TeSS (Beard et al., 2016) collects training events and materials related to the life sciences, with a special interest in bioinformatics, whilst ERuDIte focuses on materials related to data science (Ambite et al., 2017). Both portals are interested in creating a common representation of online training materials, the organizations that create and host them, and the people who create and teach them.\n\nCurrently, Bioschemas.org has defined specifications on how to represent life science training materials, people, organizations and events, based on the CreativeWork, Person, Organization and Event classes of the prevalent and popular Schema.org standard. Given ELIXIR’s membership in Bioschemas.org, in order to facilitate data exchange between TeSS, ERuDIte and other training sites and indexes, ERuDIte has aligned its learning resource descriptions to a common standard built from Schema.org. TeSS and ERuDIte have determined a core set of fields for our respective usage of the CreativeWork, Person and Organization classes, and we have documented how we use the properties of each class. These are the core properties and classes in ERuDIte, with their respective namespaces. Properties are also inherited from superclasses:\n\nschema:Thing : dcterms:identifier, schema:name, schema:alternateName, schema:description, schema:url, schema:sameAs, schema:image\n\nschema:Person (schema:Thing) : schema:givenName, schema:additionalName, schema:familyName, schema:jobTitle, schema:memberOf, erudite:teaches\n\nschema:Organization (schema:Thing) : schema:location, schema:logo, schema:member, schema:memberOf, erudite:provides\n\nschema:CreativeWork (schema:Thing) : schema:learningResourceType, schema:license, schema:datePublished, erudite:dateIndexed, schema:version, schema:author, schema:isPartOf, schema:hasPart, dcterms:subject, schema:audience, schema:aggregateRating, schema:inLanguage, schema:offers, schema:comment, schema:timeRequired, schema:provider, schema:copyrightHolder, schema:educationalUse, schema:typicalAgeRange, erudite:enables\n\nschema:Course (schema:CreativeWork) : schema:coursePrerequisites, erudite:syllabus, schema:startDate, schema:endDate, schema:instructor, lom:typicalLearningTime, xcri:contactHours, schema:grantsCredential\n\nERuDIte is aligned with Bioschemas.org and Schema.org. ERuDIte includes three additional properties not available in Schema.org.\n\nTeSS and ERuDIte provide concept tags to every material, but the two indexes use different tagging ontologies. TeSS uses EDAM (Ison, et al., 2013), which is expanded to suit the specific coverage of life science, healthcare and bioinformatics events and materials. ERuDIte uses the custom Data Science Education Ontology (DSEO), which was designed using both top-down and bottom-up approaches (Ambite et al., 2017). In a top-down fashion, the DSEO was first manually designed as an initial ontology based on expert knowledge. In the bottom-up approach, the ERuDIte team automatically computed topic models from materials’ text to identify additional concepts. Finally, the team curated and organized all these concepts. We are creating a mapping between DSEO and EDAM, which will allow us to see which materials we are interested in and then tag the materials with the terms in our respective ontologies. While this tagging ontology mapping will emerge from the TeSS and ERuDIte collaboration, it also opens up collaborations with any other institutions that use EDAM, such as GOBLET, H3ABioNet and EMBL-ABR, who also provide a rich set of training content suitable for indexing.\n\nCurrently, TeSS and ERuDIte use similar approaches to enable resource discovery. Both portals have simple, intuitive search interfaces that allow free text searching, and also provide resource tags that facilitate search filtering. Although search interfaces handle the majority of resource-discovery needs, both TeSS and ERuDIte hope to provide other meaningful ways to explore resources.\n\nFirst, both portals are actively working on initiatives that allow learners to create personalized resource collections. While these collections are formed to serve the specific learning needs of their creators, they also serve the needs of learners who are interested in similar topics and learning paths. In addition to these learner-created collections, both TeSS and the TCC portal aim to introduce expert-level collections, which will contain recommended resources that learners can use as templates for their own learning. Given the growth and rapid evolution of the learning domains of both ERuDIte and TeSS, having expert validated collections will help learners find a starting place for their learning journey through a topic or concept.\n\nThe BD2K TCC is also interested in providing interactive resource browsing interfaces. Currently, the TCC provides resource summarizations that guide learners to resources based on topics and providers of interest. In the future, the TCC aims to provide a meaningful resource-to-resource browsing interface where learners can see and select groups of similar resources. This visualization paradigm can then be shared with TeSS, if a browsing visualization is requested by its learning community.\n\nLastly, both the TCC and TeSS are interested in creating personalized recommendations for learners. Both sites are currently constrained by the size of their user bases, but as more user data are available, both portals will be able to provide resource-level suggestions that can be used on the portal directly or through community outreach methods (e.g., email campaigns or lists).\n\nUltimately, one clear way to assess the success of the TeSS and TCC web portals is the volume of resource reuse and re-indexing by other organizations. However, this relies heavily on the relevance and quality of the resources presented on TeSS’s and TCC’s respective websites. Both organizations hope that quality and relevance assessment can be accomplished by two groups: online learners and portal developers.\n\nResource quality evaluations from portal developers would need to consist of procedures that could semi-automatically check for resource status (e.g., active link or dead link), and duplication and accuracy of collected metadata. Using appropriate standards and schemas, the version history of training resources, and their author and contributor lists, can also be tracked in a transparent fashion, giving the content developers due credit and visibility for their materials. Resource quality and relevance assessments from online learners will come from user input, which could be implicit (e.g., the frequency of a resource in learners’ collections), or explicit (e.g., ratings and surveys). Both implicit and explicit feedback from learners will be essential for improving user experiences and resource reuse on both portals.\n\nOnce the reputation of the high quality of resources in the respective portals has been established, others will be encouraged to reuse and share them. Ideally, reuse will be accomplished by extracting our metadata markups, which will then spread the learning resource standards we have specified, allowing greater visibility and community use of our representations.\n\n\n6. Personalized learning\n\nThrough years of efforts, GOBLET, TeSS and ERuDIte (GTE) have built data science ontologies – hierarchical concept schemas that define knowledge networks in bioinformatics. Additionally, GTE have collected a large number of publicly accessible educational resources associated with the concepts in these ontologies. Using these resources, GTE are developing online platforms that enable people from varied professional backgrounds and levels of proficiency to efficiently and conveniently gain access to data science training resources.\n\nPersonalized training programs can recommend resources that are better aligned with learners’ interests and strengths, and may increase their motivation. Therefore, we are interested in the creation of personalized learning environments that enable users to actively customize their education using a wide variety of online resources. Throughout this process, GTE would assist users in optimizing their learning experiences by providing individualized collections of resources that properly fit their goals.\n\nThis personalization requires the creation of a systematic process to generate a learner model that identifies competency levels and learning objectives, which can then be updated with progress towards those goals. Subsequently, based on the learner model, an automatic personalization process would plan an optimal route through concepts in the ontology to identify collections of resources appropriate for users’ personal skills and educational goals. The system should also be able to adjust and update its recommendations based on user feedback, to better serve users in accomplishing data science-related training goals.\n\nThe more information a user provides to the platform, the more personalized their learning experience can be. Initially, users may be asked to fill out a learner profile survey, which should ask them to rate their current level of expertise with various concepts, their interest in learning more about those concepts or achieving broader training goals. Users who would prefer to explore resources for themselves rather than take a survey can still receive personalized experiences. They can incrementally complete a learner profile through explicit or implicit indicators: e.g., rating resources they are familiar with, or simply browsing resources related to particular topics. As learners use the platform, personalization is most obvious through resource recommendations: when looking at a resource, they may be recommended to follow it with another that continues an appropriate progression for their goals. They may also be presented with an entirely automatically generated learning plan, where the topics and the resources reflect their expertise and interests. In the following sections, we explore learner modeling and recommendations in more depth.\n\nMastery Rubric and Bloom’s Taxonomy. The Bloom Taxonomy describes six levels of cognitive skills, providing a description over a coded continuum of expertise that would be useful for resource developers. The levels are: 1) Remember/Reiterate, 2) Understand/Summarize, 3) Apply/Illustrate, 4) Analyze/Predict, 5) Evaluate/Synthesize, and (6) Create/Apply/Judge; these increase in cognitive complexity with each level (Bloom 1956). Mastery Rubric (MR) development provides a powerful methodology for developing curricula and assessments designed to teach key Knowledge, Skills and Abilities (KSAs) for specific disciplines (Tractenberg et al., 2010; Tractenberg et al., 2017). The MR formulation involves outlining how KSAs align with performance levels to provide a trajectory for students as they advance in a subject. MRs have been developed for Clinical Research (Tractenberg et al., 2017), Ethical Reasoning (Tractenberg et al., 2015), and Statistical Literacy and Evidence-Based Medicine (Tractenberg et al., 2016).\n\n6.2a High-level structure. There are many ways to model learners, but we will focus on the aspects that are important to enabling the automated personalization of learning experiences. The aspects we have identified are a) the learner, b) the concept-skill combinations comprising the domain of interest, and c) the available learning resources. We will describe these in this section.\n\n6.2b A four-dimensional learner model. Perhaps the key question when considering how we might develop user-specific capabilities for an individual within a learning portal is to think carefully about how we model that individual. We propose modeling users with a matrix that links subject domains and skills to users’ expertise and interests. This is shown diagrammatically below.\n\nThe example in the figure depicts the situation of a learner who wants to find out about the field of genetics but does not know where to start. It shows that the same learner is also an R programmer with existing expertise, but wants to learn Python to fulfill the requirements of a project. The system should recommend courses that meet these interests, whilst leveraging elements in which the learner already has expertise.\n\nThe key aspect of this underlying model is that the table shown above would need a large number of rows that codify the various different domains and skills possessed by a trained individual. Clearly, how we go about modeling each of these dimensions is crucial to the utility or effectiveness of this approach. We begin here to outline a first pass over the process of characterizing these dimensions, based on the various ontologies and conceptual schema being developed within the GTE community to date.\n\n6.2c Domains. To model people’s learning goals and their progress toward them, it is important to have a model of the domain of interest. Domain models take various forms, ranging from ontologies that domain experts and knowledge engineers construct by hand at great cost, to methods that are less reliable but fast and scalable, such as automatic topic modeling using Latent Dirichl location (LDA; Blei et al., 2003).\n\nWe do not prescribe a method for creating a domain model or a precise form, but there are several requirements for supporting personalized learning. The domain model must provide a list of discrete, named concepts. These allow the user model to represent the state of the learner’s knowledge about particular aspects of the domain, and those aspects that they have expressed interest in learning more about, either explicitly or implicitly: e.g., by browsing resources related to a particular concept. The model of the domain must also connect the identified concepts to the available resources with weighted links, giving the degree to which each resource pertains to a particular concept. Concepts can be connected to one another by various relations, including prerequisite, similarity and part-of. These relations are used to recommend concepts and their associated learning resources, described in Section 4.\n\n6.2d Skills and abilities. Following the methodology defined in the development of Mastery Rubrics3, we encode KSAs as a key element. To our knowledge, the development of taxonomies or ontologies for skills and abilities in biomedical data science is still in its infancy, but existing resources provide informal representations that support this notion. For example, Welch et al. (2016) provide a preliminary encoding of “core competencies” that broadly outline various general skill-sets required for bioinformatics.\n\nGoing beyond these general categories into more detail may require a finer-grained representation of specific skills and abilities. Potential starting points are provided by the EDAM ontology4. This resource provides a taxonomy of operations that could act as a proxy in the specific area of bioinformatics-based data analysis. Similarly, the Data Science Education Ontology (DSEO)5 provides a taxonomy of data science processes that could be elaborated into a more complete description of skills and abilities that we could use here.\n\nBeyond these resources, a starting set of high-level skill categories could include: simple comprehension, concept mastery, basic procedural execution, procedural troubleshooting, and procedural implementation. It may then be necessary to further specialize these abilities within domains, programming languages, data science processes, etc.\n\n6.2e Knowledge/expertise. Here, we follow the encoding provided by the Bloom Taxonomy/Mastery Rubric methodology (Bloom, 1956; Tractenberg et al., 2010), which specifies expertise as an ordinal scale with the following categories (in ascending order of expertise): a) beginning, b) novice, c) competent, and d) proficient. Subsequent incarnations of the Mastery Rubric specify a trajectory from a) Novice, to b) Beginner, to c) Apprentice, to d) Journeyman.\n\n6.2f Interest. Learners may have only a general idea of what they are interested in (e.g., “data science” or “bioinformatics”), or they may have specific interests (e.g., “using R to analyze genomic data”). Part of a functional learner model is to record which concepts – at different levels of granularity – a learner would like to learn more about. Wherever possible, this interest should be assigned to the particular skill: e.g., conceptual vs. procedural. Our model of learners’ interests is a mixture of what they tell us (e.g., by defining their goal), and what we infer (e.g., from identifying themes in learning resources they access).\n\n6.2g Leveraging existing resources and implementation. ERuDIte (Ambite et al., 2017; Van Horn et al., 2018) consists of an index of both manually and automatically identified data-science resources – predominantly online courses and videos – which are tagged with relevant concepts from DSEO. Ongoing work seeks to identify natural sequences of resources based on prerequisites, building on the architecture to do so from the TechKnAcq project (Gordon et al., 2016).\n\nThe EDAM ontology (Ison et al., 2013) provides an extensive representation of four branches: a) Data (with Data Identifiers) denoting types of information that may be processed computationally; b) Operations that may be applied to those data; c) the many Formats in which the data may be expressed; and d) the Topic to which the data are relevant. This provides an excellent, exhaustive collection of element classes that may be combined to describe data-science processes and structures.\n\nCurrently in development on the TeSS portal are “Training Workflows”6. These are visual tools to help users navigate sets of diverse training resources. Broadly, three distinct types of training workflow are encapsulated in TeSS:\n\n• abstract snapshots of typical data-analysis pipelines;\n\n• hands-on, step-by-step guides through specific data-analysis tasks; and\n\n• sign-posted routes charting developmental trajectories through training resources.\n\nTraining Workflows are constructed manually by domain experts and training professionals. A graphical WYSIWYG editor is provided to help trainers design their workflows, using standard components and headers. Within the editor, users may specify which training resources trainees should review in order to gain the requisite understanding of a topic before moving on; they may also select tools from the bio.tools7 registry, and/or databases, standards and policies from the FAIRSharing8 registry.\n\n\n7. Possible applications and use cases\n\nWhen learners first use this learning portal, it is proposed that they would be required to complete a “user profile”, where they can state their prior experience/expertise, career level, their interests and, if known, explain what their end goal, career-wise, will be. It is envisaged that this information could then be used in one of two ways:\n\n1. The learning portal will have predefined “learning paths” to allow learners to progress from one level of expertise to another (e.g., from bioinformatics user to bioinformatics scientist9). Using this, the system could therefore recommend a specific set of courses for the learner to take.\n\nor\n\n2. Once a learner has attended certain courses, others can be recommended to them based on previous learners attending the same series of courses.\n\nThe theory behind the second case is that learners, sometimes, do not know what courses would be complementary or useful for their work, and do not always know what other courses are available. The courses would be recommended informally, in a similar manner to shopping for items on Amazon, using phrases such as “other learners who attended this course, also attended this course…”. Just as with Amazon ‘recommendations’, users would be free to choose which of the offerings resonated most with their personal training needs. These recommendations are not intended to be required for the accomplishment of a learning goal—they are a gentle guide for people who may be new to the topic or resource or for people who are uncertain about what resources they should view next. It is important to note that interest does not equate to expertise, and before making a course recommendation, the system would check that any prerequisite courses have been completed by the learner. Additionally, by filling out a comprehensive user profile when first using the platform, it is hoped that the recommendations would be accurate, and would ultimately allow more learning paths to be created.\n\nTo create personalized learning recommendations, we must also capture the relations between concepts in the domain. For example, before recommending a resource about the concept Machine Learning, we would like to ensure that the learner has familiarity with the concepts of Probability and Algorithms, as these concepts are joined by prerequisite relations. If a learner indicates interest in Machine Learning, we should suggest resources about Supervised Learning, Unsupervised Learning, EM Algorithm, etc., as these are all connected to Machine Learning by hierarchical part-of relations.\n\n\n8. Innovating across international efforts\n\nThere is no “one size fits all” when it comes to learning, and a useful learning path for one scientist may not be useful for another. However, the creation of personalized learning paths and personalized recommendations could more readily guide learners than if they had to navigate the sea of available courses alone. So many resources and courses are offered across multiple sites that it can be daunting and overwhelming for learners to know what they should look for.\n\nThe huge volumes of data now being routinely generated from high-throughput instrumentation (imaging data, data from sequencing platforms, etc.) have generated an insatiable demand for bioinformatics and data science training among bioscientists. There are many online training resources, including videoed lectures from conference talks, tutorials, short courses and degree programs, and materials on GitHub and other online repositories. However, while much effort has been placed in creating these materials, it can still be very difficult both for students, who wish to improve their skills, and for trainers, who wish to share and re-use other trainers’ materials, to find what they need. By following the FAIR Data Principles, international initiatives like the ERuDIte and TeSS portals seek to make it much easier to find and re-use training materials by identifying, indexing and collating them into central resources. Complementing these largely automated approaches are dedicated, manually maintained training portals, such as the one provided by the GOBLET Foundation. By working together, international training efforts can both identify commonalities and complementarities, and pinpoint synergistic opportunities to drive the field forward.\n\nTo make training materials FAIR, they must be adequately annotated and described so that they can be indexed appropriately. Currently, nascent annotation standards are not yet widely adopted, and there are no widely accepted quality metrics to indicate whether courses are delivering the requisite core competencies to trainees (Welch et al., 2016). Furthermore, to make a course truly reusable, all the information required to re-use the associated materials (including data-sets and tools used for practical exercises, and the exercises themselves) needs to be made available both to students and to fellow trainers. Jupyter notebooks, for example, containing code snippets and notes describing the code, are readily reusable, and there are many Jupyter notebooks on GitHub. But how can these be described so that they can be indexed and readily found? International collaboration could help to define agreed guidelines and standards for course and material description to alleviate this problem.\n\nAlongside the challenge of making training materials FAIR is the growing demand for training, which is outstripping our ability to readily satisfy the changing skill requirements of bioscientists. To meet this demand, more instructors are needed, and new trainers must be adequately prepared and trained to train others. The Software Carpentry (SWC) Foundation has developed a model to ensure that trainers who deliver Software Carpentry- or Data Carpentry-branded workshops are trained to do so by undertaking an online or on-site (for SWC members) course, where they learn about pedagogy. Similarly, ELIXIR and GOBLET have developed a Train-the-Trainer (TtT) program (Via et al., 2017), which was piloted across Europe in 2016, to help build trainer capacity. ELIXIR has also extended these TtT course to Australia via BPA-CSIRO and EBI collaborations set up in 2012. Since then, more than 20 trainers have been trained to deliver bioinformatics training in Australia (Watson-Haigh et al., 2013). Against this background, the international data science and bioinformatics communities could together develop common guidelines for trainer training that would help to raise the quality of training delivery and increase the pool of available trainers.\n\nHaving participated in a course, a common theme among students is the desire for some form of certification. Without any aspect of assessment in the course, this will most commonly be a certificate of participation. However, many students are interested in a certificate of accomplishment; this is particularly common with MOOCs, such as those delivered by Coursera, EdX or Udacity. Here, on completion of end-of-course assessments to a set standard – and payment of a fee – students receive a certificate of accomplishment; those who opt to take the course without paying the fee are not certificated. There are practical limitations to the length of a course that determine whether and how it is meaningfully assessable. Nevertheless, for short courses, no assessment guidelines are yet available for universal adoption. Combining our joint experience and expertise, an opportunity therefore exists for the international community to make a set of recommendations for trainers who are developing course assessments, in order for some sort of universal certification to be created.\n\nWhen reusing online materials, it is important to both cite the sources and give credit to the authors. Materials can be released under several different Creative Commons licences, each of which has different features. Understanding the differences between each licence, and when a particular one should be used, would be beneficial to course developers. Across the international bioinformatics and data science communities, offering guidance to trainers on what licence to use, and allowing end-users to understand their responsibilities when using licensed materials, would be very valuable, and may encourage more trainers to make their resources available online.\n\nThere are different types of training courses, including those that i) focus on skills development, ii) tackle data-analysis problems, and iii) develop and encourage creative thinking. Skills-development workshops are the most common, where students follow lectures and tutorial exercises, and perform tasks for which tools and data have already been prescribed: e.g., how to analyze RNASeq data. In data-analysis workshops, which are outcome-driven (and sometimes also referred to as ‘hackathons’ or ‘data jamborees’), rather than follow pre-set, generic examples, participants actually create something meaningful to them using freely available data. H3ABioNet has used the hackathon concept for different scenarios, including the development of Docker containers for selected data-analysis workflows and for collaborative analysis of specific data-sets, bringing together scientists from diverse backgrounds. Extending this idea, ‘creative thinking’ workshops bring together people with different expertise (e.g., bioinformaticians and mathematicians, or bioscientists and data scientists), but without a fixed agenda. Here, the idea is that by discussing issues that one group may have with their data, the other group may propose solutions. Ultimately, the outcome may be, for example, a research proposal for a grant application, leveraging the interests and expertise of both groups.\n\nData science in biology and medicine can learn from training approaches in other disciplines: e.g., astronomy and its enormous data-sets bring familiar challenges in data analysis and stewardship, and the same basic principles and underlying skill-sets apply. Bioscientists, however, generally have little exposure to the computing and data science skills now necessary for their work, and are likely to seek training much later in the data lifecycle (Attwood et al., 2017). IT professionals play a fundamental role in supporting the life sciences, as their skills are required to install bioinformatics tools on HPC platforms, and to troubleshoot the many problems that are likely to arise. However, rapid advances in technology mean that the current best-practice software for a given task changes rapidly, making it difficult to keep pace. Understanding what bioscientists do with these tools, how quickly they change, why they may need so many tools to analyze a given data-set, the demands of storing and analyzing their data, and so on, would be hugely informative to the IT professionals who support their work.\n\nInternships offer another route via which bioscientists may augment their data science skills. The BD2K TCC supports – most commonly for junior faculty staff – the opportunity to partner with senior data scientists (for 2–3 weeks, or up to multiple months) in order to learn how best to analyze their data. At the end of these “road trips”, the bioscientists return to their own institutions and deliver training courses on what they have learned during their internship. H3ABioNet offers internships to acquire skills form experts at the host institution while working on their own data. The EU offers COST Action grants10, which exploit a similar training paradigm – so-called “short term scientific missions”. Finding out about such professional development opportunities can be very difficult, so providing a resource that indexes them would be very useful; this could also expand the program to become a globalized “road trip”, facilitating exchange of expertise internationally if sufficient funding were available. A combined international index could also offer a “matchmaking” service, connecting those who wish to gain a particular skill-set with those who are willing to host them and help them do so (rather like the Knowledge Transfer Program11 advocated by the Centre for Proteomic and Genomic Research in Cape Town). Typically, such programs yield research collaborations, funding proposals and publications that last long after the road trips have ended. Consequently, junior researchers are not just supported in developing their skills, but may also benefit from new international collaborations that support them along their career paths.\n\nIncreasing the visibility of data science training will facilitate an increase in diversity and enable under-represented groups to better exploit these opportunities. Indeed, some initiatives are already underway to represent and encourage more participation by women in the data sciences. WiDS (Women in Data Science)12 is a global initiative to inspire and educate data scientists worldwide that specifically supports women. Their 2017 conference was hosted at Stanford University, and was webcast live or had a delayed broadcast to 80 locations, in 30 countries, worldwide. Similarly, RLadies13 is an organisation that promotes gender diversity in the R community, and has multiple city-based chapters across the world.\n\nAside from connecting international training communities and efforts, a clear value of making training material more visible and discoverable is in sharing experience and expertise, reducing redundancy, and combining efforts to make individual contributions greater than the sum of their parts. Frequent meetings among the groups that represent these communities offer further opportunities for communication that will help each group to understand the other’s work, and to identify skill or knowledge gaps. However, these activities cannot continue without funding. Training bioscientists in data science is an essential investment for 21st-century bioscience. Continued investment in data science training may be possible through large funding initiatives, such as NIH’s, or through large-scale multinational projects like ELIXIR. However, data science skills are also sought after by industry, with companies like Google, Microsoft and Amazon becoming active in biomedical and bioscience research. It may be possible to leverage this interest and partner with them to increase the pool of bioscientists skilled in data science. Furthermore, charitable organizations like the Bill and Melinda Gates or Gordon and Betty Moore Foundations, which fund grand challenges in bioscience, may also be interested in partnering to address the global skills shortage in data science.\n\n\n9. Conclusions\n\nModern bioscience increasingly depends upon data science, advanced biostatistics and informatics approaches in order to manage, model and understand the ever-growing amounts of data being accumulated. This challenge has drawn the attention of major science consortia in Europe, Africa, Australia and the United States. Importantly, this necessitates consideration of how best to provide training in the modern science of “data”, which is not necessarily part of the undergraduate or even, perhaps, graduate-level bioscience curricula.\n\nAs discussed above, a multitude of factors must be considered in ensuring efficacious training and educational programs that focus on the role of training for biomedical data science, and instilling of core competencies in the development of career paths. Value exists in standardizing and sharing data science training resources across national boundaries to ensure commonalities across communities. However, challenges will always remain in encouraging adoption of training material standards, which indicates that such training should be compelling, engaging and comprehensive. In so doing, the availability of the Internet affords that resources and learning experiences can be personalized and customized to the individual. This works to accommodate the widest set of possible stakeholders in the modern bioscience enterprise, providing skills that fill important niches in learners’ understanding. International partnerships can help to inform the world on data science training in ways that are seen as positive for all such stakeholders. We look forward to continued efforts between the BD2K’s TCC, ELIXIR’s TeSS, bioCADDIE, GOBLET, EMBL-ABR and H3ABioNet to promote biomedical data science through the recommendations made herein.\n\n\nData availability\n\nNo data is associated with this article.",
"appendix": "Grant information\n\nThis work was supported in part by National Institutes of Health (NIH) grant U24 ES026465 02, part of the NIH Big Data to Knowledge (BD2K) initiative.\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 thank our respective colleagues for providing critical comments on an earlier draft of this manuscript. We also wish to acknowledge the encouragement and support of Dr. Carol Shreffler of the National Institute for Environmental Health Sciences (NIEHS).\n\n\nFootnotes\n\n1Bioinformatics is used here in its broadest sense to encompass aspects of computational biology, biocomputing, biocuration and data science\n\n2A number of existing tools facilitate parsing Schema.org metadata from web pages: https://developers.google.com/custom-search/docs/tutorial/creatingcse\n\n3https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4664514/pdf/nihms-671775.pdf\n\n4http://bioportal.bioontology.org/ontologies/EDAM?p=classes\n\n5http://bioportal.bioontology.org/ontologies/DSEO\n\n6https://tess.elixir-europe.org/workflows\n\n7https://bio.tools\n\n8https://fairsharing.org\n\n9http://journals.plos.org/ploscompbiol/article/file?id=10.1371/journal.pcbi.1004943&type=printable\n\n10http://www.cost.eu/COST_Actions\n\n11https://www.africancentreforcities.net/programme/mistra-urban-futures/the-knowledge-transfer-programme/\n\n12http://www.widsconference.org/about1.html\n\n13https://rladies.org/\n\n\nReferences\n\nAmbite JL, Fierro L, Geigl F, et al.: BD2K ERuDIte: the Educational Resource Discovery Index for Data Science. The Fourth WWW Workshop on Big Scholarly Data: Towards the Web of Scholars (BigScholar). 2017. Reference Source\n\nAmbite JL, Gordon J, Fierro L, et al.: Big Data to Knowledge (BD2K) Training Coordinating Center (TCC) Educational Resource Discovery Index (ERuDIte) as Linked Data. Zenodo. 2019. http://www.doi.org/10.5281/zenodo.2553415\n\nAttwood TK, Blackford S, Brazas MD, et al.: A global perspective on evolving bioinformatics and data science training needs. Brief Bioinform. 2017; bbx100. PubMed Abstract | Publisher Full Text\n\nAttwood TK, Bongcam-Rudloff E, Brazas ME, et al.: Correction: GOBLET: The Global Organisation for Bioinformatics Learning, Education and Training. PLoS Comput Biol. 2015; 11(5): e1004281. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBeard N, Attwood T, Nenadic A: TeSS - Training Portal [version 1; not peer reviewed]. F1000Res. 2016; 5(ISCB Comm J):1762 (poster). Publisher Full Text\n\nBloom BS, Engelhart MD, Furst EJ, et al.: Taxonomy of educational objectives: The classification of educational goals. Handbook I: Cognitive domain. (New York: David McKay Company).1956. Reference Source\n\nBlei DM, Ng AY, Jordan MI: Latent Dirichlet Allocation. J Mach Learn Res. 2003; 3(4–5): 993–1022. Reference Source\n\nBrazas MD, Blackford S, Attwood TK: Training: Plug gap in essential bioinformatics skills. Nature. 2017; 544(7649): 161. PubMed Abstract | Publisher Full Text\n\nBui AAT, Van Horn JD, NIH BD2K Centers Consortium: Envisioning the future of 'big data' biomedicine. J Biomed Inform. 2017; 69: 115–117. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCorpas M, Jimenez RC, Bongcam-Rudloff E, et al.: The GOBLET training portal: a global repository of bioinformatics training materials, courses and trainers. Bioinformatics. 2015; 31(1): 140–2. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGordon J, Zhu L, Galstyan A, et al.: Modeling Concept Dependencies in a Scientific Corpus. Proceedings of the Annual Meeting of the Association for Computational Linguistics. (ACL.). 2016. Reference Source\n\nIson J, Kalas M, Jonassen I, et al.: EDAM: an ontology of bioinformatics operations, types of data and identifiers, topics and formats. Bioinformatics. 2013; 29(10): 1325–1332. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLarcombe L, Hendricusdottir R, Attwood TK, et al.: ELIXIR-UK role in bioinformatics training at the national level and across ELIXIR [version 1; referees: 4 approved, 1 approved with reservations]. F1000Res. 2017; 6: pii: ELIXIR-952. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorgan SL, Palagi PM, Fernandes PL, et al.: The ELIXIR-EXCELERATE Train-the-Trainer pilot programme: empower researchers to deliver high-quality training [version 1; referees: 2 approved]. F1000Res. 2017; 6: pii: ELIXIR-1557. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchneider MV, Griffin PC, Tyagi S, et al.: Establishing a distributed national research infrastructure providing bioinformatics support to life science researchers in Australia.. Brief Bioinform. 2017; bbx071. PubMed Abstract | Publisher Full Text\n\nTractenberg RE, Gordon M: Supporting Evidence-Informed Teaching in Biomedical and Health Professions Education Through Knowledge Translation: An Interdisciplinary Literature Review. Teach Learn Med. 2017; 29(3): 268–279. PubMed Abstract | Publisher Full Text\n\nTractenberg RE, Gushta MM, Weinfeld JM: The Mastery Rubric for Evidence-Based Medicine: Institutional Validation via Multidimensional Scaling. Teach Learn Med. 2016; 28(2): 152–165. PubMed Abstract | Publisher Full Text\n\nTractenberg RE, Russell AJ, Morgan GJ, et al.: Using Ethical Reasoning to Amplify the Reach and Resonance of Professional Codes of Conduct in Training Big Data Scientists. Sci Eng Ethics. 2015; 21(6): 1485–1507. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTractenberg RE, Umans JG, McCarter RJ: A Mastery Rubric: guiding curriculum design, admissions and development of course objectives. Assess Eval High Educ. 2010; 35(1): 15–32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWelch L, Brooksbank C, Schwartz R, et al.: Applying, Evaluating and Refining Bioinformatics Core Competencies (An Update from the Curriculum Task Force of ISCB's Education Committee). PLoS Comput Biol. 2016; 12(5): e1004943. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVan Horn JD: Opinion: Big data biomedicine offers big higher education opportunities. Proc Natl Acad Sci U S A. 2016; 113(23): 6322–6324. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVan Horn JD, Fierro L, Kamdar J, et al.: Democratizing data science through data science training. Pac Symp Biocomput. 2018; 23: 292–303. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVia A, Blicher T, Bongcam-Rudloff E, et al.: Best practices in bioinformatics training for life scientists. Brief Bioinform. 2013; 14(5): 528–537. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVia A, Attwood TK, Fernandes PL, et al.: A new pan-European Train-the-Trainer programme for bioinformatics: pilot results on feasibility, utility and sustainability of learning. Brief Bioinform. 2017; bbx112. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "45374",
"date": "28 Mar 2019",
"name": "Ariel Rokem",
"expertise": [
"Reviewer Expertise Neuroinformatics",
"data science",
"cognitive 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\nThis article summarizes the discussions from a meeting of several organizations from different parts of the world that all develop and collect materials for education in biomedical data science.\nOne of the main strengths of the article is that it includes a world-wide perspective on training in the intersection of computing and the life sciences, with representatives from four different continents. As expected from such a forum, the paper covers a lot of ground, ranging from theories of learning, through technical development of databases for learning resources and through topics such as interactions with stakeholders. This is a major strength of the article, but it also makes the reading of the article slightly daunting, and I worry that it would make it hard for a potential reader to navigate. I propose adding a roadmap in the beginning of the article, which would direct readers towards their topic of interest. I also think that the article would benefit from relegating some technical details, such as the detailed FAIRness assessment in section 4, or the details of the schema in section 5.2 to appendices. The former (section 4) could possibly also be summarized in a table. Similarly, some theoretical background could possibly be elided, and references provided for readers who want to (e.g., Bloom's taxonomy in section 6.1).\nSimilarly, at the very end of the article, I think that it would help to summarize the main conclusions and recommendations made in the article and reiterate them.\nOne topic that I would have liked to see addressed more is the ability that relatively decentralized systems have to provide objective assessments of users' skill levels, progress and outcomes. The work of the Johns Hopkins Data Science Group on this topic is particularly compelling. For example, articles that they published about their MOOC series: Kross et al.,20171 and Hadavand et al., 20182, and work that they have done on systems for automating production of educational resources, and primarily the Swirl project (https://swirlstats.com/), which provides a uniform interface for creating data science courses that run in the RStudio console.\nOne additional topic that could be further addressed is the motivations that lead researchers to contribute training materials to the repositories that are mentioned in the article. This is addressed in 5.1.C. Another option that is not mentioned is providing venues for publication, and mechanisms to cite and provide credit and feedback, to the creators of OER. One possible venue for this is the recently-started Journal for Open Source Education, https://jose.theoj.org/ (with the full disclosure that I am a member of the editorial board of JOSE's partner journal for open source software (JOSS) and a contributor to JOSE).\nI found the assessment of interoperability between projects developed by different organizations in sections 4 and 5 to be compelling, and I was wondering why bioCADDIE's DDI was not included in these assessments.\nMinor:\nIn the abstract, the sentence: \"Discussions focused on the role of i) training for biomedical data science; ii) the need to promote core competencies, and the ii) development of career paths\" should be \"Discussions focused on i) the role of training for biomedical data science; ii) the need to promote core competencies, and iii) the development of career paths\"\nIntroduction, I believe that the sentence: \"In May 2017, during a two-day meeting between the BD2K Training Coordinating Center (TCC), ELIXIR Training/TeSS, GOBLET, H3ABioNet, EMBL-ABR, bioCADDIE and the CSIRO, convened in Huntington Beach, California, to describe our respective activities and how these might be leveraged for wider impact on a global scale\" is missing something (maybe a verb?)\nOn page 9: I believe that the sentence \"The example in the figure depicts the situation of a learner who wants to find out about the field of genetics but does not know where to start.\" refers to Figure 2. It would be good to spell that out. Also, in the following paragraph: \"The key aspect of this underlying model is that the table shown above would need a large number of rows that codify the various ...\". Because of the layout of the article, the table in the Figure is actually several rows *below* this paragraph, so it would be good to be explicit here as well.\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": "45371",
"date": "03 Apr 2019",
"name": "David N. Kennedy",
"expertise": [
"Reviewer Expertise Neuroimaging",
"Nuroinformatics",
"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\nOverall, this is an informative manuscript that sets a large number of training issues in big data in the context of the many national and international big data projects. The organization of the manuscript, however, is rather uneven, and the presentation logic is at times, confusing.\nSpecific Points:\nNeither the outline of discussion points in the abstract, nor the list of focus questions (p.3) are cleanly aligned with organizational structure of the rest of the paper. It is difficult to reference either one and then try to find the related content in the conference proceedings. Abstract does not set up problem clearly - why do the challenges to conventional methods for sharing, managing and analyzing constitute a training issue? What are the impediments on translational impact: very confusing as worded. The outline of conversations in the abstract appears to be different here from the outline of questions in the Introduction, and yet again different from the actual organization of the paper content. Even though much of the discussion ultimately considers the topics in each of the various outlines, it does make things hard to cross-reference. The structure/organization would be much more accessible if outlined by the headers, and also with the addition of headers to block some of the very long sections into meaningful subunits. Modify headers to conceptually align with the body of the text and to allow the reader to target content areas. Section 2 (Training in biomedical data science) needs a topic sentence or some overview introductory statement to orient the reader to what is coming (descriptions of 7 different big data initiatives, from different regions of the world), and why. The transition from initiative summaries in section 2 to what collectively seems to constitute a section on inter-initiative collaboration is a very awkward transition in the section as currently structured and seems like a non-sequitur. Section 3, which is a single paragraph titled ‘International interactions’ seems to be closely related to the preceding few paragraphs on inter-initiative collaborations. Re: Section 4, it is very unclear to the reader why the discussion shifts from an overview of many initiatives to a) a specific and detailed comparison of online portals from two of those initiatives, and b) the conceptual basis of a FAIRness comparison. Some topic sentence here would be important. What is the primary conclusion from the FAIR principles comparison for TeSS x ERuDIte? Identification of and utility for stakeholders are mentioned as priorities in both the Abstract and the Introduction, if this is indeed an intended priority for the readership audience, it would be helpful to summarize conclusions in a section with a corresponding header Very confusing use of the term ‘translational impact’ in the training context. Translational impact implies ‘impact of translational research’ – it isn’t obvious how this relates to training - there is a missing link between big data and the need for people to be trained to work with it in order to enable translation between current cutting edge biomedical data science and research outcomes. Section 8, pages 10-13, would be more accessible with some from some conceptual blocking/sub-headers e.g Need for Standards, Training and Trainers, Certification/Credentialing of Trainers/Trainees, Licensing of Materials, Courses and Training Paradigms, Data Science- interface between bioscientists and IT.\nOverall, the content is valuable, but the organization needs some work to improve the clarity of the presentation.\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": []
}
] | 1
|
https://f1000research.com/articles/8-251
|
https://f1000research.com/articles/8-244/v1
|
01 Mar 19
|
{
"type": "Research Article",
"title": "The effect of silver nanoparticles incorporation in the self-etch adhesive system on its antibacterial activity and degree of conversion: an in-vitro study",
"authors": [
"Heba F. Mohammed",
"Mona I. Riad",
"Mona I. Riad"
],
"abstract": "Introduction: Despite of the recent advances in the adhesive dentistry, high possibility of microbial biofilm development at the resin restoration surfaces may lead to marginal gaps and recurrent caries. Degree of conversion of the dental adhesive represents a relative assessment to its quality, and a direct correlation with its mechanical behavior. This in vitro study was carried out to investigate the minimum inhibitory concentration of antimicrobial silver nanoparticles incorporated in two forms into the self-etch adhesive system and the effect of their incorporation on the degree of conversion of the self-etch adhesive. Methods: Minimum Inhibitory Concentration of the self-etch adhesive system incorporated with nanosilver powder and solution against Streptococcus mutans was tested using an agar diffusion test. The effect of nanosilver incorporation (powder and ethanol-based solution) in the self-etch adhesive system on its degree of conversion was assessed using Attenuated Total Reflectance/ Fourier Transform Infra-Red spectrometer (ATR/ FTIR). Results: The results showed that silver nanoparticles incorporation (powder or ethanol based at 12.5 µg/ml concentration) significantly increased the antibacterial efficacy of the self-etch adhesive against Streptococcus mutans (P< 0.05). Nanosilver powder possessed higher significant antibacterial effect when compared to silver ethanol based solution (P< 0.05). Degree of conversion of self-etch adhesive containing nanosilver powder showed non-significant difference from the control group (p > 0.05). In contrast, self-etch adhesive with nanosilver solution recorded significantly lower values when compared to the control or nanosilver powder group (P< 0.05). Conclusion: The antibacterial efficacy of the adhesive system can be greatly potentiated with the addition of silver nanoparticles (12µg/mL concentration) especially the nanosilver powder. Incorporation of the antibacterial nanosilver powder in the adhesive system didn’t compromise the degree of conversion of the adhesive resin.",
"keywords": [
"Silver nanoparticles",
"Antibacterial",
"Self-etch adhesive",
"Streptococcus mutans",
"Minimum inhibitory concentration",
"Degree of conversion."
],
"content": "Introduction\n\nDental resin composite is a widely used restorative material that has superior aesthetic properties and strong bonding ability to the tooth structure in comparison to other restorative materials like amalgam1. In spite of the recent advances in the dental adhesives, there is a high possibility of microbial biofilm development at the resin restoration surface, which may lead to marginal gap and recurrent caries2.\n\nSilver is an antimicrobial agent that has broad spectrum activity against gram positive and gram negative bacteria2. Nanoparticles are insoluble particles smaller than 100 nm. Their unique size provides higher surface area, thus much more potent antimicrobial activity in comparison to the usual particles size3. Silver nanoparticles can cause cell membrane disruption and damage of the bacterial DNA1. Degree of conversion of the dental adhesive represents a relative assessment to its quality and directly correlates with its mechanical behavior. Proper polymerization of the dental adhesive can increase the longevity of the bonded restoration4. Therefore, it seems valuable to investigate the minimum inhibitory concentration of silver nanoparticles incorporated in two forms into the self-etch adhesive system and their effect on the degree of conversion.\n\n\nMethods\n\nThe materials, preparations, manufacturers, composition and batch numbers are listed in Table 1.\n\n1ml of ethanolic solution of silver nanoparticles was added to 5ml of self-etch adhesive system and serially diluted. To produce the adhesive system with the nano-silver powder incorporated, it was accurately weighed first using a scale (AE Adam, Bradford, UK). The previously measured powder (1000 µg) was added to 1 ml of adhesive solution, sonicated in an ultrasonic mixer (Eumax model: UD100SH-3LQ, China) and serially diluted using a micropipette (The final concentration of nanosilver in the adhesive resin in the two forms is 100µg/ml). The procedure was completely performed in a dark environment using X-ray processing box.\n\nThe minimum inhibitory concentration (MIC) defined as the lowest concentration of nanosilver in micrograms per milliliter (μg/ml) that inhibits the growth of an organism5. Tryptic soy agar medium; culture nutrient media for Streptococcus mutans (ATCC 25175) was poured in 10 petridish plates in a laminar flow (Telstar BIO- II-A, VWR Company, UK.). Streptococcus mutants strains; ATCC 25175 (Cairo MIRCEN, Faculty of agriculture, Ain Shams University, Egypt.) were cultured on Tryptic soy agar medium at 37°C for 24 hours6.\n\n10 petridish plates were punched by a cork-borer with a 6 mm diameter to produce rounded holes in each plate and the bacterial strain was applied equally on the agar plates7,8. The specimens were grouped as N1: Self-etch adhesive system + nanosilver solution and N2: Self-etch adhesive system+ nanosilver powder. Each agar plate contained 5 holes representing the five different concentrations of the nanosilver. For N1 and N2 groups; the self-etch adhesive system was serially diluted to produce 5 subgroups with different concentrations, C1: 100 µg/ml, C2: 50 µg/ml, C3: 25 µg/ml, C4: 12.5 µg/ml and C5: 6.25 µg/ml. The adhesive system without nanosilver incorporation was used as a control group (N0).\n\n200 µl of the adhesive agent was injected in each hole by micropipette and polymerized for 20 seconds using light emitting diode unit (Woodpecker Medical Instrument Company, Model LED. F; Model No. L14A0116F, China.) (Figure 1). The plates were incubated for 48 hours in the incubator at 37°C under completely anaerobic conditions (Shell lab Company, SMI6, Canada). The diameter of bacterial inhibition zone halo of each adhesive was measured in millimeters using a ruler7.\n\nThe specimens were prepared using a cylindrical Teflon mold surrounded by metallic ring (3mm diameter and 2mm height)9 (Figure 2). The mold was placed on a Mylar strip (Universal strips of acetate foil, Germany) that was placed on a clean flat glass slab. 200 µl of self-etch adhesive resin was injected in the hole of the mold using a micropipette then covered with the Mylar strip (to avoid the presence of oxygen inhibiting layer and pressed to obtain a uniform smooth specimen surface)4,9. The adhesive resin was cured in the presence of the top Mylar strip by LED device for 20 seconds according to manufacturer instructions. The light curing tip was applied perpendicular and with intimate contact with the top surface of the Mylar strip (Zero distance)4.\n\n15 disc shaped specimens were prepared of self-etch adhesive system. The specimens were divided into 3 equal groups (n=5) according to the form of incorporated nanosilver (N0: Adhesive system without nanosilver, N1: Adhesive system+ nanosilver solution, N2: Adhesive system+ nanosilver powder). The concentration of the incorporated nanosilver was set according to the minimum inhibitory concentration (MIC) that was tested before.\n\nDegree of conversion was measured using an Attenuated Total Reflectance/ Fourier Transform Infra-Red spectrometer (ATR/ FTIR) (Vertex 70, Bruker Company, Germany)4,10. All the data were recorded and plotted on a special computer software (OPUS Bruker Spectroscopy Software, version 7, Germany) to draw the linear graphs from which the degree of conversion of each specimen was calculated.\n\nData statistically was described in terms of mean values and standard deviation (SD) using ANOVA test (IBM® SPSS® (SPSS Inc., IBM Corporation, NY, USA, Statistics Version 22 for Windows).\n\n\nResults\n\nMean ± SD measures of the diameter of inhibition zone (DIZ) were summarized in Table 2 and Table 3 and graphically drawn in Figure 3 (underlying data available from OSF11). The largest zone of inhibition was at 100 µg/ml concentration of nanosilver powder while the smallest was at 6.25µg/ml concentration of nanosilver solution.\n\nNS; non-significant (p>0.05), *; significant (p<0.05).\n\nNS; non-significant (p>0.05), *; significant (p<0.05).\n\nAll the concentrations recorded significantly higher DIZ when compared to the control group (P≤0.05)except C5 ; 6.25µg/ml. The MIC of adhesive containing nanosilver (powder and solution form) was determined at 12.5 µgm/ml concentration. All the concentrations of nanosilver powder recorded significantly higher mean values of DIZ when compared to the different concentrations of nanosilver solution.\n\n(Mean ± SD) measures of degree of conversion in % were summarized in Table 4 and graphically presented in Figure 4. Groups of adhesive system with incorporated nanosilver solution (26.14±4.47 %) recorded a statistically significant lower degree of conversion when compared to control (50.31±4.04 %) and nanosilver powder (47.72±4.47%) (P ≤ 0.001). There was no significant difference in the degree of conversion between the control group and the group of the adhesive system containing nanosilver powder (p≥0.05).\n\nMeans with the same letter within each row are not significantly different at p=0.05. NS= Non-significant, *=Significant\n\n\nDiscussion\n\nSelf-etch adhesives exhibit limited antibacterial activity against Streptococcus mutans due to the presence of a low molecular weight monomer that possesses bacteriostatic action against Streptococcus mutans12.\n\nThe MIC for adhesive containing nanosilver in ethanol solution and powder was 12.5μg/ml concentration which was significantly different when compared to the control group (P≤0.05). These results confirmed the potential antibacterial effect of low concentrations of nanosilver. Adhesive resin with nanosilver powder at different concentrations showed significantly higher inhibition rates than that with nanosilver/ ethanol dispersion.\n\nThe higher significant efficiency of nanosilver powder may be attributed to the presence of the ethanol as a dispersion medium for nanosilver solution that can act as a diluting agent of the adhesive system. That was observed at concentration 6.25 mg/ml in nanosilver solution as it showed lower significant value than the control group despite the presence of silver nanoparticles.\n\nThe nanosilver powder group recorded a statistically non-significant difference in the degree of conversion when compared to the control group. Presence of nanofillers in the adhesive system had no harmful effect on the degree of conversion13.\n\nBesides reducing of the amount of residual monomer, the nano-particles size is less than the wavelength of the blue light of the curing units that allow the passage of the light without scattering; thus doesn’t affect the degree of conversion and the depth of cure of the adhesive system14.\n\nGroups of adhesive system with incorporated nanosilver solution recorded statistically significant lower degree of conversion when compared to the control and nanosilver powder (p≤0.05). Presence of excessive amount of ethanol (Over 10% of the neat resin blend) lead to dilution of the adhesive resin, (decrease the percent of polymerized resin). Moreover it can cause physical separation of some reactive components of the adhesive resin with subsequent reduction of the degree of conversion2. Previous research has attributed the negative effect of excess ethanol on the degree of conversion to its cooling effect (polymerization reaction is exothermic and the liberated heat increase the rate of conversion). Ethanol can absorb the liberated heat thus decrease the rate of polymerization and the degree of conversion of the adhesive15.\n\nThere were no recorded studies evaluated the effect of nanosilver (powder form or ethanol based solution) incorporation in the self-etch adhesive system on the degree of conversion.\n\n\nConclusion\n\nThe antibacterial efficacy of the adhesive system can be greatly potentiated with the addition of silver nanoparticles (12µg/mL concentration) especially the nanosilver powder. Incorporation of the antibacterial nanosilver powder in the adhesive system didn’t compromise the degree of conversion of the adhesive resin.\n\nFurther investigation is required for assessing the mechanical behavior and chemical reactions of adhesive systems containing silver nanoparticles in short and long-term. It is also recommended to evaluate the antibacterial activity of adhesive system containing silver nanoparticles against dental plaque biofilm rather than single bacterial species.\n\n\nData availability\n\nOpen Science Framework: The effect of silver nanoparticles incorporation in the self-etch adhesive system on its antibacterial activity and degree of conversion: an In-vitro Study. https://doi.org/10.17605/OSF.IO/RS4D211\n\nThis project contains the following underlying data:\n\n- Raw Data DC note pads (folder containing out files from ATR/FTIR)\n\n- raw Data final DC.docx (ATR/FTIR data with explanation of analysis pipeline)\n\n- results heba fathy.docx (inhibition zone measurements)\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 author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nKasraei S, Sami L, Hendi S, et al.: Antibacterial properties of composite resins incorporating silver and zinc oxide nanoparticles on Streptococcus mutans and Lactobacillus. Restor Dent Endod. 2014; 39(2): 109–114. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang Y, Wang Y: Photopolymerization of phosphoric acid ester-based self-etch dental adhesives. Dent Mater J. 2013; 32(1): 10–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAydin Sevinç B, Hanley L: Antibacterial activity of dental composites containing zinc oxide nanoparticles. J Biomed Mater Res B Appl Biomater. 2010; 94(1): 22–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFaria-E-Silva AL, Piva E, Moraes RR: Time-dependent effect of refrigeration on viscosity and conversion kinetics of dental adhesive resins. Eur J Dent. 2010; 4(2): 150–155. PubMed Abstract | Free Full Text\n\nKim JS, Kuk E, Yu KN, et al.: Antimicrobial effects of silver nanoparticles. Nanomedicine. 2007; 3(1): 95–101. PubMed Abstract | Publisher Full Text\n\nEspinosa-Cristóbal LF, Martínez-Castañón GA, Martínez-Martínez RE, et al.: Antibacterial effect of silver nanoparticles against Streptococcus mutans. Mater Lett. 2009; 63(29): 2603–2606. Publisher Full Text\n\nLi F, Li Z, Liu G, et al.: Long-term Antibacterial Properties and Bond Strength of Experimental Nano Silver-containing Orthodontic Cements. J Wuhan Univ Technol 2013; 28(4): 849–855. Publisher Full Text\n\nFeuerstein O, Matalon S, Slutzky H, et al.: Antibacterial properties of self-etching dental adhesive systems. J Am Dent Assoc. 2007; 138(3): 349–354; quiz 396–8. PubMed Abstract | Publisher Full Text\n\nElkorashy ME, Shalaby H, Khafagi M: Effect Of Curing Distance On The Degree Of Conversion And Microhardness Of Nano-Hybrid Resin Composites. Egyptian Dental Journal. 2013; 59(4): 4647–4653. Reference Source\n\nWegehaupt FJ, Lunghi N, Belibasakis GN, et al.: Influence of light-curing distance on degree of conversion and cytotoxicity of etch-and-rinse and self-etch adhesives. BMC Oral Health. 2016; 17(1): 12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFathy H: The effect of silver nanoparticles incorporation in the self-etch adhesive system on its antibacterial activity and degree of conversion: an In-vitro Study. 2019. http://www.doi.org/10.17605/OSF.IO/RS4D2\n\nTaha MY, Al-Shakir NM, Al-Sabawi NA: Antibacterial Effect of Dentin Bonding Agents: (An in vitro Study). Al– Rafidain Dent J. 2012; 12(2): 228–234. Reference Source\n\nKim M, Suh BI, Shin D, et al.: Comparison of the Physical and Mechanical Properties of Resin Matrix with Two Photoinitiator Systems in Dental Adhesives. Polymers J. 2016; 8(7): 250. Publisher Full Text\n\nDi Hipólito V, Reis AF, Mitra SB, et al.: Interaction morphology and bond strength of nanofilled simplified-step adhesives to acid etched dentin. Eur J Dent. 2012; 6(4): 349–360. PubMed Abstract | Free Full Text\n\nMoharam LM, Botros SA, El-Askary FS, et al.: Effect of polymerization protocol on the degree of conversion of photo- and dual-polymerized selfetch adhesives. J Adhes Sci Technol. 2015; 30(3): 262–274. Publisher Full Text"
}
|
[
{
"id": "67096",
"date": "03 Aug 2020",
"name": "Arivalagan Pugazhendhi",
"expertise": [
"Reviewer Expertise Nanoparticles",
"Anticancer",
"Photocatalytic degradation",
"Cytotoxicity",
"Organic and inorganic Nanoparticles",
"Antibacterial"
],
"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 reports the effect of silver nanoparticles incorporation in the self-etch adhesive system on its antibacterial activity and degree of conversion: an in-vitro study. Overall the experiments are performed well and also the manuscript was well written, I recommend this research manuscript is appropriate to be indexed with minor revision mentioned below.\n\nAvoid using words such as we, he, she, them, their, etc.\n\nLanguage is good, but, proof reading by a native speaker would avoid the minor errors.\n\nPlease concrete the keywords and make it formally and academically.\n\nRegarding the Introduction section, to make the reading more clear and smooth, it should be organized according to the following items: i) present state of the art; ii) literature review; iii) motivation and objective of the study proposed; iv) innovative contribution in terms of methodology developed.\n\nThe authors are suggested to read the following articles and cite them in the appropriate places. The introduction must be revised to enhance readability i.e. I would also like to see following references in the revised version:\n\nShanmuganathan et al. (20191). Jacob et al. (20192). Saravanan et al. (20183). Pugazhendhi et al. (20184). Shanmuganathan et al. (20185). Saravanan et al. (20186). Samuel et al. (20207).\n\nTypographical errors are present throughout the manuscript. The authors are required to pay keen attention to this.\n\nIn the materials and methods, divide them into nice sub-sections. Provide the details of all the equipment’s, instruments used, their model number, company of manufacture, country, etc.\n\nIn the introduction section, write the novelty of the work and the problem statement clearly.\n\n70% of the references should be from 2017, 2018 and 2019. Kindly do a careful literature review.\n\nThe results and discussion section is only to write your results and facilitate scientific/technical discussions. Provide mechanism-based reactions and refer to important/recent literatures in the results and discussions.\n\nConclusions (<100 words) should be in line with the specific objectives of your work. Do not repeat the results and the methodology here.\n\nDelete unwanted old references. Refer to references from the years 2017, 2018 and 2019. Figure numbers must be rearranged in this manuscript.\n\nThe authors must compare more results with previous publications mainly in characterization parts (Results and discussion).\n\nOnly few previous publications in discussions. Why did the authors not compare the current results with previous publications?\n\nThe authors are required to re-write the conclusion.\n\nThe authors must check the manuscript carefully before submitting to the journal. Because the manuscript still has grammar mistakes and some words are joined with some other words.\n\nThe authors must show good results in the abstract section to enhance the readers' understanding.\n\nµg/ml or µgm/ml. Check the full the manuscript.\n\nThe authors must use any one format throughout the manuscript for example minutes must be as “min”; “ml” must be as “mL” and hours must be as “h”. throughout the manuscript.\n\nPlease avoid reference overkill/run-on - do not use more than 3 references per sentence. If you need to use more, make sure you state the key relevant idea of each reference.\n\nMany sentences could be seen without enough evidences or justifications - you need appropriate references to justify your arguments.\n\nAdditionally, please pay attention to revise/check the following:\nEnglish language throughout the manuscript. Please ensure to have a proof reading to your manuscript. Similarity index shall not exceed 15% with no more than 1% from any source. Citation of references with latest publications on the topic. Please also cite some relevant papers from F1000Research. Manuscript shall have an adequate number of Figures and Tables. Figures must be in high quality format. Please ensure also to submit a high quality Graphical Abstract representing a summary of your study.\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": "68004",
"date": "10 Aug 2020",
"name": "Sashidhar Rao Beedu",
"expertise": [
"Reviewer Expertise Nanotechnology"
],
"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 details application of nano silver incorporation into a self-etch dental adhesive system. The antibacterial efficacy of the adhesive system was tested, in vitro. Even though the experimental approach seems to be fine, the experimental design and technical details appear to be poor. This manuscript requires additional details and clarification as detailed below:\n\nComments:\nWhat is the source of nanosilver powder? If silver nitrate is used, then the calculation needs to be based on elemental Ag.\n\nIf AgNPs are insoluble, then how can you get solution? It may form an emulsion or suspended particles only.\n\nWhat is the wave length (λ) of LED?\n\nTable 1: Silver Nitrate powder is not nano silver. The valency of nano silver is ‘zero’. If you have nano-silver powder, why are you reducing it again by sodium borohydride?\n\nThe authors should provide an additional representative figure showing the zone of inhibition (ZOI), in addition to table 2. Delete fig. 1.\n\nThe authors need to check the silver nitrate ions as additional control in the agar – antimicrobial assay.\n\nIf the adhesive solidifies along with nano-silver, then how does it diffuse in agar medium?\n\nWhat happens beyond 6.25µg/ml concentration of silver nano? What is the rationale to consider 10 mm ZOI as cut-off? It is appropriate to establish the IC50 value for the nano metal.\n\nRecommendation: the manuscript is not acceptable in the present form.\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? No\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-244
|
https://f1000research.com/articles/8-243/v1
|
01 Mar 19
|
{
"type": "Research Article",
"title": "The potential impact of age, gender, body mass index, socioeconomic status and dietary habits on the prevalence of dental caries among Egyptian adults: a cross-sectional study",
"authors": [
"Marwa M.S. Abbass",
"Nermeen AbuBakr",
"Israa Ahmed Radwan",
"Dina Rady",
"Sara El Moshy",
"Mohamed Ramadan",
"Attera Ahmed",
"Ayoub Al-Jawaldeh",
"Nermeen AbuBakr",
"Israa Ahmed Radwan",
"Dina Rady",
"Sara El Moshy",
"Mohamed Ramadan",
"Attera Ahmed",
"Ayoub Al-Jawaldeh"
],
"abstract": "Background: Dental caries is a major public health problem and the most widespread chronic disease to affect individuals throughout their lifetime. Little information exists about the prevalence of dental caries among Egyptian adults. Therefore, this study investigated the dental caries experience among Egyptian adults in correlation with different risk factors. Methods: A total of 359 Egyptian adults (age range, 18-74 years) were examined over a period of 3 months, starting on the 15th of November 2017 until the 13th of January 2018. Socio-demographic data, brushing frequency, body mass index (BMI) and eating habits were recorded and collected using a questionnaire. Dental examination was performed using the Decayed, Missing and Filled tooth (DMFT) index. Results: In total, 86.63% of participants had dental caries experience. Of the participants, 60.45%, 48.47% and 55.43% had at least one decayed, missing and filled tooth, respectively. The mean number of decayed, missing, filled or DMFT for the whole sample were 2.4±3.6, 1.98±3.99, 1.79±2.45, 6.09±5.7, respectively. Decayed teeth were inversely correlated with socio-economic status (SES), education level, brushing frequency and milk consumption and positively correlated with grains, junk food and soda drinks consumption. Missing teeth were inversely correlated with SES, education level and brushing frequency, while positively correlated with age, BMI and caffeinated drink consumption. Conversely, filled teeth were positively correlated with age, BMI, SES and education level, while negatively correlated with grains and sugars in drinks. Conclusion: The present study clarifies that age, BMI, SES, education level and brushing frequency are risk factors significantly associated with dental caries prevalence amongst Egyptian adults. Egyptian adults' dietary habits might lead to obesity, which indirectly causes dental caries rather than directly as in children.",
"keywords": [
"Caries",
"Prevalence",
"Age",
"Socioeconomic",
"Dietary",
"Education",
"Egyptian",
"Adults"
],
"content": "Introduction\n\nDental caries is the progressive destruction of the tooth structure by bacterial acids1. It is considered the most prevalent chronic oral disease and the primary reason for tooth loss in adults. Dental caries has been estimated to affect almost every individual during their adult life, affecting an average of 5 to 10 teeth per each individual1–3. The prevalence of caries in a population is affected by numerous risk factors, such as sex, oral hygiene and dietary habits. Moreover, the prevalence of caries tends to increase with age as it is a cumulative process3–6. Tooth decay is particularly prevalent in developing countries owing to the dietary habits, socioeconomic conditions and a lack of education7.\n\nData on the incidence of dental caries among Egyptian adults are scarce and are mostly grey literature, which makes this data hard to find. The last published report for the prevalence of caries among Egyptian adults was carried out by the World Health Organization (WHO) in collaboration with the Egyptian Ministry of Health in 20148.\n\nTherefore, the current study was carried out to investigate the prevalence of caries among Egyptian adults in correlation with different risk factors.\n\n\nMethods\n\nThis study was carried out according to the regulations of the Research Ethics Committee of Faculty of Dentistry, Cairo University, Egypt (Approval:171217). Written informed consent was obtained from patients before participating in the study.\n\nThe subjects in this study were recruited between November 2017 to January 2018, from the outpatients' clinics of Faculty of Dentistry, Cairo University, which serves patients arriving from different parts of Egypt; in addition to two private dental clinics.\n\nThe inclusion criteria for the patients were: age, 18–74 years; either gender; ethnicity, Egyptian. The exclusion criteria were: history of radiotherapy and/or chemotherapy; subjects undergoing orthodontic therapy and patients who might not comply with study procedures (as judged by those who refused to answer all questions in the questionnaire9).\n\nThe sample size for caries in adults was estimated to be 264 individuals according to the following equation:\n\n\n\nn' = sample size with finite population correction, N = Egyptian adults population size (estimated by 45,000,000), Z = Z statistic for a level of confidence which is conventional (1.96). P = Expected prevalence (78.00%) and d = Precision (5% = (0.05)). The prevalence was estimated as 78%, as in India the caries prevalence was estimated to be 82.6–91.6%11 and in Kosovo was 72.8%12.\n\nA total of 359 Egyptian adult patients were examined in this study. The collected data included name, age, gender, address, level of education; low (primary school or illiterate); moderate (diploma or high school); high (university), occupation and brushing frequency. A full assessment of the dietary habits was performed using a questionnaire9.\n\nSubjects were instructed to remove only their shoes while taking their anthropometric measurements. Weight was measured using a Beurer scale (Ulm, Germany) and their stature was measured to the nearest 0.1 cm using a stadiometer. Body mass index (BMI) was calculated using the formula: BMI = weight in kg / (height in m)2. Patients were categorized according to their BMI as follows: underweight (<18.5 kg/m2), normal (18.5–24.9 kg/m2), overweight (25.0–29.9 kg/m2) and obese (≥30.0 kg/m2)13.\n\nAdults were classified according to their age as follows: BI (18–34 years old), BII (35–44 years old), BIII (45–64 years old) and BIV (65–75 years old) and were classified according to their socioeconomic status into low, moderate and high depending on their level of education, occupation and address14.\n\nAt the beginning, the authors (M.M.S., S.E., D.R., N.A., I.A.R.) were calibrated to avoid differences in observations and to reach a consensus. Oral examination was carried out according to WHO recommendations15, as described in our previous study16. All present teeth were examined for the presence of carious lesions. Teeth were carefully inspected for the presence of any lesion with a softened floor, undermined enamel, or softened wall, in a pit or fissure or on a smooth tooth surface. Tooth surface containing temporary filling or a permanent restoration but showing an area of decay (either primary or secondary caries) were also considered carious. DMFT index was used to measure the activity of caries, where D denotes decayed teeth, M is for missing teeth and F is for filled teeth17.\n\nThe statistical package used in this study is the R statistical package, version 3.3.1 (2016-06-21). For descriptive analyses, variables were described in terms of means ± standard deviations (SD), medians and ranges. For normality, Shapiro-Wilk test was applied to assess the normality of data. All data were not normally distributed. For comparative analysis, the non-parametric Kruskal-Wallis test was performed. Spearman’s correlation coefficient was calculated for correlation analysis. The significance level was verified at P ≤ 0.05.\n\n\nResults\n\nRaw data from the present study are available on figshare18.\n\nThe mean number of decayed, missing, filled and DMFT for the whole sample were 2.4±3.6, 1.98±3.99, 1.79±2.45, 6.09±5.7, respectively. A total of 60.45%, 48.47% and 55.43% of participants had at least one decayed, missing or filled tooth, respectively. The prevalence of DMFT among participants was 86.63%.\n\nThe number and percentage of adults in different categories in each studied parameter as well as comparisons between them are presented in Table 1. The number and percentage of participants that had decayed, missing, filled or DMFT in different categories in each studied parameter are presented in Table 2.\n\n*Statistical significance at p-value ≤ 0.05.\n\nAs shown in Table 3 and Figure 1, despite the highest means of decayed teeth were recorded in adults aged 65–75 years, in males and in adults of normal weight (3.42±7.39, 2.61±3.68 and 2.83±3.77, respectively), the differences in medians were statistically insignificant (p≥0.05) and there was no correlation between age, gender or BMI and decayed teeth (Spearman’s rho=-0.04, 0.08 and -0.07, respectively; p≥0.05).\n\nThe correlation coefficient, rho, ranges from -1 to +1. Where 1= perfect positive correlation, 0=no correlation, -1 = perfect negative (inverse) correlation. *Statistical significance at p-value ≤ 0.05.\n\nN= 359: *positive correlations; **inverse correlations. A: ≤2 times/week, B: 3–6 times/week, C: 1–6 times/day.\n\nAdults with low SES, low education level and who do not brush their teeth had the highest mean number of decayed teeth (4.08±4.63, 4.78±5.97 and 4.26±5.76, respectively). The differences in medians were statistically significant (p≤0.0001). SES, education level and brushing frequency were inversely correlated with number of decayed teeth (Spearman’s rho= -0.28, -0.24, and -0.26, respectively; p≤0.0001).\n\nAdults who consume bread, eggs, fruits/vegetables, grains, crackers, junk food, chocolate, soft drinks and caffeinated drinks 1–6 times per day had the highest mean number of decayed teeth (2.48±3.70, 3.11±2.54, 2.54±3.98, 2.90±4.16, 2.67±3.2, 3.34±3.87, 2.43±3, 3.22±4.47 and 2.48±3.74, respectively). Milk consumption was significantly inversely correlated with decayed teeth (rho= -0.15, p=0.0037), whereas consumption of grains, junk food and soft drinks were significantly positively correlated (rho=0.13, 0.15 and 0.13; p=0.0143, 0.0058 and 0.0121, respectively). The differences in medians for decayed teeth in all dietary elements were statistically insignificant except for eggs, milk, grains and junk food.\n\nAs revealed in Table 4 and Figure 2, the highest mean number of missing teeth were detected in adults aged 65–75 years, besides in obese adults (6.08±6.64 and 3.51±5.56, respectively). The differences in medians were statistically significant (p≤0.0001). Age and BMI were directly correlated with number of missing teeth (Spearman’s rho=0.44 and 0.25, respectively; p≤0.0001).\n\nThe correlation coefficient, rho, ranges from -1 to +1. Where 1= perfect positive correlation, 0=no correlation, -1 = perfect negative (inverse) correlation. *Statistical significance at p-value ≤ 0.05.\n\nN= 359: *positive correlations; **inverse correlations. A: ≤2 times/week, B: 3–6 times/week, C: 1–6 times/day.\n\nHowever, males had a higher mean missing teeth than females (2.07±4.38), the difference in medians was statistically insignificant (p≥0.05) and there was no correlation between gender and missing teeth (Spearman’s rho= -0.02, p≥0.05).\n\nAdults with low SES, low education level and those who don’t brush their teeth had the highest means missing teeth (3.23±4.84, 3.78±5.09 and 3.34±5.57, respectively). The differences in medians were statistically significant (p≤0.0001). SES, education level and brushing frequency were inversely correlated with number of missing teeth (Spearman’s rho= -0.25, -0.3 and -0.23, respectively; p≤0.0001).\n\nAdults who consume eggs, milk, sugar in drinks and caffeinated drinks 1–6 times per day had the highest means missing teeth (3.11±6.01, 2.27±4.78, 2.22±4.35 and 2.10±4.01, respectively); while those who consume bread, other carbohydrates, fruits/vegetables, milk products, grains, sugar, jam, candies, crackers, junk food, chocolate, soft drinks, juices and citric juices less than or equal to two times a week had the highest means of missing teeth (2.31±4.35, 3±4.44, 2.09±3.82, 2.34±4.44, 2.13±4.13, 2.33±4.4, 2.04±4.09, 2.47±4.42, 2.57±4.62, 2.49±4.68, 2.32±4.21, 2.40±4.45, 2.42±4.18 and 2.06±3.92, respectively). The differences in medians for missing teeth in all parameters were statistically insignificant except for other carbohydrates, sugar not in drinks, candies, crackers, junk food, chocolate, soft drinks and juices. Consumption of sugars not included in drinks, candies, crackers, junk food, chocolate, soft drinks, juices and caffeinated drinks were inversely correlated with missing teeth.\n\nAs seen in Table 5 and Figure 3, adults aged 65–75 years and obese adults had the highest mean numbers of filled teeth (1.83±1.64 and, 3.51±5.56, respectively). The difference in medians was statistically insignificant for age (p->0.05), while was statistically significant for BMI (p=0.0223). Age and BMI were directly correlated with number of filled teeth (Spearman’s rho=0.13 and 0.16, and p=0.0138 and, 0.002, respectively).\n\n*P<0.05.\n\nN= 359: *positive correlations; **inverse correlations. A: ≤2 times/week, B: 3–6 times/week, C: 1–6 times/day.\n\nMales had a higher mean number of filled teeth than females (1.81±2.4), while adults who don’t brush their teeth had the lowest mean number of filled teeth (1.39±1.79). The differences in medians were statistically insignificant (p>0.05) and there was no correlation between gender or brushing frequency and number of filled teeth (Spearman’s rho=-0.02 and 0.04, respectively; p>0.05).\n\nAdults with high SES and high education levels had the highest mean number of filled teeth (2.26±2.74 and 2.24±2.72, respectively); the differences in medians were statistically significant (p=0.0016 and 0.0002, respectively). SES and education level were directly correlated with filled teeth (Spearman’s rho=0.18 and 0.21; p=0.0004 and <0.0001, respectively).\n\nAdults who consume eggs, fruits/vegetables, milk, milk products, grains, sugar in drinks, sugar not in drinks, jam, candies, junk food, chocolate, soft drinks, citric juices and caffeinated drinks 3–6 times per week had the highest means of filled teeth (2.01±2.8, 2.18±2.52, 2.48±2.56, 2.05±2.1, 3±2.69, 3.67± 2.9, 3.38±2.47, 2.3±2.14, 2.28±2.96, 3.72±3.1, 2.13±2.67, 2.77±2.78, 2.64±3.07 and 2.93±2.97, respectively). Moreover, adults who consume bread, other carbohydrates, crackers and juices less than or equal to two times a week had the highest mean numbers of filled teeth (2.62±3.36, 2.09±2.83, 2.02±2.66 and, 2.03±2.71, respectively).\n\nThe differences in medians for filled teeth regarding all dietary elements were statistically insignificant except for grains, sugar in drinks, sugar not in drinks, jam, junk food and soft drinks. Grains, sugar in drinks and crackers were inversely correlated with filled teeth, while jams were borderline positively correlated.\n\nTable 6, Figure 4 show that the highest means DMFT was in adults aged 65–75 years and obese adults (11.42±7.63 and 7.4±5.89, respectively). The differences in medians were statistically significant (p≤0.0001). Age and BMI were directly correlated with DMFT (Spearman’s rho=0.33 and 0.15; p≤0.0001 and 0.0053, respectively).\n\nThe correlation coefficient, rho, ranges from -1 to +1. Where 1= perfect positive correlation, 0=no correlation, -1 = perfect negative (inverse) correlation. *Statistical significance at p-value ≤ 0.05.\n\nN= 359: *positive correlations; **inverse correlations. A: ≤2 times/week, B: 3–6 times/week, C: 1–6 times/day.\n\nMales had a higher mean DMFT teeth than females (6.5±5.47). The difference in medians was statistically insignificant (p≥0.05) and there was no correlation between gender and DMFT (Spearman’s rho= -0.09, p≥0.05).\n\nAdults with low SES, low education level and those who don’t brush their teeth had the highest mean DMFT numbers (8.2±6.24, 9.48±6.93 and 8.95±6.71, respectively). The differences in medians were statistically significant (p<0.0001). SES, education level and brushing frequency were inversely correlated with DMFT (Spearman’s rho=-0.19, -0.21 and -0.3; p=0.0002, <0.0001 and <0.0001, respectively).\n\nThe highest means of DMFT, were recorded in adults who consume eggs, jam and caffeinated drinks 1–6 times per day (7.69±8.06, 6.14±6.04 and 6.25±5.83, respectively) and those who consume grains and soft drinks 3–6 times per week (6.41±5.8 and 6.8±4.59, respectively), as well as those who consume bread, other carbohydrates, fruits/vegetables, milk, milk products, sugar in drinks, sugar not in drinks, candies, crackers, junk food, chocolate, juices and citric juices two or fewer times a week (6.69±5.23, 7.05±6.42, 6.29±5.25, 6.43±5.75, 6.66±5.95, 6.34±4.76, 6.51±6.06, 6.8±6.17, 6.85±6.4, 6.42±6.2, 6.51±5.98, 6.78±5.54 and 6.25±5.84, respectively). The differences in medians for DMFT in all dietary elements were statistically insignificant except for candies and juices. Consumption of candies, chocolate and juices were inversely correlated with DMFT.\n\n\nDiscussion\n\nTo our knowledge, the present study is the first to clarify the prevalence of dental caries and treatment needs among Egyptian adults and their correlation with different risk factors. Egyptian adults proved to be at greater risk of developing caries, with a total prevalence of 86.63%; as compared to children and adolescence, for whom a value of 74% was recorded16. The recorded DMFT in the present investigation is lower than that recorded in North-West Russia (96%)19, higher than that recorded in Turkey (62%)20 and much higher than the prevalence of caries in England (31%)21.\n\nIn the current work, DMFT with its components M and F showed a significant increase with age (P<0.0001), this is in agreement with previous studies carried out in Australia and China22,23, but contrary to a study performed in Turkey, which reported a decrease in dental caries with age20.\n\nNo significant correlation between gender and caries has been reported in the current investigation. This is in contrast with other studies, which reported higher caries indices in females23,24, while another study reported a significant increase in decayed teeth in males25.\n\nBoth dental caries and BMI are diet-related health measures. The incidence of caries and obesity has raised in the last two decades due to changes in lifestyle and diet26,27. The reported significant positive correlation between BMI, DMFT and missing teeth in the current study is in accordance with findings by Sheiham et al., who showed that British people that had less than 20 teeth were more likely to be obese28; however, a study performed on Korean adults revealed an inverse correlation29.\n\nSES and education level are thought to be strongly associated with dental caries rate. In the present work, there were inverse correlations between socioeconomic, education levels and number of decayed, missing teeth and DMFT, while a positive correlation with filled teeth. These findings reinforce those of previous studies30–33 and are partially in agreement with those of Ceylan et al.34, who found that the mean number of filled teeth was strongly correlated with income level, while DMFT was not correlated with income and education level.\n\nRegarding oral hygiene, DMFT and the mean number of missing and decayed teeth were significantly higher in adults who don’t brush their teeth (8.95±6.71, 3.34±5.57, 4.26±5.76, respectively). These results are similar to studies conducted by Fukuda et al.35 and Levin et al.36, who reported that regular tooth brushing improved the oral health.\n\nDietary habits are important contributors to the heath or disease of a population. Alteration in dietary habits, like increased consumption of refined sugars, soft drinks and fast food, cause caries as well as obesity37. In the present study, it was found that adults who consumed bread, grains, crackers, junk food, chocolate, soft drinks and caffeinated drinks 1–6 times per day had the highest mean number of decayed teeth, with a significant positive correlation observed with grain, junk food and soft drink consumption. This is in accordance with results obtained by Jones et al., who reported a significant correlation between soft drink consumption and DMFT index38.\n\nDespite evidence from previous studies, which revealed a correlation between caries incidence and sugar intake in children39,40 and in adults34, the present investigation reported weak correlations or even negative correlations between cariogenic food and the number of missing, filled teeth and DMFT, which are in accordance with other studies41–43. Consumption of sweetened foods and drinks between meals usually leads to the development of caries in children, while is associated with obesity in adults44. An indication of an answer to the controversial question “how obesity causes dental caries?” could be found in a recent cross-sectional study carried out on adolescences. The authors investigated the relationship between obesity and bite force. They inferred that the decreased bite force reported in obese males and females might result in a preference for soft food stuffs and a reduction in chewing, which in turn might cause caries. In contrast, individuals with normal body weight, have increased bite force and choose harder foodstuffs45.\n\nOn the contrary, some diets may favour remineralization when their content is high in calcium, phosphate and protein46. This is confirmed by the significant negative correlation between milk consumption and the number of decayed teeth reported in the current study.\n\nA limitation of this cross-sectional study is that the collected dietary information covered only a few short periods in time, which may not be an accurate representation of an individual's actual lifetime dietary habits. This could explain the lack of significant differences between some caries indices and cariogenic dietary elements. The DMFT index is the most widely used index worldwide, but it has some limitations. The F or M component might not only display teeth that were previously decayed and could include other conditions not related to dental caries47.\n\nFrom this study, it could be concluded that age, BMI, SES, education level and brushing frequency are risk factors significantly associated with caries prevalence amongst Egyptian adults.\n\n\nData availability\n\nRaw data, including all answers to the questionnaire and data on caries incidence amongst the sampled population, are available on figshare. DOI: https://doi.org/10.6084/m9.figshare.7609832.v118.\n\nA copy of the study questionnaire is available on figshare. DOI: https://doi.org/10.6084/m9.figshare.7609733.v19.\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\nThis study was supported by the World Health Organization.\n\n\nAcknowledgments\n\nWe would like to acknowledge the support and technical guidance of nutrition unit at World Health Organization office for Eastern Mediterranean region.\n\n\nReferences\n\nKidd EA, Giedrys-Leeper E, Simons D: Take two dentists: a tale of root caries. Dent Update. 2000; 27(5): 222–230. PubMed Abstract | Publisher Full Text\n\nPetersen PE, Bourgeois D, Ogawa H, et al.: The global burden of oral diseases and risks to oral health. Bull World Health Organ. 2005; 83(9): 661–9. PubMed Abstract | Free Full Text\n\nShingare P, Jogani V, Sevekar S, et al.: Dental Caries Prevalence among 3 to 14 years old school children, Uran, Raigad District, Maharashtra. J Contemporary Dent. 2012; 2(2): 11–14. Publisher Full Text\n\nGriffin SO, Griffin PM, Swann JL, et al.: New coronal caries in older adults: implications for prevention. J Dent Res. 2005; 84(8): 715–720. PubMed Abstract | Publisher Full Text\n\nBurt BA: Concepts of risk in dental public health. Community Dent Oral Epidemiol. 2005; 33(4): 240–247. PubMed Abstract | Publisher Full Text\n\nPetersen P, Razanamihaja N, Poulsen VJ: Surveillance of Oral health among children and adults in Madagascar. WHO, Geneva, Switzerland. 2004.\n\nKulkarni SS, Deshpande SD: Caries prevalence and treatment needs in 11-15 year old children of Belgaum city. J Indian Soc Pedod Prev Dent. 2002; 20(1): 12–15. PubMed Abstract\n\nhttp://www.emro.who.int/egy/egypt-events/results-of-epidemiological-study-on-oral-health-status-released.html. 2014.\n\nAbbass M: questionnaire adult.pdf. figshare. 2019. http://www.doi.org/10.6084/m9.figshare.7609733.v1\n\nDaniel WW: Biostatistics: A Foundation for Analysis in the Health Sciences. 7th edition. New York: John Wiley & Sons. 1999. Reference Source\n\nPatro BK, Ravi Kumar B, Goswami A, et al.: Prevalence of dental caries among adults and elderly in an urban resettlement colony of New Delhi. Indian J Dent Res. 2008; 19(2): 95–8. PubMed Abstract | Publisher Full Text\n\nKamberi B, Koçani F, Begzati A, et al.: Prevalence of Dental Caries in Kosovar Adult Population. Int J Dent. 2016; 2016: 4290291. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorld Health Organization: Expert Committee on Physical Status: the use and interpretation of anthropometry physical status. WHO Techniques Report Series, Geneva: World Health Organization. 1995; 854. Reference Source\n\nEl-Gilany A, El-Wehady A, El-Wasify M: Updating and validation of the socioeconomic status scale for health research in Egypt. East Mediterr Health J. 2012; 18(9): 962–968. PubMed Abstract | Publisher Full Text\n\nWorld Health Organization: Oral Health Surveys: Basic Methods. World Health Organization, Geneva, Switzerland. 1997. Reference Source\n\nAbbass MM, Mahmoud SA, El Moshy S, et al.: The prevalence of dental caries among Egyptian children and adolescences and its association with age, socioeconomic status, dietary habits and other risk factors. A cross-sectional study [version 1; referees: awaiting peer review]. F1000Res. 2019; 8. Publisher Full Text\n\nWorld Health Organization: Oral health surveys: Basic methods. 5th ed., Geneva, Switzerland, World Health Organization. 2013. Reference Source\n\nAbbass M: Raw data for caries incidence in correlation to risk factors in egyptian adults.xlsx. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.7609832.v1\n\nDrachev SN, Brenn T, Trovik TA: Dental caries experience and determinants in young adults of the Northern State Medical University, Arkhangelsk, North-West Russia: a cross-sectional study. BMC Oral Health. 2017; 17(1): 136. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNamal N, Can G, Vehid S, et al.: Dental health status and risk factors for dental caries in adults in Istanbul, Turkey. East Mediterr Health J. 2008; 14(1): 110–8. PubMed Abstract\n\nWhite DA, Tsakos G, Pitts NB, et al.: Adult Dental Health Survey 2009: common oral health conditions and their impact on the population. Br Dent J. 2012; 213(11): 567–72. PubMed Abstract | Publisher Full Text\n\nBrennan DS, Spencer AJ: Changes in caries experience among Australian public dental patients between 1995/96 and 2001/2. Aust N Z J Public Health. 2004; 28(6): 542–8. Publisher Full Text\n\nWang HY, Petersen PE, Bian JY, et al.: The second national survey of oral health status of children and adults in China. Int Dent J. 2002; 52(4): 283–90. PubMed Abstract | Publisher Full Text\n\nVarenne B, Petersen PE, Quattara S: Oral health status of children and adults in urban and rural areas of Burkina Faso, Africa. Int Dent J. 2004; 54(2): 83–9. PubMed Abstract | Publisher Full Text\n\nHenriksen BM, Ambjørnsen E, Axéll T: Dental caries among the elderly in Norway. Acta Odontol Scand. 2004; 62(2): 75–81. PubMed Abstract | Publisher Full Text\n\nBooth ML, Dobbins T, Okely AD, et al.: Trends in the prevalence of overweight and obesity among young Australians, 1985, 1997, and 2004. Obesity (Silver Spring). 2007; 15(5): 1089–1095. PubMed Abstract | Publisher Full Text\n\nDye BA, Shenkin JD, Ogden CL, et al.: The relationship between healthful eating practices and dental caries in children aged 2-5 years in the United States, 1988-1994. J Am Dent Assoc. 2004; 135(1): 55–66. PubMed Abstract | Publisher Full Text\n\nSheiham A, Steele JG, Marcenes W, et al.: The relationship between oral health status and Body Mass Index among older people: a national survey of older people in Great Britain. Br Dent J. 2002; 192(12): 703–706. PubMed Abstract | Publisher Full Text\n\nSong IS, Han K, Ryu JJ, et al.: Obesity is inversely related to the risks of dental caries in Korean adults. Oral Dis. 2017; 23(8): 1080–6. PubMed Abstract | Publisher Full Text\n\nFukuda Y, Nakamura K, Takano T: Accumulation of health risk behaviours is associated with lower socioeconomic status and women's urban residence: a multilevel analysis in Japan. BMC Public Health. 2005; 5: 53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTafere Y, Chanie S, Dessie T, et al.: Assessment of prevalence of dental caries and the associated factors among patients attending dental clinic in Debre Tabor general hospital: a hospital-based cross-sectional study. BMC Oral Health. 2018; 18(1): 119. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSenna A, Campus G, Gagliani M, et al.: Socio-economic influence on caries experience and CPITN values among a group of Italian call-up soldiers and cadets. Oral Health Prev Dent. 2005; 3(1): 39–46. PubMed Abstract\n\nSkuduyte R, Aleksejuniene J, Eriksen HM: Dental caries in adult Lithuanians. Acta Odontol Scand. 2000; 58(4): 143–7. PubMed Abstract | Publisher Full Text\n\nCeylan S, Açikel CH, Okçu KM, et al.: Evaluation of the dental health of the young adult male population in Turkey. Mil Med. 2004; 169(11): 885–9. PubMed Abstract | Publisher Full Text\n\nFukuda H, Shinsho F, Nakajima K, et al.: Oral health habits and the number of teeth present in Japanese aged 50-80 years. Community Dent Health. 1997; 14(4): 248–52. PubMed Abstract\n\nLevin L, Shenkman A: The relationship between dental caries status and oral health attitudes and behaviour in young Israeli adults. J Dent Educ. 2004; 68(11): 1185–91. PubMed Abstract\n\nAlm A, Fahraeus C, Wendt LK, et al.: Body adiposity status in teenagers and snacking habits in early childhood in relation to approximal caries at 15 years of age. Int J Paediatr Dent. 2008; 18(3): 189–196. PubMed Abstract | Publisher Full Text\n\nJones C, Woods K, Whittle G, et al.: Sugar, drinks, deprivation and dental caries in 14-year-old children in the north west of England in 1995. Commun Dent Health. 1999; 16(2): 68–71. PubMed Abstract\n\nKalsbeek H, Verrips GH: Consumption of sweet snacks and caries experience of primary school children. Caries Res. 1994; 28(6): 477–83. PubMed Abstract | Publisher Full Text\n\nSzpunar SM, Eklund SA, Burt BA: Sugar consumption and caries risk in school children with low caries experience. Commun Dent Oral Epidemiol. 1995; 23(3): 142–6. PubMed Abstract | Publisher Full Text\n\nRugarabamu P, Frencken JE, Amuli JA, et al.: Caries experience amongst 12- and 15-year-old Tanzanian children residing on a sugar estate. Community Dent Health. 1990; 7(1): 53–8. PubMed Abstract\n\nLarsson B, Johansson I, Ericson T: Prevalence of caries in adolescents in relation to diet. Community Dent Oral Epidemiol. 1992; 20(3): 133–7. PubMed Abstract | Publisher Full Text\n\nStecksén-Blicks C, Arvidsson S, Holm AK: Dental health, dental care, and dietary habits in children in different parts of Sweden. Acta Odontol Scand. 1985; 43(1): 59–67. PubMed Abstract | Publisher Full Text\n\nMatsuoka Y, Fukai K: Adult dental caries and sugar intake. Health Sci Health Care. 2015; 15(1): 22–9. Reference Source\n\nSun KT, Chen SC, Li YF, et al.: Bite-force difference among obese adolescents in central Taiwan. J Formos Med Assoc. 2016; 115(6): 404–10. PubMed Abstract | Publisher Full Text\n\nTouger-Decker R, van Loveren C: Sugars and dental caries. Am J Clin Nutr. 2003; 78(4): 881S–892S. PubMed Abstract | Publisher Full Text\n\nAbreu MH, Drummond SN, Paixão HH, et al.: Correction factor for the M-component in the DMFS index in an adult brazilian population. Rev Odontol Univ Sao Paulo. 1998; 12(4): 323–328. Publisher Full Text"
}
|
[
{
"id": "45145",
"date": "12 Mar 2019",
"name": "Mohammed Grawish",
"expertise": [
"Reviewer Expertise Professor of Oral Biology",
"my research interest biomaterials",
"stem cells and integrative 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 paper by Abbass et al., evaluated clearly and accurately the potential impact of age, gender, body mass index, socioeconomic status and dietary habits on the prevalence of dental caries among Egyptian adults. The title and abstract are appropriate for the content of the text. The cross-sectional study design is appropriate and the work is technically sound. Sufficient details of methods and analyses were provided and it has an external validity as well as internal one and thus it can be reproduced. Non parametric statistical tests were performed as the data are categorical in type. The data obtained are interpreted appropriately and the results are valid. All the source data underlying the results are available and thus full reproducibility can be attained. The conclusions drawn are adequately supported by the results. In addition, the paper is well written and precisely formatted.\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": "45143",
"date": "15 Mar 2019",
"name": "Tahra Elobeid",
"expertise": [
"Reviewer Expertise Nutrition",
"food safety",
"food policy and planning"
],
"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\nWith the high prevalence of obesity and Non communicable diseases (NCDs), it is important to investigate how it affects the other underlying problems associated with it. This study provides very important findings as research in this area is still virgin in the region. The change in the food habits and nutrition transition further complicates this problem. Oral health and NCDs risk factors is established in several studies however the correlations were not significant. The association between dental health and obesity has not been studied in the region although there has been a significant change in the food habits and socioeconomic levels of the different populations. This will further increase the burden on public health system.\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": "45144",
"date": "10 May 2019",
"name": "Wafa El-Badrawy",
"expertise": [
"Reviewer Expertise Dental Materials and restorative"
],
"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 subject of the research is new for the area and it will be of value to the profession. The study is well designed, the manuscript is well written and edited however from the research point of view the study is lacking objectives and hypotheses. The methods are clear and well explained. In the result section: the number of tables is excessive which makes it slightly complicated for the reader to follow. Conclusion are clear and well written.\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? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-243
|
https://f1000research.com/articles/8-242/v1
|
01 Mar 19
|
{
"type": "Systematic Review",
"title": "Oximetry and neonatal examination for the detection of critical congenital heart disease: a systematic review and meta-analysis",
"authors": [
"Hernán Camilo Aranguren Bello",
"Dario Londoño Trujillo",
"Gloria Amparo Troncoso Moreno",
"Maria Teresa Dominguez Torres",
"Alejandra Taborda Restrepo",
"Alejandra Fonseca",
"Nestor Sandoval Reyes",
"Cindy Lorena Chamorro",
"Rodolfo José Dennis Verano",
"Hernán Camilo Aranguren Bello",
"Dario Londoño Trujillo",
"Gloria Amparo Troncoso Moreno",
"Alejandra Taborda Restrepo",
"Alejandra Fonseca",
"Nestor Sandoval Reyes",
"Cindy Lorena Chamorro",
"Rodolfo José Dennis Verano"
],
"abstract": "Background: Undiagnosed congenital heart disease in the prenatal stage can occur in approximately 5 to 15 out of 1000 live births; more than a quarter of these will have critical congenital heart disease (CCHD). Late postnatal diagnosis is associated with a worse prognosis during childhood, and there is evidence that a standardized measurement of oxygen saturation in the newborn by cutaneous oximetry is an optimal method for the detection of CCHD. We conducted a systematic review of the literature and meta-analysis comparing the operational characteristics of oximetry and physical examination for the detection of CCHD. Methods: A systematic review of the literature was conducted on the following databases including published studies between 2002 and 2017, with no language restrictions: Pubmed, Science Direct, Ovid, Scopus and EBSCO, with the following keywords: oximetry screening, critical congenital heart disease, newborn OR oximetry screening heart defects, congenital, specificity, sensitivity, physical examination. Results: A total of 419 articles were found, from which 69 were selected based on their titles and abstracts. After quality assessment, five articles were chosen for extraction of data according to inclusion criteria; data were analyzed on a sample of 404,735 newborns in the five included studies. The following values were found, corresponding to the operational characteristics of oximetry in combination with the physical examination: sensitivity: 0.92 (CI 95%, 0.87-0.95), specificity: 0.98 (CI 95%, 0.89-1.00), for physical examination alone sensitivity: 0.53 (CI 95%, 0.28-0.78) and specificity: 0.99 (CI 95%, 0.97-1.00). Conclusions: Evidence found in different articles suggests that pulse oximetry in addition to neonatal physical examination presents optimal operative characteristics that make it an adequate screening test for detection of CCHD in newborns, above all this is essential in low and middle-income settings where technology medical support is not entirely available.",
"keywords": [
"Oximetry",
"screening",
"critical congenital heart disease",
"specificity",
"sensitivity",
"physical examination",
"newborn"
],
"content": "Introduction\n\nCongenital Heart Disease (CHD) in the prenatal stage can affect approximately 5 to 15 out of 1,000 live births1. A significant proportion of these cases will suffer from critical congenital heart disease (CCHD)2 requiring surgical treatment of intervention before the first year of life, as a late postnatal diagnosis is associated with a worse prognosis, a higher number of hospital admissions, long stays in hospital, and consequently, increased costs during childhood3.\n\nThe accuracy of cutaneous oximetry as a noninvasive measure to detect arterial oxygen saturation (SaO2) has been studied for more than two decades in neonatology, and its importance in the care of patients with respiratory and cardiovascular compromise has been recognized4. Evidence shows that the standardized, systematic measurement of SaO2 in newborns (NB) through transcutaneous oximetry may be a safe method with satisfactory operating characteristics in the detection of CCHD5–10. A systematic literature review (SLR) assessing the use of oximetry in the screening of CHD in NB showed that oximetry is a highly specific technology, with a very low level of false positive results, in the detection of congenital heart defects in NB5. Likewise, a meta-analysis showed similar findings6, highlighting the low rate of false-positive results in oximetry, primarily when the screening was conducted after the 24 hours following birth. It is worth highlighting that there are no recent studies of this type which include the evidence available on oximetry screening in the newborn for the detection of CCHDs.\n\nTogether with the use of oximetry, the role of the neonatal physical examination in the detection of CCHD has been studied7,8, revealing that NBs with CCHDs have been discharged without a timely diagnosis. This fact has been reported in the literature as relevant, suggesting that the neonatal physical examination fails to detect almost half of the cases of NB with CCHDs8. The above added to the low sensitivity and a high rate of false positives in the neonatal physical examination, has aroused more interest in including oximetry in CCHD screening, and it has been found that CCHD screening has a higher sensitivity when combining the neonatal physical examination and oximetry as compared with the individual use of any of these two methods9.\n\nEvidence has shown that CCHD screening through oximetry may have optimal operating characteristics10–12 that may allow for the identification of cases that would otherwise be impossible to detect; the above may allow as well for the implementation of national screening programs for early detection of CCHD, a step that would have an impact in the reduction of neonatal mortality and the costs associated with the assistance to complications deriving from late diagnosis3.\n\nThe purpose of this review was to define the operating characteristics of oximetry combined with physical examination in the detection of CCHD in NB younger than 37 weeks without suspicion or prior diagnosis of CCHDs.\n\n\nMethods\n\nThis review followed the criteria for reporting systematic literature reviews and meta-analysis as defined by the PRISMA strategy13. A completed PRISMA checklist is provided on OSF14.\n\nNewborns born after 37 weeks who underwent screening with oximetry, and who were analyzed for operative characteristics (sensitivity and specificity) were included. Studies on NB requiring neonatal intensive care or enduring infectious processes at birth were excluded.\n\nThe studies selected compared cutaneous oximetry screening and physical examination with physical examination alone.\n\nNewborns with CCHDs (undiagnosed in the prenatal stage), who are diagnosed early (at birth in the hospital) and not late (after hospital discharge at birth). CCHDs that can be diagnosed by screening include the following 12 CCHDs: interrupted aortic arch, coarctation of the aorta, dextro-transposition of the great arteries, double outlet right ventricle, Ebstein’s anomaly, hypoplastic left heart syndrome, pulmonary atresia, single ventricle, tetralogy of Fallot, total anomalous venous return, tricuspid atresia and truncus arteriosus.\n\nAn SLR was conducted on the following databases: Pubmed, Science Direct, Ovid, Scopus, and EBSCO, with the following keywords: oximetry screening, critical congenital heart disease, newborn OR oximetry screening, heart defects, congenital, specificity, sensitivity, physical examination. Cohort and case and control observational studies were included, as well as cross-sectional studies and prospective multicenter studies published between January 2002 and December 2017, with no language restrictions (Figure 1).\n\nThe selection and extraction of information, as well as the quality assessment of the articles, was performed independently by a revisor, considering the criteria of eligibility, and evaluating the bias risk, as well as the quality criteria adjusted to every type of study. Studies complying with more than 60% of the quality criteria were selected; the selection was made utilizing the STARD 2015 checklists15 for diagnostic test studies, and STROBE16 for observational studies. The methodological quality was also assessed employing the criteria included in the QUADAS-2 instrument17.\n\nThe following data related with the characteristics of the study and the outcomes of interest were selected from the studies selected: author, year, type of study, sample size, screening age, cut-off point, false positives, false negatives, positive predicting value, negative predicting value, sensitivity, and specificity.\n\nTo assess the bias risk, the following data were extracted: design of the study, blinding (in case it applies to the study of interest), losses to follow-up, outcome reports, contamination risk, and any other aspect affecting validity according to the type of the study.\n\nThe extraction was conducted for each one of the studies selected for meta-analysis, both for the physical examination and oximetry: sensitivity, specificity, predictive values, false positives, false negatives, true positives, and true negatives. In the cases where some of these data were not specified in the studies, an approximate calculation of the missing data was performed using RevMan 5.3 (Review Manager RevMan program calculator [Computer program] Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) and a meta-analysis of diagnostic tests was performed using the STATA software version 12 (StataCorp, College Station, Texas). Confidence intervals were calculated utilizing the Midas18 and metandi19 programs using STATA; likelihood ratios and heterogeneity values utilizing I2 statistic calculation20, and publication bias probability assessment through Deeks’ asymmetry test21, the most highly recommended current strategy to assess publication bias22\n\nA statistical heterogeneity assessment was conducted using the I2 statistic calculation, which describes the total variation percentage among the studies, which may be attributed to heterogeneity and not chance20. I2 may take values between 0% and 100%, where 0% is the absence of heterogeneity observed; values above 50% suggest substantial heterogeneity. The advantage of I2 calculation lies in the fact that it does not depend on the number of studies in the meta-analysis20.\n\n\nResults\n\nA total of 419 papers were identified as follows (Figure 1): 111 in Pubmed, 76 in EBSCO, 104 in Ovid, 118 in Science Direct and 10 in Scopus; 73 were duplicates, and 69 papers were selected for full-text revision by their title and abstract review. Out of these, 40 papers were excluded, as they did not include the sensitivity or specificity of the physical examination; 17 papers were excluded because they were narrative literature reviews; four papers were letters to the editor, and one paper included newborns with non-cardiac pathologies. After verifying compliance with the inclusion criteria, seven articles were included in the final analysis for quality assessment (Table 1 and Table 2).\n\nFor every article selected, the 12 CCHDs previously mentioned were considered as outcomes of interest. Following the quality assessment, a low risk of bias was observed in individual studies due to compliance with the four criteria described in five of the articles (Figure 2). The echocardiogram was applied as a pattern of reference in all reviews, and its results were assessed independently from the result of the oximetry screening in two of the papers23,24. These were considered, as the remaining reviews included echocardiogram only for newborns who showed alterations in the oximetry screening. It is important to mention that in most reviews no aspects were found that might hinder the applicability of the screening, both when selecting the population participating in the study and when performing the screening of the reference pattern. (Figure 3)\n\nOnce the application of the quality criteria was performed, seven articles were selected for data extraction; one article was excluded after conducting the quality assessment; one of the articles complying with quality criteria was excluded, as it did not include comparative data between the physical examination and oximetry. Finally, five articles were selected for the meta-analysis.\n\nFigure 4 shows the grouped values corresponding to the operative characteristics of the physical examination: sensitivity: 0.53 (95% confidence interval (95%CI), 0.28-0.78), specificity: 0.99 (95%CI, 0.97-1.00). Upon adding the use of oximetry (Figure 5) it was found that sensitivity increased: 0.92 (95%CI, 0.87-0.95) and specificity remained constant: 0.98 (95%CI, 0.89-1.00). For the physical examination, a positive diagnostic likelihood ratio (DLR) of 46.2 (95%CI, 15.2-140.2), and a negative DLR of 0.47 (95%CI, 0.26-0.85) were obtained. Regarding physical examination plus oximetry, a positive diagnostic likelihood ratio (positive DLR) of 43.7 (95%CI, 8.0-239.8), and a negative diagnostic likelihood ratio (negative DLR) of 0.08 (95%CI, 0.05-0.14) were found.\n\nFigure 6 and Figure 7 show the receptor operative characteristics (ROC) area under the curve (AUC) for physical examination, and physical examination plus oximetry. The AUC value found was 0.96 (95%CI, 0.94-0.97) for the physical exam, and 0.95 (95%CI, 0.93-0.97) when combining physical examination and oximetry, showing a similar diagnostic accuracy for both screening strategies. The following limits have been suggested to establish a diagnostic accuracy as per the AUC values as follows: low accuracy: AUC >0.5 and <0.7, moderate accuracy: AUC >0.7 and <0.9, and high accuracy: AUC >0.9 and ≤125.\n\nAn I2 of 94.91 (95%CI, 91.9-97.91) was found for the sensitivity of the physical examination, and an I2 value of 82.42 (95%CI, 69.23-95.61) was found for the sensitivity of the physical examination in combination with oximetry. Likewise, the heterogeneity proportion, probably due to the threshold effect, was high (1.00), indicating the presence of a diagnostic threshold effect on the performance of the physical examination and oximetry screening.\n\nPublication risk of bias was assessed through the Deeks’ regression test21. Upon performing this evaluation, a non-statistically significant value was found for the coefficient corresponding to the slope (p = 0.89), which suggests symmetry in the data, and hence, a low probability of publication bias (Figure 8).\n\nFigure 9 shows an example of the results that would be obtained after the application of the physical examination on a hypothetical cohort of 10,000 live asymptomatic NBs and an expected prevalence of 17 cases of CCHD per 10,000 NBs2,26. Figure 10 shows the results that would be obtained in the same cohort when adding the use of oximetry to physical examination, it can be seen how the number of cases diagnosed in hospital increases almost twofold, and in a proportionate number, the number of false positive results increase.\n\n\nDiscussion\n\nCCHD represents a considerable cause of morbidity and mortality in newborns. Their early diagnosis has become an essential objective at the time of implementing a screening strategy involving oximetry, an aspect of the utmost importance, as late detection is associated with neurological and cognitive sequels in the NB, in addition to cardiac sequels, and social and economic repercussions.\n\nThe medical literature recommends the detection of CHD classified as \"life-threatening\" due to the risk of collapse and long-term sequels in the development of the NB27. Among the CCHD that can be diagnosed through oximetry screening we find the following: interrupted aortic arch, coarctation of the aorta, dextro-transposition of the great arteries, double outlet right ventricle, Ebstein’s anomaly, hypoplastic left heart syndrome, pulmonary atresia, single ventricle, tetralogy of Fallot, total anomalous venous return, tricuspid atresia and truncus arteriosus.\n\nThanks to the inclusion of recent studies, this review tends to validate the importance of oximetry as a screening tool for this type of cardiopathies through the definition of their main operative characteristics. The data described in this study suggest that the physical examination in isolation does not offer reasonable levels of sensitivity for the diagnosis of CCHD in NBs, a feature observed in most of the studies. However, the specificity of the physical examination was high, a fact that was clinically expected. According to our data, when complementing the physical examination with oximetry, the sensitivity of the screening process is notably higher. The likelihood values described also tend to favor the complementary use of oximetry.\n\nThe degree of heterogeneity observed among the different study estimates is statistically significant; this variability might be explained as a result of the different cut-off points of \"threshold effect\", a finding that has been described as one of the most frequent primary causes of heterogeneity in meta-analyses of diagnostic tests. It occurs when differences in sensitivity and specificity are the result of there being different cut-off points or thresholds22. As previously explained, the proportion of heterogeneity in this study is probably due to the threshold effect.\n\nAnother cause for the heterogeneity found in our study may be attributed to the different sample sizes found, which may also condition the presence of variability between studies; although this variation is meaningful from the statistic perspective, its clinical importance regarding diagnostic performance of physical examination and oximetry is objectionable, due to the operative characteristics already described and its diagnostic accuracy in the detection of CCHDs.\n\nLikewise, the low sensitivity of physical examination found in most of the included articles may also influence the global estimate for heterogeneity. Among the causes for these low sensitivity values, the one provided by Saxena et al.23 stands out, as it reports lower sensitivity values as the severity of the heart disease increases; they also report technical and human types of errors when conducting the screening, which might also have an effect on these low sensitivity figures.\n\nRegarding the clinical usefulness of including oximetry in CCHD screening, we may conclude that it contributes to the early diagnosis of few cases, thus reducing the number of false negative results of the physical examination. According to Figure 9, it may be observed that 100 out of 109 NB with a positive screening would correspond to false positives and 9 would be true positives; out of the 9,891 NBs with a negative screening, eight would be false negatives with a form of CCHD, but they would not be detected by routine physical examination alone.\n\nFurthermore, when adding the use of oximetry (Figure 10), the number of cases diagnosed in hospital increases almost twofold, and in a proportionate number, the number of false positive results increase. Similarly, oximetry screening in addition to the physical examination reduces the number of NB with false-negative results (Figure 10), going from 8 NBs with CCHDs that would not have been detected by the physical examination alone, to 1 NB when using physical examination in combination with oximetry. Although the number of false positives increases when oximetry is added, we consider this aspect as minor as compared with the reduction in the number of false negatives resulting from the application of screening, which would be reflected on better survival and in a reduction of the associated costs derived from additional medical and surgical interventions arising from a late diagnosis9.\n\nThe above data are consistent with what previous studies have reported31, whereby oximetry as an additional tool to the physical examination provides a timely diagnosis for almost 30 additional cases of CHD per 100,000 live NBs as compared to the use of the physical examination alone. At the same time, oximetry is considered a potentially efficient and appropriate tool to identify cases of CCHD that would otherwise go undetected after the physical examination of the NB31,32.\n\nThis review also confirms the finding described in a prospective multicenter study24, but it is worth highlighting that the same study reports a global rate of false positives of less than 1% in the detection of CCHD; however, in this study the rate of false positives was affected by the time at which the oximetry was conducted, as it was significantly lower when the screening was conducted 24 hours after birth, as compared to those conducted before the 24 hours.\n\nAmong the strengths of this review, we may highlight the rigorous search conducted on the recent literature, and the standardized quality assessment performed on the articles included. Among the limitations we may include the small number of studies assessing the use of oximetry together with physical examination as a screening strategy for the detection of CCHD, and also the fact that one single evaluator conducted the assessment for selection, quality measurement, and data extraction, although formats previously standardized for this process were used.\n\n\nConclusions\n\nFrom this SLR and meta-analysis, it may be concluded that the use of oximetry, added to the conventional physical examination helps to detect a more significant number of NBs with CCHDs, without significantly increasing the number of false-positive results, a finding that may reduce the morbidity and mortality associated with hospital discharge of NB without a timely diagnosis33.\n\nThis review also provides updated information which sets the bases for determining whether the impact of including this noninvasive technology as part of NB screening is cost-effective in low- and middle-income countries.\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\nReporting guidelines\n\nA completed PRISMA checklist is available from Open Science Framework: PRISMA checklist. DOI: https://doi.org/10.17605/OSF.IO/9SNCQ14.",
"appendix": "Grant information\n\nThis study was supported by the Fondo Nacional de Financiamiento para la Ciencia, la Tecnología y la Innovación, Francisco José de Caldas – COLCIENCIAS, Programa para la Innovación en Cardiopatías Congénitas Humanas Infrecuentes para Colombia, PINOCCHIO (Program for Innovation in Rare Congenital Heart Diseases in Humans in Colombia) – Contract number: 662-2015.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe wish to thank the PINOCCHIO (Program for the innovation of rare human congenital cardiopathies for Colombia) team at Fundacion Cardioinfantil- Instituto de Cardiologia, as well as the Public Health Shaft from Fundación Santa Fe de Bogotá for all the support provided to the development of this review. We wish to thank Martin Rondon, MSc, Professor at the Department of Clinical Epidemiology and Biostatistics at the School of Medicine at Pontificia Universidad Javeriana for his support and guidance in statistical analysis, and Juan Carlos Villar, MD, MSc, PhD in the Research Department at Fundacion Cardioinfantil- Instituto de Cardiologia for his contribution and feedback in the final stages of manuscript development.\n\n\nReferences\n\nLiu S, Joseph KS, Lisonkova S, et al.: Association between maternal chronic conditions and congenital heart defects: A population-based cohort study. Circulation. 2013; 128(6): 583–589. PubMed Abstract | Publisher Full Text\n\nOster ME, Lee KA, Honein MA, et al.: Temporal Trends in Survival Among Infants With Critical Congenital Heart Defects. Pediatrics. 2013; 131(5): e1502–e1508. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPeterson C, Dawson A, Grosse SD, et al.: Hospitalizations, costs, and mortality among infants with critical congenital heart disease: How important is timely detection? Birth Defects Res A Clin Mol Teratol. 2013; 97(10): 664–672. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBowes WA 3rd, Corke BC, Hulka J: Pulse oximetry: a review of the theory, accuracy, and clinical applications. Obs Gynecol. 1989; 74(3 Pt 2): 541–546. PubMed Abstract\n\nThangaratinam S, Daniels J, Ewer AK, et al.: Accuracy of pulse oximetry in screening for congenital heart disease in asymptomatic newborns: A systematic review. Arch Dis Child Fetal Neonatal Ed. 2007; 92(3): 176–180. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThangaratinam S, Brown K, Zamora J, et al.: Pulse oximetry screening for critical congenital heart defects in asymptomatic newborn babies: A systematic review and meta-analysis. Lancet. 2012; 379(9835): 2459–2464. PubMed Abstract | Publisher Full Text\n\nWren C, Richmond S, Donaldson L: Presentation of congenital heart disease in infancy: implications for routine examination. Arch Dis Child Fetal Neonatal Ed. 1999; 80(1): F49–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChang RK, Gurvitz M, Rodriguez S: Missed Diagnosis of Critical Congenital Heart Disease. Arch Pediatr Adolesc Med. 2008; 162(10): 969. PubMed Abstract | Publisher Full Text\n\nde-Wahl Granelli A, Wennergren M, Sandberg K, et al.: Impact of pulse oximetry screening on the detection of duct dependent congenital heart disease: a Swedish prospective screening study in 39 821 newborns. BMJ. 2009; 338: a3037. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMahle WT, Newburger JW, Matherne GP, et al.: Role of Pulse Oximetry in Examining Newborns for Congenital Heart Disease: A Scientific Statement from the AHA and AAP. Pediatrics. 2009; 124(2): 823–836. PubMed Abstract | Publisher Full Text\n\nKemper AR, Mahle WT, Martin GR, et al.: Strategies for Implementing Screening for Critical Congenital Heart Disease. Pediatrics. 2011; 128(5): e1259–e1267. PubMed Abstract | Publisher Full Text\n\nEwer AK, Furmston AT, Middleton LJ, et al.: Pulse oximetry as a screening test for congenital heart defects in newborn infants: a test accuracy study with evaluation of acceptability and cost-effectiveness. Health Technol Assess (Rockv). 2012; 16(2): v–xiii. PubMed Abstract | Publisher Full Text\n\nLiberati A, Altman DG, Tetzlaff J, et al.: The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: Explanation and elaboration. PLoS Med. 2009; 62(7): e1–34. PubMed Abstract | Publisher Full Text\n\nAranguren Bello HC, Trujillo DL, Troncoso GA, et al.: PRISMA Checklist. OSF. 2019; Accessed January 25, 2019. http://www.doi.org/10.17605/OSF.IO/9SNCQ\n\nCohen JF, Korevaar DA, Altman DG, et al.: STARD 2015 guidelines for reporting diagnostic accuracy studies: Explanation and elaboration. BMJ Open. 2016; 6(11): e012799. PubMed Abstract | Publisher Full Text | Free Full Text\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. Epidemiology. 2007; 18(6): 800–804. PubMed Abstract | Publisher Full Text\n\nWhiting PF, Rutjes AWS, Westwood ME, et al.: QUADAS-2: A Revised Tool for the Quality Assessment of Diagnostic Accuracy Studies. Ann Intern Med. 2011; 155(4): 529–536. PubMed Abstract | Publisher Full Text\n\nDwamena B: MIDAS: Stata module for meta-analytical integration of diagnostic test accuracy studies. Stat Softw Components. 2009. Reference Source\n\nHarbord RM, Whiting P: metandi: Meta-analysis of diagnostic accuracy using hierarchical logistic regression. 2009. Publisher Full Text\n\nHiggins JP, Thompson SG: Quantifying heterogeneity in a meta-analysis. Stat Med. 2002; 21(11): 1539–1558. PubMed Abstract | Publisher Full Text\n\nDeeks JJ, Macaskill P, Irwig L: The performance of tests of publication bias and other sample size effects in systematic reviews of diagnostic test accuracy was assessed. J Clin Epidemiol. 2005; 58(9): 882–893. PubMed Abstract | Publisher Full Text\n\nCronin P, Kelly AM, Altaee D, et al.: How to Perform a Systematic Review and Meta-analysis of Diagnostic Imaging Studies. Acad Radiol. 2018; 25(5): 573–593. PubMed Abstract | Publisher Full Text\n\nSaxena A, Mehta A, Ramakrishnan S, et al.: Pulse oximetry as a screening tool for detecting major congenital heart defects in Indian newborns. Arch Dis Child Fetal Neonatal Ed. 2015; 100(5): F416–F421. PubMed Abstract | Publisher Full Text\n\nZhao QM, Ma XJ, Ge XL, et al.: Pulse oximetry with clinical assessment to screen for congenital heart disease in neonates in China: a prospective study. Lancet. 2014; 384(9945): 747–754. PubMed Abstract | Publisher Full Text\n\nSwets JA: Measuring the accuracy of diagnostic systems. Science. 1988; 240(4857): 1285–1293. PubMed Abstract | Publisher Full Text\n\nDolk H, Loane M, Garne E: Congenital heart defects in Europe: prevalence and perinatal mortality, 2000 to 2005. Circulation. 2011; 123(8): 841–849. PubMed Abstract | Publisher Full Text\n\nGriebsch I, Knowles RL, Brown J, et al.: Comparing the clinical and economic effects of clinical examination, pulse oximetry, and echocardiography in newborn screening for congenital heart defects: a probabilistic cost-effectiveness model and value of information analysis. Int J Technol Assess Health Care. 2007; 23(2): 192–204. PubMed Abstract | Publisher Full Text\n\nHu XJ, Ma XJ, Zhao QM, et al.: Pulse Oximetry and Auscultation for Congenital Heart Disease Detection. Pediatrics. 2017; 140(4): pii: e20171154. PubMed Abstract | Publisher Full Text\n\nMeberg A, Brügmann-Pieper S, Due R Jr, et al.: First day of life pulse oximetry screening to detect congenital heart defects. J Pediatr. 2008; 152(6): 761–765. PubMed Abstract | Publisher Full Text\n\nOakley JL, Soni NB, Wilson D, et al.: Effectiveness of pulse-oximetry in addition to routine neonatal examination in detection of congenital heart disease in asymptomatic newborns. J Matern Fetal Neonatal Med. 2015; 28(14): 1736–1739. PubMed Abstract | Publisher Full Text\n\nRoberts TE, Barton PM, Auguste PE, et al.: Pulse oximetry as a screening test for congenital heart defects in newborn infants: a cost-effectiveness analysis. Arch Dis Child. 2012; 97(3): 221–226. PubMed Abstract | Publisher Full Text\n\nPeterson C, Grosse SD, Oster ME, et al.: Cost-effectiveness of routine screening for critical congenital heart disease in US newborns. Pediatrics. 2013; 132(3): e595–e603. PubMed Abstract | Publisher Full Text | Free Full Text\n\nValmari P: Should pulse oximetry be used to screen for congenital heart disease?. Arch Dis Child Fetal Neonatal Ed. 2007; 92(3): F219–224. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "48727",
"date": "07 Jun 2019",
"name": "Andrew K Ewer",
"expertise": [
"Reviewer Expertise Clinical Neonatology"
],
"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\nA statistician will need to review the methodology and results. I would remove the term 'in the prenatal stage' from the abstract, introduction and throughout. Include the recent Cochrane review on pulse oximetry screening.1 I do not understand the phrase ‘It is worth highlighting that there are no recent studies of this type which include the evidence available on oximetry screening in the newborn for the detection of CCHDs. Together with the use of oximetry,’. Please explain. P3: ‘in NB younger than 37 weeks’. Do you mean greater than 37 weeks? The PISMA Flow chart does not make sense as currently presented. Need to look again at the numbers coming of at each stage. ‘The echocardiogram was applied as a pattern of reference’ – This should be changed to gold standard test. In 'Clinical impact', include numbers identified in both text and figure.\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? I cannot comment. A qualified statistician is required.\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": []
},
{
"id": "51065",
"date": "24 Jul 2019",
"name": "Praveen Kumar",
"expertise": [
"Reviewer Expertise Clinical Neonatology"
],
"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\nBello et al. completed a systematic review and meta-analysis to evaluate if pulse oximetry screening improves the diagnostic accuracy over physical examination alone for early detection of CCHD. Authors conclude that pulse oximetry in addition to neonatal physical examination improves sensitivity and may be particularly helpful in early detection of CCHD in low-resource settings.\nAuthors have used appropriate search and statistical tools to complete their review. The manuscript is well written and easy to follow.\nPlease correct the typing error at the end of introduction section - authors state their objective as \"to define ..... CCHD in NB younger than 37 weeks without suspicion or prior diagnosis of CCHDs\" but the review is based on data collected on term infants.\nAuthors provide no data to support their conclusion that pulse oximetry is likely to be more helpful in early detection of CCHD in low-resource settings.\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/8-242
|
https://f1000research.com/articles/8-238/v1
|
01 Mar 19
|
{
"type": "Opinion Article",
"title": "Reinventing the antimicrobial pipeline in response to the global crisis of antimicrobial-resistant infections",
"authors": [
"Andrew C. Singer",
"Claas Kirchhelle",
"Adam P. Roberts",
"Claas Kirchhelle",
"Adam P. Roberts"
],
"abstract": "The pipeline for new antibiotics is dry. Despite the creation of public/private initiatives like Combating Antibiotic Resistant Bacteria Biopharmaceutical Accelerator (Carb-X) and the Antimicrobial Resistance (AMR) Centre, the current focus on ‘push-pull’ incentives for the pharmaceutical industry still relies on economic return. We propose a joint, internationally-funded antimicrobial development institute that would fund permanent staff to take on roles previously assigned to pharmaceutical companies. This institute would receive ring-fenced, long-term, core funding from participating countries as well as charities, with the aim to focus on transforming the largely dormant antimicrobial pipeline. Resulting drugs would be sold globally and according to a principle of shared burdens. Our proposed model for antimicrobial development aims to maximise society’s investment, through open science, investment in people, and the sharing of intellectual property.",
"keywords": [
"antibiotics",
"antibiotic resistance",
"antimicrobial",
"drug pipeline"
],
"content": "\n\nThe UK’s new five-year national antimicrobial resistance (AMR) action plan highlights that society is at a tipping point: not only are we running out of effective antibiotics, but the pipeline for new drugs is dry and novel diagnostics are slow to come to market (HM Government, 2019). Over the past two decades, decision-makers have tried to overcome this dry spell by relying on the market and creating ‘push-pull’ incentives for the pharmaceutical industry (Renwick et al., 2016). The fundamental premise of ‘push-pull’ incentives, is to make it economically attractive for major pharmaceutical companies to invest their infrastructure, staff and skills into the research, development and manufacture of novel antibiotics. The reality for pharmaceutical companies is that there is too little money to be made with antibiotics: any new antibiotic will be reserved as a ‘last-line defence’, which means it will (hopefully) be an infrequently used antibiotic and as such, a poor source of income.\n\nWhile recent public-private initiatives like Combating Antibiotic Resistant Bacteria Biopharmaceutical Accelerator (Carb-X) and the AMR Centre (UK), are important steps in the right direction, we strongly contend that using public funds to support private venture and profit is not the only way to refill the antibiotic pipeline. In the current model for drug development, the tax payer pays for everything up to the point where the pharmaceutical companies may invest in later stages of development (Galkina Cleary et al., 2018). Society invests in the students, post-docs, and professors through public funding of the university and/or the research grant that pays their salaries and equipment. However, at the point where knowledge becomes patentable, it quickly disappears down pharmaceutical pipelines.\n\nCommercial pipelines are not always efficient producers of new drugs. Potentially promising drugs are abandoned when commercial or global health priorities change. In some cases, patents are only filed to deter competitors from developing them further. Not only are society’s investments in antimicrobial innovation wasted, but the intellectual property associated with a drug’s development, i.e., the countless serendipitous leads as well as dropped, yet promising, antimicrobials, remain locked-up within the private pharmaceutical company—inaccessible to the public who invested in the initial development. Ultimately, the current mode of subsidised antibiotic development means that society pays for 100% of all the drugs that are developed and not developed, but, importantly, owns and controls nothing (discussed here (Anon, 2019)). It is this obvious disconnect in societal investment, which needs to be reformed.\n\nWe propose a radical change to the current paradigm of antimicrobial development and manufacture—one that reflects humanity’s shared interests and global health challenges. We propose a joint, internationally-funded antimicrobial development Institute that would fund permanent staff to take on the role previously assigned to pharmaceutical companies. This institute would receive ring-fenced, long-term, core funding from participating countries as well as charities, with the aim to focus on transforming the largely dormant antimicrobial pipeline.\n\nThe international centre would aim to sustainably develop a breadth of new antimicrobials to address both immediate and emerging global health challenges, respond to needs across organisms, i.e., viral, bacteria, fungal, protist, and develop novel modes of action. Prospective antimicrobials submitted to the centre would be developed in an open and transparent manner, so that innovation can be immediately shared and serendipitous findings can be leveraged by the wider research community. The centre would also conduct clinical trials on prospective drug candidates, manufacture all antimicrobials through existing generic drug manufacturers, and contract research organisations.\n\nResulting drugs would be sold globally and according to a principle of shared burdens. This would mean that high-income countries would pay more for research and the drugs themselves than low-income countries facing ongoing access problems. Signing up to enforceable stewardship requirements would be a precondition to receiving new drugs. Since no development costs need recouping and no share holder incentives need to be satisfied, most drugs could be sold at the cost of manufacture. Nearly 40 years of lacking commercial interest in new antibiotic development means that our ‘not-for-profit’ antibiotic pipeline would not compete with established manufacturers.\n\nA look back at the 20th century shows that our proposed approach is not as far-fetched as it may sound. While a turn to more profitable ‘lifestyle’ drugs ended pharmaceutical investment in antibiotic development from the late 1970s onwards (Gradmann, 2016), the 1940s saw many of the same companies work hand-in-hand with nation states and universities to screen antimicrobial compounds, test them in clinical settings, and upscale production. There was no patent on penicillin – only on the process developed to mass-produce it (Bud, 2009). In the case of antimalarials, military interests led to a long history of state-directed and subsidized development (Lezaun, 2018).\n\nOur proposed model for antimicrobial development aims to maximise society’s investment, through open science, investment in people, and the sharing of intellectual property. Antimicrobial resistant infections pose a global challenge. It is time to realise that the challenge of solving the global problem of AMR exceeds the capacity of commercial actors. We are already financing the development of antibiotics, why not collectively own and manage the resulting drugs?\n\n\nData availability\n\nNo data are associated with this article",
"appendix": "Grant information\n\nThis work was funded by the Natural Environment Research Council [NE/N019687/1].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nAnon: Examining the Actions of Drug Companies in Raising Prescription Drug Prices. House Committee on Oversight and Reform [Online]. 2019; [Accessed: 12 February 2019]. Reference Source\n\nBud R: Penicillin: triumph and tragedy. Oxford: Oxford University Press, 2009. Reference Source\n\nGalkina Cleary E, Beierlein JM, Khanuja NS, et al.: Contribution of NIH funding to new drug approvals 2010-2016. Proc Natl Acad Sci U S A. 2018; 115(10): 2329–2334. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGradmann C: Re-Inventing Infectious Disease: Antibiotic Resistance and Drug Development at the Bayer Company 1945-80. Med Hist. 2016; 60(2): 155–180. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHM Government: Tackling antimicrobial resistance 2019-2024: The UK’s five-year national action plan. HM Government. 2019. Reference Source\n\nLezaun J: The deferred promise of radical cure: pharmaceutical conjugations of malaria in the global health era. Econ Soc. 2018; 47(4): 547–571. Publisher Full Text\n\nRenwick MJ, Brogan DM, Mossialos E: A systematic review and critical assessment of incentive strategies for discovery and development of novel antibiotics. J Antibiot (Tokyo). 2016; 69(2): 73–88. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "45186",
"date": "13 Mar 2019",
"name": "Christoph Gradmann",
"expertise": [
"Reviewer Expertise History of 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 authors have written a bold, interesting and well argued opinion piece. It discusses alternatives to reviving antibiotics drug development beyond the ‘push-pull’ incentives offered to pharmaceutical industry. They would essentially result in making novel antibiotics a public commodity. This should trigger much needed debate. I have two comments. One is in detail. In the opening of the third last paragraph we find the sentence \"Resulting drugs would be sold globally ...\". Is that consistent with the approach. After all selling could again make the invented drugs private property and prohibit keeping their further development in open areas. Would not licensing be better?\nMy second question is about comparable stories. I am no great expert in vaccines but it seems to me that in this case we have seen rather successful cases of public-private partnerships in the last decades. Maybe worth considering. Historical reading could start from \"Galambos, Louis, and Jane Eliot Sewell. Networks of Innovation: Vaccine Development at Merck, Sharp and Dohme, and Mulford, 1895-1995. Cambridge: Cambridge University Press, 1996.\"\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": "45713",
"date": "19 Mar 2019",
"name": "Dean Baker",
"expertise": [
"Reviewer Expertise I am economist that has been researching drug development for two decades."
],
"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 proposes to establish an international center for antimicrobial research and development. The center would develop and then sell drugs to according to a principle of \"shared burden.\" There is a need for such a center, since the private pharmaceutical industry has largely abandoned research in this area since it sees little prospect for profit.\n\nWhile this is an innovate idea, it would be helpful to see it flushed out a bit more in two directions. First, while it seems the authors' intentions that all research findings be open, a slightly clearer statement on this principle would be helpful. For example, do they envision something like the Bermuda Principle, where results are posted to the web, if not nightly, as quickly as practical.\nThe second area that could benefit from clarification is the meaning of the principle of \"shared burden.\" Since the research costs are being paid upfront, it is unclear why there would be any reason to charge a price above the marginal cost. (There could an issue where countries that did not contribute to research costs get charged more, but that should be explicitly stated.)\nAnyhow, this is a worthwhile proposal which should advance the debate on the best way to finance the development of prescription drugs.\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": []
}
] | 1
|
https://f1000research.com/articles/8-238
|
https://f1000research.com/articles/8-230/v1
|
28 Feb 19
|
{
"type": "Software Tool Article",
"title": "shiny-pred: a server for the prediction of protein disordered regions",
"authors": [
"Mauricio Oberti",
"Iosif Vaisman",
"Iosif Vaisman"
],
"abstract": "Intrinsically disordered proteins or intrinsically disordered regions (IDR) are segments within a protein chain lacking a stable three-dimensional structure under normal physiological conditions. Accurate prediction of IDRs is challenging due to their genome wide occurrence and low ratio of disordered residues, making them a difficult target for traditional classification techniques. Existing computational methods mostly rely on sequence profiles to improve accuracy, which is time consuming and computationally expensive. The shiny-pred application is an ab initio sequence-only disorder predictor implemented in R/Shiny language. In order to make predictions, it uses convolutional neural network models, trained using PDB sequence data. It can be installed on any operating system on which R can be installed and run locally. A public version of the web application can be accessed at https://gmu-binf.shinyapps.io/shiny-pred",
"keywords": [
"Disordered proteins",
"machine learning",
"convolutional neural networks",
"R",
"Shiny"
],
"content": "Introduction\n\nExperimental structure resolution of intrinsically disordered proteins/intrinsically disordered regions (IDP/IDRs) is complex, lengthy and expensive, leading to a variety of computational approaches being developed (He et al., 2009). Over 60 computational protein disorder prediction servers are currently available, although not all publicly. Methods can be classified in one of the following categories (Atkins et al., 2015): (i) Ab initio or sequence-based, (ii) clustering, (iii) template based, and (iv) meta or consensus.\n\nshiny-pred is an ab initio predictor, which means it relies exclusively on amino acid sequence information to make disordered predictions. It uses prediction models based on convolutional neural networks and reduced protein alphabets. Currently there are three available models, each one built using the same training protein data from PDB (Berman et al., 2000) but differing on the convolutional neural network architecture. Since it doesn't rely in sequence profiles to make predictions, it is fast to be used in proteome-wide disorder scenarios. It performs at the same level or outperforms other state of the art sequence-only methods, achieving accuracy levels of 0.76 and AUC of 0.85 on the publicly available CASP10 dataset (Monastyrskyy et al., 2014), at faster speeds.\n\n\nMethods\n\nshiny-pred is written in the R programming language (R Core Team, 2017) and the shiny web application framework is implemented using the Shiny R package v1.1.0 (Chang, 2018).\n\nCurrently, three convolutional neural network models are made available by our application:\n\n(i) cnn-64-ker-local, is a one layer convolutional network (step size 1 and window size of 32) with 64 kernels and local max pooling model; (ii) cnn-128-ker-local, implements one convolutional layer (step size 1 and window size of 32) with 128 kernels and local max pooling model; and (iii) cnn-2-conv-local implements two convolutional layers (64 and 32 kernels) with local max pooling.\n\nThe models were created, trained and accessed using the keras R package v2.1.6 (Allaire & Chollet, 2018).\n\nOur tool has two operation modes; predicting disordered residues in protein sequences (prediction) and benchmarking the predictor performance against sequences with known disorder information (benchmark). The mode is selected automatically based on the format of the input sequences. Users can either upload a sequence file, type/paste a sequence into the text area or select pre-loaded examples from a list.\n\nWhen in prediction mode, the amino acid sequences are expected to be in FASTA format (Figure 1). In benchmark mode, input sequences in FASTA format are expected to have an additional line containing the disorder information (D=disorder, O=ordered). Multiple sequences can be submitted at once; several examples for different types of submissions (prediction and benchmark modes) are made available as examples. In both modes, the application will show a result panel, where for each input sequence a graph with the probability of disorder per residue is plotted (Figure 2).\n\n(1) Prediction mode\n\nThe workflow for protein disorder prediction is:\n\n(i) Input the target sequences (in FASTA format) in the text area;\n\n(ii) Select the model to use for the prediction (default is cnn-128-ker-local) and submit the sequence for prediction;\n\n(iii) Visualize and download results.\n\n(2) Benchmark mode\n\nIn benchmark mode, input sequences are expected to have an extra line with the actual disorder information to be used as benchmark. Result tables will populate two extra columns (actual class and match) with the actual disorder information and if the prediction was accurate for the current residue. An extra panel (Benchmark) shows the ROC curve along with other common binary metrics (sensitivity, specificity, balance accuracy and Matthews correlation coefficient).\n\n\nUse cases\n\nWe use shiny-pred to predict disordered regions within the publicly available CASP10 benchmark dataset. The dataset contains 94 target sequences, each one annotated with the disorder/order information at the residue level. The annotated dataset is provided as an example (‘CASP_all’) and it can be selected form the example selection list on the ‘Sequence Input’ tab. Figure 3 shows the input panel after the dataset is selected and loaded. Predictions per sequence can be viewed and downloaded from the ‘Results’ tab while the ‘Benchmark’ tab provides a summary of the performance using binary and statistical metrics. Figure 4 shows the server performance for the input dataset, achieving an AUC value of 0.85 and balance accuracy of 0.75.\n\n\nSummary\n\nThis article presents shiny-pred, a sequence-only ab initio web application for predicting protein disorder. It's based on reduced amino acid alphabets and convolutional neural networks, being fast and accurate, it is suitable for large proteome-wide experiments.\n\n\nSoftware availability\n\nSoftware available from: https://gmu-binf.shinyapps.io/shiny-pred\n\nSource code available from: https://github.com/mauricioob/shiny-pred\n\nArchived source code as at time of publication: https://doi.org/10.5281/zenodo.2567259 (Mauricio, 2019).\n\nLicense: GNU public license (GPL-3)",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgments\n\nThe authors are grateful for the computational facilities provided by Novartis Institutes of Biomedical Research.\n\n\nReferences\n\nAtkins JD, Boateng SY, Sorensen T, et al.: Disorder Prediction Methods, Their Applicability to Different Protein Targets and Their Usefulness for Guiding Experimental Studies. Int J Mol Sci. 2015; 16(8): 19040–19054. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBerman HM, Westbrook J, Feng Z, et al.: The Protein Data Bank. Nucleic Acids Res. 2000; 28(1): 235–242. PubMed Abstract | Free Full Text\n\nHe B, Wang K, Liu Y, et al.: Predicting intrinsic disorder in proteins: an overview. Cell Res. 2009; 19(8): 929–949. PubMed Abstract | Publisher Full Text\n\nAllaire JJ, Chollet F: keras: R Interface to “Keras”. [Accessed: 15 January 2019]. 2018. Reference Source\n\nMauricio: mauricioob/shiny-pred: Initial release (Version v1.0). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.2567259\n\nMonastyrskyy B, Kryshtafovych A, Moult J, et al.: Assessment of protein disorder region predictions in CASP10. Proteins. 2014; 82 Suppl 2: 127–137. PubMed Abstract | Publisher Full Text | Free Full Text\n\nR Core Team: R: A Language and Environment for Statistical Computing. [Accessed: 13 January 2019]. 2017. Reference Source\n\nChang W: shiny: Web Application Framework for R. [Accessed: 13 January 2019]. 2018. Reference Source"
}
|
[
{
"id": "46176",
"date": "11 Apr 2019",
"name": "Jinbo Xu",
"expertise": [
"Reviewer Expertise Computational biology",
"machine learning",
"optimization."
],
"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 a new protein disorder prediction web server that makes use of (shallow) convolutional neural networks. There are already many disorder predictors, some of which are based upon deep convolutional neural network and can do prediction directly on amino acid sequence instead of sequence profile. This manuscript does not have sufficient results to justify why one more web server for disorder prediction is needed. Here are some concerns:\nPlease compare with existing, similar methods. It is better to test the method on more recent CASP datasets and make sure that there is no redundancy between training and test data. Ideally, a much larger test set shall be used to evaluate the method. AUC may not be a good metric for disorder prediction since the ratio of disordered residues is quite small. Precision and Recall may be better. Existing work shall be cited.\n\nIs the rationale for developing the new software tool clearly explained? Partly\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? Partly",
"responses": []
},
{
"id": "50204",
"date": "08 Jul 2019",
"name": "Appadurai Rajeswari",
"expertise": [
"Reviewer Expertise Membrane Biophysics",
"Protein Structures and Folding",
"Mechanotransduction",
"Statistical mechanics of Biological Systems",
"Integrative Modeling",
"Multiscale Biomolecular Simulations"
],
"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 presented yet another neural network-based disorder prediction tool written in R, trained on PDB data and benchmarked on CASP10 dataset and they claim that the tool outperforms other existing tools in terms of both calculation speed and performance.\n\nWe tried using the tool for predicting the known disordered sequences and found that the predictions are accurate and similar to other tools such as IUPRED, DISOPRED3 servers for the well-known disordered sequences such as p53 and Histatin5.\n\nIn terms of concerns, I have following comments to make:\nIn general, I find the paper does not describe the motivation, methods and the results in a self-sufficient manner and these could be elaborated further.\n\nAs the authors state in the paper, there are over 60 tools already existing for disorder prediction.The justification for requiring another tool is not clearly stated.\n\nThe authors mention they have used PDB data for training the neural network. Do they take all the currently available PDB datasets for training? Does any overlap exist between the datasets trained and benchmarked? The reason why I am asking this is the CASP10 dataset that the authors used for benchmarking has been released in 2012, which would be a subset of the training PDB dataset if they have taken all the PDB data published till date.\n\nThe authors claim that their method is faster than the existing methods. It would be nice to provide evidence towards that and provide some benchmarking data.\n\nAUC and balance accuracy are the two metrics used for evaluating the performance of the tool. However, a clear definition of these terms are not described in the method section.\n\nThe tool should be tested and bench marked against a larger data set such as Disorder-723, which contains 723 disorder sequences.\n\nIs the rationale for developing the new software tool clearly explained? No\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? No\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-230
|
https://f1000research.com/articles/8-223/v1
|
27 Feb 19
|
{
"type": "Research Note",
"title": "Incidence of the thermal transition in the range of 45-5°C in chromatophores present in the wings of Schistocerca americana",
"authors": [
"María Belén Cañizares",
"Nathaly Naranjo",
"Bence Mátyás",
"María Belén Cañizares",
"Nathaly Naranjo"
],
"abstract": "The variation of the color intensity of the chromatophores present in the wings of Schistocerca americana was analyzed by exposing 31 specimens to thermal transitions within the range of 45 - 5 °C. The adult specimens were collected using a mini-terrarium of dimensions 40x40x30 cm. As a substrate, a layer of soil, stones, and finally a layer of grass were used along with branches of bushes and leaves; hydroponic lettuce, cabbage and the grass were used as food for the specimens. Optical microscopy of the wings of the insects was used for live observation without coverslips or contrasting substances. At 45°C, degradation of color intensity was observed in the chromatophores present in the wings. At 5°C, chromatophores intensify their color to brownish-black. This temperature was the extreme minimum that S. americana could tolerate. We found negative correlation between the temperature and the degree of darkness (R2 = 0.8038). Our results are in accordance with a previously published study in which Phaulacridium vittatum was examined, as the decrease of temperature caused darkening color change in melanin-type chromatophores. The present investigation can be considered as the first initial study of its kind for S. americana, in terms of examining the changes in the color intensity of the chromatophores present in the wings caused by thermal transition under laboratory conditions.",
"keywords": [
"Schistocerca americana",
"Thermal transition",
"Chromatophores",
"Melanism",
"Pigmentary coloration"
],
"content": "Introduction\n\nMelanism, the occurrence of darker pigmentation in specimens within species is well-known in insects1,2. It is also known that chromatophores to those cells present in the animal integumentary system containing pigments, and these cells respond to hormonal factors and/or neuronal (response to stimuli) factors2. There are several studies focusing on why melanism has evolved, and a hypothesis about this is known as thermal melanism hypothesis2,3. This states that at low temperatures, darker insects are at an advantage compared to light ones because they warm up more quickly. There are several studies that support the thermal melanism hypothesis related to butterflies3–5 and hoverflies6.\n\nInsects have a chemical or pigmentary coloration represented mainly by carotene and melanin, which are found in the cuticle, hypodermis or sub-hypodermis7. In the exocuticle, it is more common to find the variants of carotene and melanin as yellow, brown and black colorings, and in the hypodermis as yellow, red, green and orange colorings8. Adult Schistocerca americana have wings with large brown spots (melanin) on a light-colored background9. The study of the Arthropoda phylum and its Insecta class is highly wide-spread, however, the majority of the studies regarding grasshoppers' behavior under different temperature conditions, focus on habitat selection only10,11.\n\nThe main objective of this study was to analyze the effect of thermal transition on the intensity of the color of the chromatophores present in the wings of the species S. Americana. This study was inspired by the observation in the literature3–6 that temperature has a direct effect on change and/or color variations of the chromatophores located in other species and phyla of the Animalia kingdom.\n\n\nMethods\n\nA total of 31 adult S. americana were collected by hand according to Harris et al. 20133 in Tumbaco- Quito Ecuador (GPS coordinates: 0°13'19.1\"S 78°22'18.2\"W) in May, 2017. No permits were needed for collecting the specimens considering that the S. americana does not appear in the red list (IUCN), and the sampling site is not located in a protected area. Nevertheless, prior consultation was carried out with the local authorities (Mayor's office and Ministry of Environment of Ecuador). All efforts were made to ameliorate any suffering to the animals following the protocol described by the Ecuadorian code of practice for the care and use of animals for scientific purposes of the Ministry of Environment of Ecuador.\n\nThree terrariums were used in the experiment: one with the dimensions 40x40x30 cm and two with the dimensions 70x70x50 cm (Figure 1A). The first terrarium was used for the collection of specimens in the and the others of greater longevity for the creation of an adaptation habitat and for experimentation, respectively. As a substrate, a layer of soil (approx. 4cm) collected in the same location, one layer of stones (approx. 2 cm), and finally a layer of grass (approx. 2cm) was used along with branches of bushes and leaves (Figure 1A). A plastic mesh was used to cover the terrariums. Hydroponic lettuce, cabbage and the grass were used as food for the specimens.\n\n(A top) Mini terrarium for collection; (A bottom) habitat for the specimens during the experiment; (B left) Evidence of adaptation, molt; (B right) Evidence of adaptation, reproduction.\n\nTo ensure adaption from nature to laboratory conditions, 25°C was applied for 7 days before the heat treatments. To observe the color changes, in the laboratory, the specimens were exposed to temperatures of: 45°C, 40°C, 35°C, 30°C, 25°C and 5°C for which a system of five 100w lightbulbs were used as a source of heat. Every two days a light bulb was switched off in order to ensure the decrease of temperature. For low temperature (5°C) all the lights were turned off in the system and dry ice was placed directly on the substrate.\n\nFor the microscopic observation, the specimens were sacrificed at -4℃, and the primary wings were removed from the insects using entomological tweezers, avoiding tearing them, at room temperature according to the method by Paredes et al.12. Optical microscopy (Carl Zeiss, model: Laboval 4) was used for live observation of the color change without coverslip or contrast.\n\nFor color interpretation, we used hexadecimal classification by selecting the wing spot areas using iVinci Express v.4.6 on a 32 scale where the degree of darkness is interpreted in percentage using 0to255 software tool (see the color codes used in the analysis in Underlying data13). Linear regression was applied for analyzing the correlation between the temperature and the degree of darkness in Microsoft Office 365.\n\n\nResults and Discussions\n\nAdaption to the terrarium environment was considered successful as after a one-week period the specimens began molting (Figure 1B left) and reproducing (Figure 1B right).\n\nAt 45°C, degradation of color intensity can be observed in chromatophores present in the wings (Figure 2A). At 40°C chromatophores of the melanin-type begin to darken (Figure 2B); however, between the temperature of 45°C and 40°C there are no huge differences in color change. At 35°C chromatophores of the wings take a light brown color, intensify their color (Figure 2C). At 30°C the chromatophores turn dark brown, indicating an increase in their color intensity (Figure 2D). A temperature of 25°C corresponds to the optimum temperature for the normal development of the species in the region14 (Figure 2E). This is the reference color of the chromatophores used for all analysis. At 5°C chromatophores of the melanin-type intensify their color to brownish-black (Figure 2F). This temperature was the extreme minimum that S. americana could tolerate.\n\nChromatophores (melanin) at a temperature of (A) 45 ⍜C; (B) 40 ⍜C; (C) 35⍜C; (D) 30⍜C; (E) 25 ⍜C; (F) 5⍜C. Magnifications: 3.5 x (left), and 10x (right).\n\nTemperature, precipitation and solar radiation are the meteorological elements that most affect the distribution, rate of growth, reproduction, migration and adaptation of insects15. The thermal melanism in tests with Phaulacridium vittatum and its exposure to the heat of lights within the range 300 to 700 nm represented an intensification in its color with a variation in percentage of between 2.49% to 5.65%3. The majority of the studies that examine the thermal melanism have focused on species with very distinct color morphs representing a wide range in melanism.\n\nWe found negative correlation between the temperature and the degree of darkness (R2 = 0.8038). Our results are in accordance with a previously published study3 in which P. vittatum was examined - a decrease of temperature caused darkening color change in melanin-type chromatophores.\n\n\nConclusions\n\nThe present investigation can be considered as the first initial study of its kind for Schistocerca americana, in terms of examining the changes in the color intensity of the chromatophores present in the wings caused by thermal transition under laboratory conditions. The color change can be considered as an indicator of a reaction to the increase or decrease of temperature2,3 as the clear surface (in this case the wings) reflects the radiation emitted by the lightbulbs and thus absorb less heat, while in the case of a decrease in temperature, the wings absorb more heat.\n\n\nData availability\n\nFigshare: Microscopy photos of the wings. https://doi.org/10.6084/m9.figshare.7749404.v116\n\nFigshare: Raw data for Figure 2. https://doi.org/10.6084/m9.figshare.7749395.v113\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nMajerus MEN: Melanism: Evolution in Action. Oxford University Press. 1998. Reference Source\n\nSadava D: Vida, la ciencia de la Biología. Octava ed., Buenos Aires: Médica Panamericana. 2009. Reference Source\n\nHarris R, McQuillan P, Hughes L: A test of the thermal melanism hypothesis in the wingless grasshopper Phaulacridium vittatum. J Insect Sci. 2013; 13(1): 51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nClusella-Trullas S, van Wyk JH, Spotila JR: Thermal melanism in ectotherms. J Therm Biol. 2007; 32(5): 235–245. Publisher Full Text\n\nLewis JE: Temperature Induced Seasonal Melanism in the Wings of Copaeodes minima (Lepidoptera: Hesperiidae). Fla Entomol. 1985; 68(4): 667–671. Publisher Full Text\n\nJ Holloway G, Marriott CG: Colour Pattern Plasticity in the Hoverfly, Episyrphus balteatus: The Critical Immature Stage and Reaction Norm on Developmental Temperature. J Insect Physiol. 1998; 44(2): 113–119. PubMed Abstract | Publisher Full Text\n\nKarl I, Geister TL, Fischer K: Intraspecific variation in wing and pupal melanization in copper butterflies (Lepidoptera: Lycaenidae). Biol J Linn Soc. 2009; 98(2): 301–312. Publisher Full Text\n\nDe la Cruz Lozano J: ENTOMOLOGÍA, MORFOLOGÍA Y FISIOLOGÍA DE LOS INSECTOS. Palmira: Universidad Nacional de Colombia. 2006. Reference Source\n\nContinente: [En línea]. 2014. Reference Source\n\nBonello S, Piskorowski K, Spotts A: The effects of habitat color and surface temperature on habitat selection of various Melanoplus grasshopper color morphs. University of Michigan Biological Station. 2017. Reference Source\n\nAhnesjo J, Forsman A: Differential Habitat Selection by Pygmy Grasshopper Color Morphs; Interactive Effects of Temperature and Predator Avoidance. Evol Ecol. 2006; 20(3): 235–257. Publisher Full Text\n\nParedes E, Márquez G, Juárez MP, et al.: Quimiotaxonomía de Triatoma longipennis por Análisis de Hidrocarburos Cuticulares. INVURNUS. 2008; 3(2): 3–10. Reference Source\n\nCañizares MB, Naranjo N, Mátyás B: Raw data for Figure 2. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.7749395.v1\n\nMurriagui Ibarra M: EXPANSIÓN URBANA Y DEMANDA DE TRANSPORTE PÚBLICO DE BUSES: CASO DE ESTUDIO PARROQUIAS DE CUMBAYÁ, TUMBACO Y PUEMBO. [En línea]. 2016. Reference Source\n\nRetana J: Relación entre algunos aspectos climatológicos y el desarrollo de la langosta centroamericana Schistocerca piceifrons piceifrons en el Pacífico Norte de Costa Rica durante la fase cálida del fenómeno El Niño-Oscilacion Sur (ENOS). Instituto Metereológico Nacional. 2000; 7(2): 73–87. Reference Source\n\nCañizares MB, Naranjo N, Mátyás B: Microscopy photos of the wings. figshare. Figure. 2019. http://www.doi.org/10.6084/m9.figshare.7749404.v1"
}
|
[
{
"id": "45030",
"date": "14 Mar 2019",
"name": "Ana Danitza Peñafiel-Vinueza",
"expertise": [
"Reviewer Expertise Entomology",
"Ecology",
"Genetics",
"Taxonomy"
],
"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 presented research note addresses the effect of temperature on melanin pigment of wing chromatophores of S. americana by manipulating radiant heat. Despite the results found being in accordance with previous studies, there are a number of issues with the methods and results that need to be clarified. I present more specific comments, as follows:\nIntroduction:\nIt would be helpful to have more information about the ecology and distribution of S. americana, to enrich the introduction and to correlate with the conclusions.\nMethods:\nHousing of specimens: \"first terrarium used for the collection of specimens in the (_) and the others…\" – I think you wanted to state “the first terrarium used for the collection of specimens in the field”, if this is right and since the specimens were collected by hand, I understand that this first terrarium was used for carrying the specimens from the field to laboratory.\n\nThermal transition: Not clear why, if thermal transition ranged from 45° to 5°C, the temperatures tested did not include the ranges from 20° to 10° and instead went directly from 25° to 5°C. Since the distribution of S. americana includes places where the temperature occurs in the range that is missing, testing these temperatures could complete the overview of thermal effects from the whole 45° to 5°C range. Otherwise, please clarify and state your reasons.\n\nIf temperature is induced by manipulating radiant heat to reach the range 45° to 25°C, why did you use dry ice to induce the 5°C temperature? The methodology used in the laboratory should be homogeneous, otherwise it can increase variability in results.\n\nData collection: How many wings were analyzed for each treatment? The number should be stated here, it seems that only one wing was analyzed for each treatment.\nResults and Discussions:\nThe statement “the majority of the studies that examine the thermal melanism have focused on species with very distinct color morphs representing a wide range in melanism” (third paragraph, line 7-9). Is taken almost exactly as is stated by Harris et al. (20131) (reference #3 in your manuscript) in their conclusions.\n\nA discussion about how the negative correlation between temperature and pigmentation that you have found, relates with the adaptation to the environment where S. americana lives.\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? Partly\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? No\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "45936",
"date": "09 Apr 2019",
"name": "David Terán",
"expertise": [
"Reviewer Expertise Biology",
"ecology",
"medical entomology",
"biochemistry",
"structural biology",
"drug discovery"
],
"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\nIncidence of the thermal transition in the range of 45-5°C in chromatophores present in the wings of Schistocerca Americana\n\nThe manuscript presented by Cañizares and co-authors reports the effect of temperature (radiant heat) in wing chromatophores of S. americana. There are other publications about this topic, and the present manuscript is in accordance of reported results. However, the manuscript requires addressing of different issues to be accepted. The suggestions are the following:\n\nIntroduction:\nThe citations 11-12 could have a peer review source.\n\nThe description of the natural environment of S. americana is missing. This information will enrich the description of the hypothesis of the manuscript.\n\nMaterial and methods:\nThe description of house-keeping is not clear.\n\nThe range of temperatures is between 45-5˚C however there is a gap between 25-5˚C.\n\nThere is no statement about how the temperature was controlled.\n\nThere is no statement of how long the S. Americana was kept at certain conditions.\n\nResults and discussion:\nIt’s true that “Temperature, precipitation and solar radiation are the meteorological elements that most affect the distribution, rate of growth, reproduction, migration and adaptation of insects” as stated in paragraph 3 lines 1-3. However, are the authors sure that they can mention this in a very short time of exposure in their experiment?\n\nThe last paragraph about the negative correlation requires more discussion; as it is the most interesting part of the paper.\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? I 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? Yes",
"responses": []
},
{
"id": "47128",
"date": "15 Apr 2019",
"name": "Daniel A. Lowy",
"expertise": [
"Reviewer Expertise Analytical 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\nThe authors report their study on the variation of color intensity of the chromatophores present in the wings of Schistocerca americana. For this, 31 specimens were subjected to thermal transitions within a range of decreasing temperatures, from 45 to 5°C. Darkening and intensifying of color was revealed by optical microscopy of the insect wings.\nThe reviewer believes that this study represents a novelty and is of interest to the scientific community. The authors' approach may have an impact in the monitoring of local climate changes. The manuscript is well written, has a good flow, and its English style is good. The observations listed below are intended to serve for further improving the quality of the paper.\n\nObservations regarding the content:\nPlease, add to the Abstract, as well, the following statement of the Conclusion: \"The color change can be considered as an indicator of a reaction to the increase or decrease of temperature as the clear surface (in this case the wings) reflects the radiation emitted by the light bulbs and, therefore, absorb less heat, while at decreasing temperature, the wings absorb the more and more heat.\"\n\nPlease add to the Introduction: \"It is believed that the detected color in natural environment may become an indicative of local climate change.\" The reviewer believes that instructing school students to survey the color of insects may represent a statistical means to reveal possible micro changes in climate.\n\nThree additional references would be welcome: (i) at the end of section on Specimen collection: Reference (or web site) for the protocol of the Ministry of Environment of Ecuador; (ii) for iVinci Express v.4.6. (iii) for 0to225 software.\n\nHousing of specimens – the second sentence is incomplete: \"...the collection of specimens in the (???) and the others of greater longevity...etc.\"\n\nSuggestions for style improvements:\nIntroduction:\nThe following sentence reads difficultly: “It is also known that chromatophores (to those) OF cells present in the animal integumentary system containing pigments, (and these cells) respond to hormonal factors and/or neuronal (response to stimuli) factors2.” Please, delete the words in parentheses and add \"OF\".\n\nLast sentence of paragraph 1: \"insects are at (an) advantage...because they ABSORB SOLAR LIGHT MORE EFFICIENTLY, SO THAT THEY warm up more quickly.\" Please, delete the words in parentheses and add the capitalized part for more accuracy.\n\nParagraph 3: ...by the observation REPORTED in the literature (3-) etc.\n\nThermal transition:\nThe temperature was set to 25°C (rather than \"was applied\")...were exposed to DECREASING temperatures of 45°C, 40°C, etc.\n\nResults and Discussions:\n...at 45°C and 40°C there are no (huge) SIGNIFICANT differences etc. At 35°C chromatophores of the wings become light brown, intensifyING their color (rather than \"intensify\").\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": []
}
] | 1
|
https://f1000research.com/articles/8-223
|
https://f1000research.com/articles/7-1825/v1
|
20 Nov 18
|
{
"type": "Clinical Practice Article",
"title": "Metachronous or synchronous male breast and prostate cancers a duality to lookout for.",
"authors": [
"Alain Mwamba Mukendi",
"Eunice Van Den Berg",
"Sugeshnee Pather",
"Rushen Siva Padayachee",
"Eunice Van Den Berg",
"Sugeshnee Pather",
"Rushen Siva Padayachee"
],
"abstract": "Introduction: Breast cancer is well known as the stereotypical women's cancer, and prostate cancer represents the well-known stereotypical male counterpart. While prostate cancer carries the potential to metastasize to the breast, the synchronous or metachronous co-occurrence of primary breast and primary prostate cancers is quite unusual. Prostate cancer in men of African descent may have its own behavior with regards to its relationship with male breast cancer.\n\nCase presentation:\n\nCase 1: A 64 year old male presented to Chris Hani Baragwanath Hospital (CHBAH) with a 2 years history of a painless left breast lump. A core biopsy was confirmed breast carcinoma. Tamoxifen was started but, due to disease progression, he underwent left modified radical mastectomy followed by chemotherapy. Prostate biopsy was done for raised Prostate Specific Antigen (PSA) and suspicious prostate on digital rectal examination. A prostatic adenocarcinoma was subsequently diagnosed with bone metastasis on bone scan. He was started on Androgen deprivation therapy and followed up every 3 months.\n\nCase 2: A 68 year old male presented to CHBAH with a 1 year history of a painless right breast lump. A core biopsy confirmed breast cancer. Tamoxifen was started, followed by right modified radical mastectomy and chemotherapy for disease progression. A raised PSA and suspicious prostate on digital rectal examination prompted a prostate biopsy revealing a prostatic adenocarcinoma. Bone scan was negative for metastasis. He is currently on 3 monthly Androgen deprivation therapy and awaiting radiation.\n\nConclusion: This clinical practice article not only presents this exceptionally rare duality but highlights that both cancers can coexist either as sporadic conditions, or as a result of genetic mutations. Thus, we suggest that men with prostate cancer be screened clinically, biochemically and genetically for breast cancer and vice versa.",
"keywords": [
"Male breast cancer",
"prostate cancer",
"duality",
"metachronous",
"synchronous",
"men of African descent."
],
"content": "Introduction\n\nProstate cancer is one of the leading causes of male cancer death1. It is the most common conditions seen in patients in our outpatient department. It represents about one third of conditions encountered on a daily basis, out of about 200 consultations.\n\nMale breast cancer is very rare, and comprises only about 1% of all breast cancers1,2. Prostate cancer and breast cancer synchronous or metachronous duality is quite unusual3. Most of the reported cases of this dual pathology were prostate cancer patients on estrogen therapy who later on developed breast cancer4,5.\n\nMen with a family history of breast or prostate cancer have elevated prostate cancer risks, including risk of very aggressive disease6. Sporadic cases are exceptionally rare1.\n\nWe present 2 cases of primary prostate and primary breast cancers in patients of African descent, who were diagnosed first with breast cancer without any family history of breast cancer, and then with prostate cancer. This is the first such case series from Africa.\n\n\nCase presentation\n\nA 64 year old black male, retired forensic pathology auditor, was referred to urology clinic at CHBAH from medical oncology at the end of June 2018 with a prostate specific antigen (PSA) of 43.82 ng/dL. His hospital attendance had begun in October 2016 when he presented to the breast unit with a 2 years history of a painless progressively enlarging left breast lump with further investigations revealing carcinoma of the left breast. He was diabetic, hypertensive on treatment, and HIV negative. There was no known family history of breast, prostate or any other cancers. There was no history of undescended testes, testicular injury or mumps orchitis. A left modified radical mastectomy was performed in April 2017. He was previously a smoker, smoking 6 cigarettes per day for 40 years (12 pack-year), and quit in May 2017.\n\nOn physical examination, performed in June 2018, scars from left mastectomy and axillary lymph nodes dissection were noted. He had a normal right breast with no palpable lumps. Digital rectal examination revealed an approximately 30 g firm prostate with a nodule on the left lobe. The rest of the examination was unremarkable.\n\nMammography (October 2016). Mammograms showed the left breast was Breast Imaging Reporting and Data System (BIRADS) category 6 and BIRADS category 1 for the right breast.\n\nCore biopsy of the left breast (October 2016). Pathological diagnosis: An invasive carcinoma of no special type.\n\nImmunohistochemistry: Estrogen receptors: Strong positivity in 67–100% of cells.\n\nProgesterone receptors: Strong positivity in 34–66% of cells. HER2neu: Score 1+ Negative. Ki-67: 20%\n\nLeft breast and axillary dissection histopathology assessment (April 2017). On macroscopic examination the left breast nipple and skin were unremarkable. On section a tumour was noted in the lower outer and inner quadrants. It measured 40mm in maximal diameter. Microscopically, the tumour comprised infiltrating epithelioid cells in a nested growth pattern, with some tubule formation. There was marked nuclear pleomorphism, and eight mitotic figures were noted in ten high power fields. The modified Bloom and Richardson classification was Grade 2. Five lymph nodes were identified, and all of these contained tumour metastases.\n\nImmunohistochemical stains were positive for estrogen receptors, negative for progesterone receptors, and equivocal for HER2. FISH testing was negative for HER2 amplification. Ki67 staining showed a proliferation index of 20% (Figure 1). The PSA immunochemistry staining was negative confirming primary breast carcinoma.\n\n(H&E stain, X400 magnification).\n\nBloods tests.\n\n-Prostate specific antigen: 43.82ng/dL (normal range=0–4.5)\n\n-Lactate dehydrogenase: 254 U/L (normal range=48–115)\n\n-Calcium: 2.63 mmol/L (normal range=2.15–2.50)\n\n-Alkaline phosphatase (ALP): 49 U/L (normal range=53–128)\n\nProstate biopsy. Microscopic examination of the prostate core biopsies demonstrated an invasive prostatic adenocarcinoma. The tumour comprised predominately infiltrative, poorly formed, fused glands, and a minor component of more well differentiated glandular areas that splayed muscle fibres. The poorly formed fused glands were compatible with a Gleason pattern of 4 and the well-formed glands were compatible with a Gleason pattern of 3 (Gleason score = 7; WHO grade group 3). The tumour cells had large nuclei, conspicuous nucleoli and abundant pale eosinophilic cytoplasm. The immunohistochemical profile of the tumour (CK 7 and CK20 negative; PSA monoclonal positive) confirmed primary prostatic origin (Figure 2).\n\n(X400 magnification).\n\nComputerised tomography (CT) scan of the brain, chest and abdomen. Global cerebral involutional changes, postsurgical changes in the left chest wall and axilla, and lytic sclerotic lesions at L3 and right iliac bone were found on the CT scans.\n\nBone scan. Active pathology at L3 and right iliac bone were found on Bone scan.\n\nInitially assessed as T2N0M0 in October 2016, he was put on tamoxifen 20 mg per os daily. 6 months later (April 2017) in view of disease progression to T4bN1Mx, a left mastectomy and left axillary lymph nodes dissection was done. Tamoxifen was replaced by anastrozole 1 mg daily and He was then referred for chemo-radiation. He received 4 cycles of chemotherapy AC-T (Adriamycin, cyclophosphamide and taxol) completed in January 2018. A second primary cancer, a prostatic adenocarcinoma was diagnosed in July 2018, almost 2 years after the initial one, for which he is currently on Goserelin 10.8 mg subcutaneously every 3 months. He has so far received 2 injections (July and October of 2018) and will follow up in 3 months for the next injection, and PSA/testosterone monitoring will be done at 6 months as per the hospital policies (funds restrictions). He is unlikely to be given radiation therapy for curative intent as the disease was found to be metastatic.\n\nA 68 year old black male, retired teacher, was referred to urology clinic at CHBAH in May 2018 from medical oncology with a PSA of 113 ng/dL. He first presented in December 2016 to CHBAH breast unit with a 1 year history of a painless right breast lump with further investigations revealing carcinoma of the right breast. He reported that his father died of cancer, but does not know which cancer it was. He had no medical history, was HIV negative, and had no history of undescended testes, mumps orchitis or testicular injury. He was a heavy smoker who smoked 20 cigarettes per day for 40 years (40 pack-year) and quit in July 2016. Right mastectomy and axillary lymph nodes dissection was performed in April 2017.\n\nOn physical examination, scars from right mastectomy and axillary lymph nodes dissection were observed. He had a normal left breast with no palpable lumps. Digital rectal examination revealed an approximately 40 g hard nodular prostate. The rest of the examination was unremarkable.\n\nMammography (December 2016). Mammograms showed the left breast was BIRADS category 1 and BIRADS category 5 for the right breast.\n\nCore biopsy of the right breast (December 2016). Pathological diagnosis: Infiltrating duct carcinoma displaying cribriform features.\n\nImmunohistochemistry: Estrogen receptor positive; strong 3+ intensity staining in >91% of tumour cells. Progesterone Receptor positive; strong 3+ intensity staining in 70% of tumour cells. HER2 negative score 1+ in approximately 10% of tumour cells. Ki-67 the tumour proliferation index approaches 30%.\n\nThe right breast histopathological assessment. Histopathological examination showed an infiltrating duct carcinoma (Figure 3), with a Nottingham combined histopathological grade of 1/3 (tubules 1, pleomorphism 2 and mitoses 1). Estrogen and progesterone receptors were expressed diffusely within the tumour cells (Figure 4). Foci of low-grade cribriform ductal carcinoma in situ were also evident (Figure 5). The PSA immunochemistry staining was negative confirming primary breast carcinoma.\n\nBloods tests.\n\n-Prostate specific antigen: 113ng/dL (normal range=0–4.5)\n\n-Lactate dehydrogenase: 289 U/L (normal range=48–115)\n\n-Calcium: 2.56 mmol/L (normal range=2.15–2.50)\n\n-ALP: 73 U/L (normal range=53–128)\n\nProstate biopsy. Biopsy results showed an invasive prostatic adenocarcinoma (Figure 6), with a modified Gleason score of 9, and grade group 5, which was infiltrating into skeletal muscle (Figure 7). The proportion of tumour infiltration was >90% within the cores. There were areas of lymphatic invasion and perineural invasion.\n\nBone scan. No evidence of osteoblastic metastases was found.\n\nHe was assessed from the beginning in December 2016 as T4bN0Mx, he was put on tamoxifen 20 mg per os daily. 4 months later (April 2017), due to disease progression on tamoxifen from T4bN0Mx to T4bN1Mx, a right mastectomy and right axillary lymph nodes dissection was done prior to referring him for chemo-radiation. He received 6 cycles of chemotherapy FAC, (Fluorouracil, Adriamycin and cyclophosphamide) completed in January 2018; he is still on tamoxifen. A second primary non metastatic cancer arising from the prostate, a prostatic adenocarcinoma, was diagnosed in May 2018, 17 months after the first primary. He is currently on Goserelin 10.8 mg subcutaneously every 3 months and has already received 2 doses. He is awaiting radiation therapy to the right chest. PSA/testosterone monitoring will be performed at 6 months as per the hospital policies (funds restrictions).\n\n\nDiscussion\n\nProstate cancer is the most common male cancer in the U.S. and second leading cause of cancer death. Male breast cancer is very rare, and comprises only about 1% of all breast cancers1,2. Prostate cancer and breast cancer, as synchronous or metachronous dual primary cancer, is quite infrequent3. They are both part of the hereditary breast and ovarian cancer (HBOC) syndrome which is associated with gene mutations in BRCA1 and BRCA27. The BRCA alteration can be inherited from either of the parents and be responsible for a familial type of breast and prostate cancers which has also been previously reported6,8.\n\nBRCA2 and BRCA1 mutation carriers have respectively 40% and 20% lifetime risk of prostate cancer, whereas the lifetime risk of male breast cancer is approximately 5% to 10%, and 1% to 5% for BRCA2 and BRCA1 mutation carriers, respectively9. Mutations in BRCA2, in particular, have been associated with more aggressive clinicopathologic characteristics of prostate cancer and worse outcomes7. The BRCA germline mutations in male cancers are found in only 1% – 2 % of sporadic prostate cancer10. However, prostate and male breast cancers have some similarities with regards to etiology and therapeutic approaches2,9,11.\n\nProstate cancer carries the potential to metastasize to the breast3. Male primary breast cancer and breast metastases from prostatic carcinoma may have similar histological features12. Therefore it is crucial to exclude breast metastases from prostatic adenocarcinoma through a thorough histopathological assessment. PSA immunohistochemistry may help to distinguish between primary and metastatic disease within the breast specimen. This step is very important because of its therapeutic and prognostic implications12. Breast specimens from our 2 cases were evaluated for possible prostatic metastases using PSA immunochemistry staining and were both negative.\n\nAndrogen deficiency or excess estrogen increases the risk of male breast cancer in men with a history of testicular injury, mumps orchitis, or undescended testes3. Klinefelter's syndrome also increases the risk of male breast cancer. Estrogen treatment for prostate cancer was reported as another risk factor for the development of male breast cancer3,4. The latter risk factor is associated with most of the reported cases of this dual pathology4,5. None of these risk factors were found in our 2 cases.\n\nMetachronous or synchronous occurrences of both breast and prostate cancer are scarce in the current literature. Lee et al. evaluated 161 male breast cancer patients for prostate cancer between 1977 and 2000, and found that only 10 patients were also diagnosed with prostate cancer. Breast cancer was the first primary tumour in 8 of these 10 patients3. The association of a sporadic male breast cancer with any other primary tumours, including prostate cancer, is exceptionally uncommon1,9. As with our 2 patients, just a few cases of this combination without any family history of breast cancer or any genetic mutations have been reported13–16. Treatment for each cancer includes hormone manipulation, which may increase the risk for subsequent cancers17. There is no available guideline for treatment of this association, hence further research is necessary.\n\nThis is in our knowledge the first reported cases from patients of African descent living in Africa. Unfortunately, we could not evaluate our patients in terms of genetics because of the lack of some technical facilities. Prostate cancer from men of African descent may have its own behaviour with regards to its relationship with male breast cancer as seen in our 2 cases, both are high risk for metastases but one seems to be more aggressive than the other one.\n\nProstate cancer is in general a slow growing cancer and on that account many years have to pass before developing any symptomatology8. We strongly believe that in those 2 patients even though prostate cancer was the second diagnosed primary they both coexisted undiagnosed at some point in time. Therefore it is absolutely difficult to say which one started first. Most likely if prostate cancer screening was done when they were in their 50s, prostate cancer would have been detected first.\n\n\nConclusion\n\nBreast and prostate cancer as synchronous or metachronous primary cancers are very rare as dual primary malignancy probably because cases are not being reported. Having 2 patients with both primary malignancies may indicate that there may be more patients with similar diagnosed or undiagnosed conditions out there. Both cancers can be either sporadic or as a result of genetic mutations.\n\nWhen either a breast or prostate cancer is diagnosed in a male patient, clinicians should be alert for the other, however, because such a co-occurrence is rarely diagnosed or reported this rarely happens. More studies and research may be helpful to understand the behaviour of these synchronous or metachronous breast and prostate cancers particularly in an African setting.\n\nA suggestion is therefore made that men with prostate cancer be screened clinically, biochemically and genetically for breast cancer, or vice versa, in view of possible common genetic factors contributing to the pathogenesis of both cancers.\n\n\nTake-away lessons\n\nSynchronous or metachronous Breast and prostate cancer are rarely reported, and sporadic cases are extremely rare.\n\nBRCA mutation is found to be a common genetic pathway in both cancers.\n\nBecause of the possible link between the 2, clinicians should lookout for this duality regardless of which one is diagnosed, being investigated or worked up first.\n\nIt is crucial to do PSA staining on both specimen (breast and prostate) to exclude metastases.\n\n\nConsent\n\nWritten informed consents were obtained from both patients for publication of this manuscript and accompanying results.\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\nAcknowledgements\n\nMrs. Joelle Bukumbabu Mukendi (Registered Nurse at Chris Hani Baragwanath Academic Hospital /paediatric oncology department) for helping with referencing and uploading files.\n\nMrs. Mosase Lehloka-Mashabane (Registered Nurse/ Oncology trained at Chris Hani Baragwanath Academic Hospital/ Urology department) for helping with obtaining consent from the patients and providing feedback from the patients during routine follow ups.\n\n\nReferences\n\nOzet A, Yavuz AA, Kömürcü S, et al.: Bilateral male breast cancer and prostate cancer: a case report. Jpn J Clin Oncol. 2000; 30(4): 188–190. PubMed Abstract | Publisher Full Text\n\nCutuli BF, Lacroze M, Dilhuydy JM, et al.: [Breast cancer in men: incidence and types of associated previous synchronous and metachronous cancers]. Bull Cancer. 1992; 79(7): 689–96. PubMed Abstract\n\nLee UJ, Jones JS: Incidence of prostate cancer in male breast cancer patients: a risk factor for prostate cancer screening. Prostate Cancer Prostatic Dis. 2009; 12(1): 52–56. PubMed Abstract | Publisher Full Text\n\nKarlsson CT, Malmer B, Wiklund F, et al.: Breast cancer as a second primary in patients with prostate cancer--estrogen treatment or association with family history of cancer? J Urol. 2006; 176(2): 538–543. PubMed Abstract | Publisher Full Text\n\nTajika M, Tuchiya T, Yasuda M, et al.: A male case of synchronous double cancers of the breast and prostate. Intern Med. 1994; 33(1): 31–35. PubMed Abstract | Publisher Full Text\n\nBarber L, Gerke T, Markt SC, et al.: Family History of Breast or Prostate Cancer and Prostate Cancer Risk. Clin Cancer Res. 2018. PubMed Abstract | Publisher Full Text\n\nGiri VN, Hyatt C, Kelly WK, et al.: Current BRCA1/2 Genetic Testing Guidelines for Prostate Cancer and the Implications for Oncologists. 2017. Reference Source\n\nEsserman L, Shieh Y, Thompson I: Rethinking screening for breast cancer and prostate cancer. JAMA. 2009; 302(15): 1685–1692. PubMed Abstract | Publisher Full Text\n\nLecarpentier J, Silvestri V, Kuchenbaecker KB, et al.: Prediction of breast and prostate cancer risks in male BRCA1 and BRCA2 mutation carriers using polygenic risk scores. J Clin Oncol. 2017; 35(20): 2240–2250. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSong WH, Kim SH, Joung JY, et al.: Prostate Cancer in a Patient with a Family History of BRCA Mutation: a Case Report and Literature Review. J Korean Med Sci. 2017; 32(2): 377–381. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPienta KJ, Goodson JA, Esper PS: Epidemiology of prostate cancer: molecular and environmental clues. Urology. 1996; 48(5): 676–83. PubMed Abstract | Publisher Full Text\n\nCarder PJ, Speirs V, Ramsdale J, et al.: Expression of prostate specific antigen in male breast cancer. J Clin Pathol. 2005; 58(1): 69–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarter BS, Bova GS, Beaty TH, et al.: Hereditary prostate cancer: epidemiologic and clinical features. J Urol. 1993; 150(3): 797–802. PubMed Abstract | Publisher Full Text\n\nAkashi-Tanaka S, Fukutomi T, Fukami A, et al.: Male breast cancer in patients with a familial history of breast cancer. Surg Today. 1996; 26(12): 975–9. PubMed Abstract\n\nTulinius H, Olafsdottir GH, Sigvaldason H, et al.: Neoplastic diseases in families of breast cancer patients. J Med Genet. 1994; 31(8): 618–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAnderson DE, Badzioch MD: Familial effects of prostate and other cancers on lifetime breast cancer risk. Breast Cancer Res Treat. 1993; 28(2): 107–13. PubMed Abstract | Publisher Full Text\n\nAbhyankar N, Hoskins KF, Abern MR, et al.: Descriptive characteristics of prostate cancer in patients with a history of primary male breast cancer - a SEER analysis. BMC Cancer. 2017; 17(1): 659. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "40909",
"date": "04 Dec 2018",
"name": "Pankaj Joshi",
"expertise": [
"Reviewer Expertise Reconstructive Urology"
],
"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's an interesting observation.\nI think authors should mention BRCA Gene mutation's role in breast and prostate cancer.\nWhy is the second patient on ADT if bone scan was negative?\nDetails are important when discussing cancer: What was the breast pathology? Ductal or lobular? What were the patients' Gleason scores and PSAs?\nWhich men would you screen for breast cancer since it is so rare - less than 1/100? Which genetic tests? Be clear with screening goals because you are talking about screening parameters because prostate cancer affects a massive population.\nThose are my thoughts to the author(s). It is an idea which needs refinements and more details regarding how and why we should carry their message into practice.\n\nIs the background of the cases’ history and progression described in sufficient detail? No\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? No\n\nIs the conclusion balanced and justified on the basis of the findings? No",
"responses": [
{
"c_id": "4277",
"date": "17 Dec 2018",
"name": "Alain Mwamba Mukendi",
"role": "Author Response",
"response": "Thank you for your review. I am not sure if you could access the entire article most of the points raised have their answers in the article. All the PSA, Gleason scores etc. Are under case presentation. The role of BRCA was also highlighted under the discussion. Both breast malignancies were ductal as shown in figures 3 and 5. We will on the next version specify that for the first case."
}
]
},
{
"id": "41758",
"date": "15 Jan 2019",
"name": "Bannakij Lojanapiwat",
"expertise": [
"Reviewer Expertise Urooncology",
"endourology"
],
"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\nBoth breast cancer and prostate cancer are hormonal dependent and have genetic alteration in recent studies, This article is not perfect due to lack of genetic testing that is difficult to explain the co-incident of both cancer.\nConclusion is quite long, author should rewrite. Conclusion has this sentence 'suggest that men with prostate cancer be screened clinically, biochemically and genetically for breast cancer and vice versa.' Maybe write 'only suggests that men with prostate cancer can be screened for breast cancer.'\nIf possible summary of the case report in Table maybe more understand the natural history and treatment of this rare condition.\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? 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 conclusion balanced and justified on the basis of the findings? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1825
|
https://f1000research.com/articles/7-1352/v1
|
29 Aug 18
|
{
"type": "Software Tool Article",
"title": "The Mega2R package: R tools for accessing and processing genetic data in common formats",
"authors": [
"Robert V. Baron",
"Justin R. Stickel",
"Daniel E. Weeks",
"Robert V. Baron",
"Justin R. Stickel"
],
"abstract": "The standalone C++ Mega2 program has been facilitating data-reformatting for linkage and association analysis programs since 2000. Support for more analysis programs has been added over time. Currently, Mega2 converts data from several different genetic data formats (including PLINK, VCF, BCF, and IMPUTE2) into the specific data requirements for over 40 commonly-used linkage and association analysis programs (including Mendel, Merlin, Morgan, SHAPEIT, ROADTRIPS, MaCH/minimac3). Recently, Mega2 has been enhanced to use a SQLite database as an intermediate data representation. Additionally, Mega2 now stores bialleleic genotype data in a highly compressed form, like that of the GenABEL R package and the PLINK binary format. Our new Mega2R package now makes it easy to load Mega2 SQLite databases directly into R as data frames. In addition, Mega2R is memory efficient, keeping its genotype data in a compressed format, portions of which are only expanded when needed. Mega2R has functions that ease the process of applying gene-based tests by looping over genes, efficiently pulling out genotypes for variants within the desired boundaries. We have also created several more functions that illustrate how to use the data frames: these permit one to run the pedgene package to carry out gene-based association tests on family data, to run the SKAT package to carry out gene-based association tests, to output the Mega2R data as a VCF file and related files (for phenotype and family data), and to convert the data frames into GenABEL format. The Mega2R package enhances GenABEL since it supports additional input data formats (such as PLINK, VCF, and IMPUTE2) not currently supported by GenABEL. The Mega2 program and the Mega2R R package are both open source and are freely available, along with extensive documentation, from https://watson.hgen.pitt.edu/register for Mega2 and https://CRAN.R-project.org/package=Mega2R for Mega2R.",
"keywords": [
"database",
"genotypes",
"genome-wide association studies",
"linkage",
"Mega2",
"phenotypes",
"R",
"SQLite",
"gene-based association tests"
],
"content": "Introduction\n\nDuring an association or linkage analysis project, one may need to analyze the data with several different programs. Unfortunately, it can often be quite difficult to get one’s data in the proper format required by each different computer program. Not only must the data be converted to the proper format, but also the loci must be reordered into the desired order. Writing custom reformatting scripts can be error-prone and very time-consuming. To address these problems, we created Mega21,2, which can be obtained from the University of Pittsburgh site or via BitBucket. The early Mega2 could read input data in only a few formats: LINKAGE format3–5 and Mega2 format. The Mega2 format allowed you to specify additional information in a more straightforward way, via a simple tabular format, than can be done with LINKAGE format. The earliest versions of Mega2 were written in the C computer language.\n\nOver time, Mega2 was upgraded to the C++ computer language and some of the internals were rewritten. Reformatting data for more analysis programs was gradually added to Mega2. The formats for genetic data changed over time, and Mega2 was enhanced to read input data in PLINK format6, VCF format7, and IMPUTE2 format8–11. The volume of genome-wide marker data increased significantly, so Mega2 was extended to support the compression of biallelic genotype data (though still supporting microsatellite and other polymorphic data as non-compressed legacy data). As the volume of genotype data increased, it became slow and tedious to validate the genotype data and generate allele frequency data for each separate analysis that Mega2 was used for. Rather, Mega2 was restructured so that the validation and meta data generation were performed once for a given data set.\n\nUntil recently, Mega2 used C language structures to store its intermediate data before analysis. Several alternative strategies were considered to save these data for subsequent reuse without having to recompute, reanalyze, and revalidate the data. The first choice was to dump the raw structure data to disk. Reloading the data is fast but it is much harder to inspect the data for debugging. Also, the internal pointers would have to be relocated to new storage areas. A better solution, for many reasons, was to create a SQLite table for each C structure and “insert” the C structure data into the table. Mega2 later uses this database, via a menu-driven interface, to provide the data needed for any particular analysis. In addition, the data in the database are inherently inspectable and there are database tools to help to selectively extract data. The data are portable and can be shared on different platforms. Since interfaces for many languages are provided for SQLite, the data are accessible in many languages besides C. Of particular interest is R12 because there is a wealth of existing genetic analysis programs already created, well documented, and available in standard repositories such as The Comprehensive R Archive Network (CRAN) and Bioconductor. Also, researchers regularly use R to develop new analysis algorithms because R provides a rich environment for statistical computing with many useful libraries.\n\nWe describe here our Mega2R package, which makes it easy to load SQLite Mega2 databases directly into R as data frames for further analysis and manipulation. In addition, we document several R functions that illustrate how to use the Mega2R data frames as well as perform useful functions: the Mega2pedgene function to run the pedgene R package13 to carry out gene-based association tests on family data using selected marker subsets, the Mega2SKAT function to run the SKAT R package14 to carry out gene-based association tests on family data using selected marker subsets, the Mega2VCF function to output the Mega2R data as a VCF file and related files (for phenotype and family data), and the Mega2GenABEL function to convert the data frames into GenABEL R objects15. Typically, these R functions are designed to process a small collection of markers at a time. Our versions process all the markers in a transcript for as many genes as desired. Alternatively, the transcripts can be abstracted to just a table of chromosome/start/end base pair positions. Mega2R eases this process of looping over genes, efficiently pulling out genotypes for variants within the transcript boundaries. In addition to describing the functionality of our Mega2R package, we also provide a Use Case illustrating how to apply it in practice.\n\n\nMethods\n\nMega2R is an R package which loads a Mega2-created SQLite database of genetic information into R as data frames for further analysis. Parts of Mega2R are written in C++ for speed and efficiency. Mega2R is memory efficient, keeping its genotype data in a compressed format, portions of which are only expanded when needed. Mega2R has functions that ease the process of applying gene-based tests by looping over genes, efficiently pulling out genotypes for variants within the desired boundaries.\n\nThe Mega2R package is available from the CRAN repository, and can be run within the R environment on Windows, Mac OS X, and Linux computers. The Mega2R CRAN page documents package dependencies. To prepare an input database for Mega2R, the C++ program Mega21,2 is needed; it can be obtained from the University of Pittsburgh site or via BitBucket.\n\nBefore we discuss the Mega2R package, we must describe one important convention of how the database information is stored and how it is accessed later. The Mega2R R package function, read.Mega2DB, reads the tables needed by Mega2R from a Mega2 SQLite database into a collection of data frames. It returns an R environment containing these data frames. (If you are unfamiliar with environments, you can think of them as data frames. For example, ’ENV’$locus_table will access the locus_table variable from the environment, ’ENV’; this is similar to fetching an \"observation\" from a data frame. The difference is when you change data in an ordinary data frame passed to a function, the change does not affect the original data frame; only the function’s local data is changed and all changes are forgotten when the function exits. If you change the data in a data frame of an environment passed to a function, the change is permanent.)\n\nAll Mega2R functions that do not return an environment need to have an environment supplied as an argument; the environment is used to store the data frames that contain the SQLite database. There are two ways to pass the environment to a Mega2R function. If you assigned the result of read.Mega2DB to the variable ’seqsimr’, then you could supply the value ’seqsimr’ as the named argument, ’envir’:\n\n\n\nThe second choice is a bit of a “hack” but it is very convenient. Every Mega2R function (that does not return an environment) has a named argument, ’envir’, defined to take on the default value ’ENV’, as in\n\n\n\nThe code above, assigns the local variable, ’envir’ the default value ’ENV’. Thus, if ’envir’ is not provided in the function call, R will use the default assignment and look up the value of ’ENV’. This “hack” does not handle the case where ‘ENV‘ is defined in an outer frame, not the global environment. In this situation, we search backwards from the calling frame to find the first ‘ENV‘ and use it.\n\nIn the current Mega2 design, there are SQLite tables for loci data, marker data, allele data, map data, phenotype data, and genotype data as well as pedigree and person data. For completeness, Supplementary File 1 shows all the tables and their fields. Here, we will make a few general observations about the tables. The first issue was mentioned earlier: How should we handle pointers. Where necessary, we added a field to a table, which we call a “link” field, viz. locus_link. This field is a unique integer. When the data are read back in by Mega2 via C++, these links are looked up to find a pointer to use. Consider the locus_table and the allele_table. They both have a locus_link. The link in the locus_table is hashed and a pointer to the corresponding entry in the locus_table is stored. When the allele_table is read, the locus_link is looked up, then a pointer is inserted into a new field in the “in memory” locus_table (using the look up pointer) to point to the memory location of the allele_table entry, i.e. the locus_table in memory will point to the allele_table. These “links” are also very useful in R for performing merges of related tables. For example, in the above case, a merge can produce a table with rows containing both locus and allele data. Similar links are used to associate family data with person data, as well as to associate person data with phenotype data and genotype data.\n\nMost tables are composed of integer, numeric and string data (see Supplementary File 1). The phenotype table is an exception. The normal trick of adding an extra column (e.g., the phenotype id) per row to indicate which phenotype’s data is contained in the row becomes more complicated because affection status traits and quantitative traits have different data types and number of values. Instead, the phenotype data for each person is stored as one long raw vector composed of multiple values with 8 bytes of data per value either representing an affection status value or a quantitative value. Similarly, the genotype data are stored as a raw vector with a pair of bytes representing a genotype for non bialleleic markers and a raw compressed bit vector with two bits per genotype for bialleleic markers. Note that the internal compressed genotype vector used by Mega2 is one long vector, while the genotype data written to the database is split into separate vectors for each chromosome. The end of each vector may have 0, 1, 2, or 3 unused bit pairs representing no markers. When reading back in from the database, the Mega2 C++ program reassembles the vectors without these gaps. The R version is not as sophisticated: it reassembles all the chromosome vectors and includes these gaps. It then takes care when referencing the nth marker, to add in the gaps of all the chromosomes that come before the nth marker’s chromosome. Note that reassembling the data with no “virtual markers” would require a lot of bit manipulations in R which would not be terribly efficient; while keeping track of the virtual markers introduced by the gaps is not particularly hard.\n\nAfter using Mega2 to compress and store one’s genetic data in a Mega2 SQLite database, then one can use the Mega2R read.Mega2DB function to read the database into R data frames within the specified R environment.\n\n\n\nIf the verbose flag is true, summary statistics will be printed as the data frames are created. After running the read.Mega2DB function, the ENV environment will contain a set of data frames ending in “_table”, viz. person_table, locus_table. These contain raw data copied directly from the Mega2 SQLite database (although some of the original fields that are only needed for Mega2 are not copied).\n\nThe read.Mega2DB function also creates other data frames containing derived information. The function mkfam is called by read.Mega2DB to create the fam data frame, the equivalent of a PLINK .fam sample information file. It merges the rows from the pedigree_table, person_table, and phenotype_table to make a fam data frame. The pedigree_table has a pedigree_link, person_table has a pedigree_link and person_link, and phenotype_table has a person_link field. These link fields are used by the R ’merge’ function to assemble the data. Later, you may want to prune the fam data frame to restrict the samples processed. For, example you might want to eliminate samples with unknown case/control status. After you prune the fam information, the function, setfam, will set the fam data frame while also appropriately pruning the phenotype_table and unified_genotype_table to contain only those persons left in the fam data frame.\n\nThe unified_genotype_table collects all the raw byte vectors from the genotype_table stored in the input SQLite database by chromosome for each person and concatenates them into a new vector for each person. The concatenation can introduce up to three “virtual markers” (as mentioned above) at the end of each chromosome if the number of markers contained is not an exact multiple of 4 (genotypes per byte). The marker_table contains the offset of the marker genotype data assuming all the markers across all the chromosomes are contiguous. There is an analogous offset column needed for the unified_genotype_table accounting for the “virtual markers” in the gaps; this offset column is computed and added to the marker_table. The markers data frame is a subset of the marker_table data containing only the name, chromosome, position, original marker_table offset and marker offset into the unified_genotype_table vector.\n\nWe might have genotype data for thousands of samples and millions of markers in a study, but we are unlikely to have the memory to support the associated genotype matrix in R. The Mega2 genotype data is compressed in the SQLite database and remains compressed in its data frame. Most of the time, we do not need to decompress all of it at once. We typically want to examine the markers on a specific chromosome or in the region of some genes of interest. We only decompress the set of markers we need. The key idea is we want to find the set of markers that lie within some base pair range and process just those markers and then we repeat this for the “next” range. We provide two ways to specify these ranges, via the setRanges and setAnnotations functions. In addition, we want to iterate through multiple ranges as would appear in a set of selected gene transcripts and process the markers. To do this, we process each range of markers by invoking a callback function on the range, markers and related data, via one of the applyFnToRanges, applyFnToGenes, or applyFnToMarkers functions.\n\nOne of Mega2R’s strengths is its ability, via the applyFnToRanges and applyFnToGenes functions, to loop over a list of ranges or genes, for each one selecting markers that lie within the region/gene of interest. To do this, one has to have well-defined gene boundaries. By default, Mega2R provides an internal list of the chromosome and base pair ranges for the transcripts defined in the refGene table from the UCSC Genome Browser reference assembly GRCH37. The list was modified to eliminate multiple records of the same gene with the exact same transcript start and transcript end. The list contains 29,062 records. But there are several reasons that the default ranges may not be to your liking. The setRanges function lets you use custom ranges; it has two arguments: a range data frame that specifies at least the chromosome, the start position, end position (and optionally a name) of each range, and an index vector of three integers that indicates which columns in the range data frame contain the chromosome, start position and end position. If the range data frame also contains a name column, the index vector should have a final fourth integer that indicates the column containing the name.\n\nThe applyFnToRanges function (and also the applyFnToGenes function, described below) take an initial argument that is a callback function that is called repeatedly for each transcript that has markers present. They also take a final argument that is an R environment. The applyFnToRanges function, with no additional arguments, will iterate through each row of the default Mega2R ranges, determine the markers contained between the start and end and then invoke a callback function. Alternatively, applyFnToRanges may be given a range argument and a selector argument which will cause it to use the ranges supplied rather than the default. These two arguments are the same format as the corresponding arguments provided to setRanges.\n\nThe applyFnToRanges function requires a callback function of three arguments: the first is a data frame of the markers in the range, the second is a single row from the data frame that defines the range (minimally it has a name, ID, start position, end position, and chromosome), and the third argument is the R environment that contains all the Mega2R data frames. The callback function can apply whatever analysis is required and it can access the other data frames available in the R environment and even store intermediate results back into the R environment. Typically, the callback function will call either getgenotypes or getgenotypesraw to get a matrix of genotypes values for the markers in the range.\n\nMega2R can also load a list of genes and their transcripts from the Bioconductor data annotations: “org.Hs.eg.db”and “TxDb.Hsapiens.UCSC.hg19.knownGene”. The former translates a gene name or alias to the entrez ID of the gene. The latter gives start position, end position, and chromosome for each transcript known for every gene. You might want to use the transcript data from another build that is available in Bioconductor or follow their instructions to make your own mapping. The Mega2R function, setAnnotations, lets you use alternate tables. Mega2R will load the selected tables when you access the applyFnToGenes function.\n\nThe applyFnToGenes function specifies a list of genes (via the genes_arg argument) from which to extract the transcripts and the corresponding ranges (If the special gene name, “*”, is passed as the only gene argument, the ranges for all transcripts will be chosen). Alternatively, applyFnToGenes supports other ways to indicate transcript ranges: on sub-ranges of chromosomes (via the ranges_arg argument), on the entire chromosome (via the chrs_arg argument), or on an arbitrary list of markers (via the markers_args argument). The callback function of three arguments (markers, range, environment) and a list of ranges is passed to applyFnToRanges and processed as described above.\n\nThe applyFnToMarkers function is the simplest function that uses the aforementioned callback function. It takes one argument that contains selected rows of the markers data frame. It then invokes the callback function with the marker argument data frame, a NULL range, and the R environment (The range is NULL because there is no range information; this implies that the callback function for applyFnToMarkers must check for a range value of NULL).\n\nThe functions getgenotypes and getgenotypesraw return a matrix of nucleotide pairs or a matrix of encoded integers with a column for each marker and containing a row per sample. Both functions take two arguments: rows of the markers data frame and an environment. The first argument is usually computed by one of the applyFnToRanges, applyFnToGenes, or applyFnToMarkers functions. The markers data frame has two offsets that are used internally by the decompression code; it also has a name, chromosome number and position that can be used to identify a marker to the user. The two getgenotypes functions are dispatches that call Rcpp code. The R code collects genotype and allele data from the R environment’s data frames and passes these data arguments to the Rcpp code; it also calls the proper function for the level of compression: 2 bits or 2 bytes per marker. For each marker specified, getgenotypes returns for each sample a string of the corresponding pair of nucleotides, while getgenotypesraw retruns for each sample a single integer with the integer for the first allele shifted 16 bits and or’ed to the integer for the second allele.\n\nWe now give a brief overview of several processing functions that Mega2R makes available and that illustrate how to build new functions using Mega2R. A much more detailed example may be found in the “Use Case” section.\n\nThe ’pedgene’ R package performs gene-level association tests with disease status for pedigree data13. Mega2R enhances ’pedgene’ via the Mega2pedgene function, which automates and eases computation of the ’pedgene’ statistics for all the genes in the genome. The Mega2pedgene function does this by invoking applyFnToRanges on the specified ranges or calls applyFnToGenes on gene names. Before running the Mega2pedgene function, we first call init_pedgene rather than read.Mega2DB; the former uses the function dbmega2_import to load the Mega2 database. Then it generates a modified fam data frame, a kind of family file where the case/control values of 0/1/2 are translated to NA/0/1, as required by pedgene. The pedgene package requires recoding of the genotypes 1/1, 1/2 (and 2/1), and 2/2 (with raw encodings: 65537, 65538 (and 131073), 131074) to 0, 1, and 2 respectively; plus the alleles must be flipped as needed so that 1 is the major allele. The callback function, DOpedgene, transforms the raw genotype matrix and then it calls the pedgene function with several different marker weights (e.g., unweighted, Madsen-Browning weights, and β density weights). It returns the p-value of the kernel and burden association statistics for each weight. The results are appended to a data frame in the environment, pedgene_results.\n\nIf one has a sample of unrelated individuals, the Mega2SKAT function eases the process of applying the region-based association tests that are implemented in the SKAT package14,16,17. The SKAT package uses kernel regression approaches to compute association statistics such as the Sequence Kernel Association Test (SKAT) and its optimized version (SKAT-O), both for quantitative and dichotomous traits. The Mega2SKAT function is rather similar to Mega2pedgene (above). The init_SKAT function behaves very much like init_pedgene; in addition, it computes a phenotype data frame. The Mega2SKAT function has the signature:\n\n\n\nThe parameters f and ty specify the formula and phenotype type that are used to invoke SKAT_Null_model. The skat argument indicates the particular SKAT package function to call and the ... arguments are place holders for any additional arguments that the chosen skat function needs.\n\nThe VCF data format was originally defined by the 1000 Genomes Project18 for data storage19. The current version of data format is available from GitHub20. The Mega2VCF function serves several purposes. It allows VCF format files to be generated for analysis by programs that require VCF input. It shows how to extract data provided in the Mega2R data frames for subsequent analysis by other R packages. Finally, it provides a regression for Mega2R by producing the same files from an SQLite database via R code that Mega2 can produce from the identical database via C++ code. The VCF file and related files created both ways should be the same except for time stamps.\n\nBefore running the Mega2VCF function, you should first call the read.Mega2DB function to load a Mega2 database. A .vcf VCF file is the main output of Mega2VCF; it is created from the genotype matrix, supplemented by columns from the fam table and allele_table table. Besides the .vcf file, Mega2VCF generates several additional complementary files. Below, we indicate the file suffix, the internal function that generates the file, and the file contents are presented in Table 1.\n\nThe code in each function illustrates how to assemble the corresponding data from the Mega2R data frames. The mkphenotype function is a bit subtle. The database contains a raw vector of 8 bytes times the number of phenotypes (per person). The 8 bytes encode either one double precision float number or two integers depending on the encoding of the trait phenotype in the locus_table (quantitative or affection status). The readBin function is used to convert the raw bytes in each 8 byte vector to the correct form.\n\nAnother interesting aspect of the code design is how we compute genotype data for a large numbers of markers without needing large amounts of memory. We build the .vcf file a chunk at a time, where a chunk contains a relatively small number of markers (currently 1,000). We generate the genotype matrix for a chunk then append the results for the chunk (of markers) to the .vcf file. Then we get a new chunk and repeat. This looping strategy might prove useful for other situations where the results can be written incrementally, a block of markers at a time.\n\nThe GenABEL15 package is available in the archive section of CRAN (https://CRAN.R-project.org/package=GenABEL). The GenABEL package provides many functions for genome-wide association analysis and it accepts data in several formats. But Mega2 accepts input in still more formats, notably VCF, PLINK, IMPUTE2 and even Linkage format. Thus Mega2GenABEL can be a bridge to easily convert data for GenABEL analysis.\n\nBefore running the Mega2GenABEL function, you must call the read.Mega2DB function to load your database and you should save the returned environment into ’ENV.’ The function, Mega2GenABEL, returns a gwaa.data-class object for the selected markers. It operates by re-writing the Mega2R data frames into one of the input formats that GenABEL supports; currently, this is PLINK .tped format. This requires a .tped file, a .fam file, and a .phe file; they are stored in a temporary directory. The .tped file looks very much like a VCF .vcf file with entries in each marker’s row for the samples’ genotype data. (But, in comparison to the .vcf file, there are fewer fields of metadata in a .tped file and the genotype data specify real allele labels rather than VCF’s REF/ALT field references.) Mega2GenABEL calls the GenABEL “glue” functions, convert.snp.tped, to process the a .tped file and .fam file into a generic GenABEL “raw” format file, then it converts that raw format file and a phenotype file into a GenABEL gwaa.data-class object via the function, load.gwaa.data, and then it deletes the temporary files.\n\nThe function Mega2ENVGenABEL produces the same object as Mega2GenABEL. It does not write out any temporary files, but rather converts the Mega2R genotype and related data in memory to a GenABEL gwaa.data-class object. It is mainly written in Rcpp and runs 10 to 20 times faster than the Mega2GenABEL function.\n\n\nUse case\n\nWe now show some examples of how to use Mega2R. First, we illustrate the base functionality of the Mega2R package (e.g., iterating over gene ranges and getting genotypes from the database). Then we demonstrate one of the extended functions, Mega2pedgene. The other extended functions, Mega2SKAT, Mega2VCF and Mega2GenABEL are illustrated in the Mega2R package vignette. Further, details of all the Mega2R functions are available in the Mega2R manual Finally, the package source can be obtained from CRAN or it can be obtained via git from Bitbucket.\n\nThe R code in the examples below was executed during the creation of this document using the knitr R package (version 1.20)21–23.\n\nWe used the SeqSIMLA224 program, which is available from SourceForge, to simulate an example dataset. To do this, we started with the first example under the “Prevalence” tab within the “Simulate by Disease status” tab on the tutorial page. This simulation creates 1,380 samples of 500,000 markers on a subrange of chromosome 1. To illustrate the Mega2 and Mega2R operations that follow, we subsampled the data down to 1,380 people and 1,000 polymorphic markers.\n\nInstallation R. We will assume that you have started an R session in which you type the commands in this Use Case (We used R version 3.5.0).\n\nYou should first install the package Mega2R.\n\n\n\nBioconductor. Below we will use ’pedgene’ carry out gene-based association tests, where ’genes’ are defined according to a database containing the boundaries of the gene transcripts. This requires two Bioconductor Annotations databases to be installed. The first line below loads the Bioconductor biocLite loader and the next two lines install two annotation databases. One annotation database provides the gene transcript locations and the other maps gene names to Entrez gene IDs. (As described in the section about pedgene below, you may choose a different transcript database from Bioconductor or construct one of your own.) Please type in R:\n\n\n\nThe above steps are run once.\n\nUse case data. First, you will need to load the Mega2R package. Type:\n\n\n\nThe files you will need for this Use Case are provided in the Mega2R package. You need to extract these files to the current directory via this command:\n\n\n\nYou should see the following files:\n\n\n\n\n\nNote: The use of the “mega2” executable in our examples expects the Mega2.BATCH.<name> files to be in the working directory and the latter files expect their data files to be in the working directory. When you are done with these exercises, the “clean” command will remove these files and other temporary files:\n\n\n\nCreating a Mega2 database. We have provided files in this package that contain the data from the simulation. These files are in PLINK ped format data:\n\n• Mega2r.ped\n\n• Mega2r.map\n\nIf you do not wish to install Mega2 right now, you can skip this database creation step and instead later use the seqsimr.db database that was placed in your directory by the dump_mega2rtutorial_data(“.”) command. You can obtain the Mega2 program from the University of Pittsburgh site. Then, you will invoke Mega2 on your data. To make matters simple, we will use a pre-constructed Mega2 batch file to automate the processing by Mega2. To run Mega2 to process and create the ’SQLite’ database ’seqsimr.db’, we issue this command at the Unix prompt from within the directory containing the Use Case example files:\n\n\n\nNote: The vignette associated with the Mega2R package illustrates what the results of this Mega2 run would look like.\n\nIf you do not provide the command-line argument giving the name of a BATCH file, Mega2 will proceed to ask a series of interactive questions to collect the information needed to produce a database. In addition, it will create a Mega2.BATCH file, similar to the one we suggested you use. You can look at the “Quick Start” section of the Mega2 documentation to better understand the interactive process.\n\nThe MEGA2.BATCH.seqsimr file begins with a rather long comment section indicating the keyword values that may be set and their default values. Towards the end of the file, we set the inputs to Mega2r.ped and Mega2r.map, indicate the input is PLINK ped format with parameters, and indicate that Mega2 should produce a database called seqsimr.db, etc. (The initial comment section is not shown and unimportant lines are elided.)\n\nInput_Database_Mode=1\n\nInput_Format_Type=4\n\nInput_Pedigree_File=Mega2r.ped\n\nInput_PLINK_Map_File=Mega2r.map\n\n...\n\nPLINK_Args= --cM --missing-phenotype -9 --trait default\n\n...\n\nValue_Marker_Compression=1\n\nAnalysis_Option=Dump\n\n...\n\nDBfile_name=seqsimr.db\n\n...\n\nIf you wish to use any of the Mega2R functions described here on your own data, you will have to first run “mega2” to convert your data into an ’SQLite’ database.\n\nReading and examining a Mega2 database. The Mega2R package facilitates reading genetic data from a Mega2-created SQLite database.\n\nReading a Mega2 database After you have created the SQLite database “seqsimr.db”, start up R, load the Mega2R package, then use the function read.Mega2DB to read a Mega2 database.\n\n\n\nThe first argument should be the path to the database. If ’verbose’ is TRUE, for each data frame created in ’ENV’, read.Mega2DB emits two lines: one with the number of rows and number of columns of the data frame and the other with the column names of the data frame. Finally, an ’environment’, which contains the data frames, is returned.\n\nVerbose Flag When verbose is set in the initial read.Mega2DB, the value will be remembered. It may be used by any subsequent function. If verbose is TRUE, Mega2R functions will print diagnostic information.\n\nUse of an Environment The environment returned is used to store the data frames that contain the ’SQLite’ database. The code above stores the environment in global variable ’ENV’. If the named argument, ’envir’, is not provided in any subsequent Mega2R function call, R will look up the value of ’ENV’ starting at the calling frame and chaining up the call stack to the global environment.\n\nBack to more examples. The ’ls(ENV)’ function will show you all the variables in the ’ENV’ environment. (You probably have used it without arguments to show you the variables in the global environment .GlobalEnv.) Type:\n\n\n\nA more informative overview of the database can be had with:\n\n\n\nUse standard R operations to examine the created data frames Try typing:\n\n\n\nMega2R provides two ways to compute a function on the genotypes (or markers) in each of the transcripts. These are further illustrated via examples in section about ’pedgene’ below.\n\nsetRanges The default ranges contains 29,062 records taken from the UCSC Genome Browser reference assembly GRCH37. We show a bit of the data frame below. Each row contains 5 values: a transcript id, the gene id and three position values: chromosome, start base pair and end base pair.\n\n\n\nYou may load your own range set instead of the default. You create a data frame that lists, for each range, the chromosome, the start position, and the end position. And you create an integer index vector that indicates which column contains which data. These two become the arguments to the setRanges function as shown in the example below. When the index vector contains only three entries, a range name is generated by concatenating the position information and adding it to the range data frame.\n\n\n\nIf you provide an index vector with four entries, the fourth one indicates the column containing the names of the ranges.\n\n\n\napplyFnToRanges The function, applyFnToRanges, goes through each range and finds the markers that fall within the bounds. The first argument of the applyFnToRanges function is a callback function with three arguments: the markers in range, the selected range entry and the environment. The callback is invoked repeatedly for each transcript range that contains any markers. If there are no markers contained and the ’verbose’ flag is set, a warning will be printed. The callback function builds a genotype matrix for the samples and each marker in the range. If necessary, the environment (’envir’) can be used to store information between successive invocations.\n\nFor the examples that follow, we use “show” as the callback function, which simply prints, for each range, the range itself, the first three lines of the markers within the range, and the first three lines of the genotype matrix, in that order. It also prints a little banner before each argument. Finally, it does not print the environment argument value; it does not change.\n\n\n\nA simple example is shown below using the ranges value that were most recently set. We see that the range named “A” contains markers in our example data set.\n\n\n\napplyFnToRanges can also be provided explicit ranges by using the ’ranges_arg’ argument:\n\n\n\nsetAnnotations If you are iterating/selecting via genes, the default transcript database is “TxDb.Hsapiens.UCSC.hg19.knownGene” from Bioconductor; it is stored in the environment as shown below:\n\n\n\nOf course, you can change this database. Suppose we want to use build “hg18”, we would run:\n\n\n\napplyFnToGenes The function applyFnToGenes can apply the ’show’ function to the transcripts of a particular gene via the genes_arg argument. For example, here we apply the ’show’ function to transcripts of the CEP104 gene:\n\n\n\nThe applyFnToGenes function has several other optional arguments that can request complete chromosomes, (multiple) ranges of base pairs on chromosomes, or collections of markers. All these arguments define ranges that are passed to applyFnToRanges for evaluation. Note: If the genes_arg argument is set to the special \"gene\" string \"*\", then all transcripts in the Bioconductor database, will be processed.\n\n\n\nEncoded Genotypes The heart of the above two functions is the calculation of the genotype matrix of the samples. We will build the matrix for the first 10 markers. Remember our database has 1,380 samples so we will just show the head of the matrix.\n\n\n\nRaw Genotypes The function getgenotypesraw will be called with the same 10 markers as above. Recall it returns an integer encoding for each genotype. The integer’s high 16 bits are the index for allele 1 and the low 16 bits are the index for allele 2.\n\n\n\n\n\nMega2R provides functions that permit one to run the ’pedgene’ function by Schaid et al.13 to carry out gene-based association tests on family data looping over selected transcripts.\n\nRather than read the Mega2 SQLite database with the read.Mega2DB function as described previously, here we use a specialized init_pedgene function to read the Mega2 database. This latter function calls a utility function also used by read.Mega2DB. Then it creates, edits, and rewrites the family (.fam) data, storing it in a ’pedgene’-compatible data frame, fam. (fam merges data from the pedigree_table, person_table and phenotype_table.) (When fam is updated, filtering is automatically done to the persons in the phenotype_table and the unified_genotype_table to remove any persons removed from fam.) Finally, init_pedgene calculates some values that will be used repeatedly and stores them in the environment that is returned.\n\n\n\nMega2R has an internal default list of the chromosome and base pair ranges for a number of gene transcripts. These transcripts come from the UCSC Genome Browser reference assembly GRCH37. The list was further modified to eliminate multiple records of the same gene with the exact same transcript start and transcript end. These data contain about 29,000 records. You may use the setRanges function to load your own range set instead of the default, as described above.\n\nBy default, the init_pedgene function sets the transcript database to “TxDb.Hsapiens.UCSC.hg19.knownGene” and the Entrez gene name mapping database to “org.Hs.eg.db”. If you wish to use different databases to select transcripts by gene name, you must use the setAnnotations function to load them from Bioconductor, as discussed above.\n\nBy default, the function Mega2pedgene examines the first 100 transcripts and prints the results. (Internally, it calls the function applyFnToRanges.) For the seqsimr.db database, the first 100 transcripts contain only one transcript with several markers. To make this Use Case run faster, we noticed that the identified transcript appeared at transcript 54; so we will restrict Mega2pedgene to a small range of transcripts around 54, viz. 53 through 55.\n\nNote: ’verbose’ needs to be TRUE, for the diagnostics to be printed.\n\n\n\nYou will see some reports of “No markers in range”, because the database only contains markers on a sub range of chromosome 1 whereas the transcripts span the entire genome. Then you will see a listing of a gene name, a marker name, and count of the 1/1, 1/2, and 2/2 genotypes (where ’1’ is the major allele), e.g.,.\n\n\n\nThe markers, the range used, and the environment are passed to the callback function DOpedgene. DOpedgene converts the raw genotype representation returned by getgenotypesraw to the values 0, 1 and 2. Then it runs ’pedgene’. The results are automatically stored in a data frame with columns: chromosome, gene, number of markers and base pair range followed by ’pedgene’ data: kernel and burden, value and p-values, four values for each of three weightings of the markers. These data are saved in the data frame, ’pedgene_results’, in the environment. They are also printed when ’verbose’ is TRUE.\n\nNote: The results are always appended to the ’pedgene_results’ data frame. You should truncate it when necessary.\n\nYou could run Mega2pedgene on all the transcript entries, but it takes a rather long time. You would type:\n\n\n\nIf you run the above test, you will see that genes DISP1 and KIF26B have at least one p-value less than 0.01 and AK5 and STL7 at least one less than 0.03.\n\nYou may try searching for transcripts of specific genes. Here, the default transcript database is \"TxDb.Hsapiens.UCSC.hg19.knownGene\" from Bioconductor. Of course you can change it, if you install a new database from Bioconductor, as shown earlier.\n\nWe leave the command below as an exercise, as it runs a bit slowly. It needs to find all the transcripts for each gene, then to find all the markers between each pair of transcript start/end ranges, then to compute the genotype matrix for these markers, and finally to call the callback function with the appropriate arguments.\n\n\n\nBut let us run this function for a few genes:\n\n\n\n\nSummary\n\nAs with all software projects, Mega2R is a work in progress, and so there are a number of possible improvements that could be made in the future. These include:\n\n• defer constructing the unified_genotype_table until the set of chromosomes that are needed can be computed and then only read the required chromosome genotype vectors.\n\n• analogously to Mega2SKAT, add support for the R famSKATRC package25,26. famSKAT_RC is another package for family or unrelated gene-based association analysis for a mix of rare variants and common variants.\n\n• analogously to Mega2GenABEL, add support for the CoreArray data storage system (e.g., Genomic Data Structure (GDS) format27). This will support conversion to GDS format in both the SNPRelate and the SeqArray data representations28.\n\n• apply more aggressive compression of the database. For one, the genotype raw vector can be further compressed by gzip or a similar compression protocol. Additionally, since C does not have a way to represent missing data, we use -99.99 to represent it which is stored in the database as a floating point number (8 bytes). But the database (via NULL) and R (via NA) can efficiently represent a missing values, so we will will convert C missing (-99.99) to SQLite (NULL).\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\nSoftware name: Mega2R\n\nSoftware available from: https://CRAN.R-project.org/package=Mega2R\n\nBitBucket repository: https://bitbucket.org/dweeks/mega2r/\n\nDocumentation page: https://watson.hgen.pitt.edu/register/docs/mega2R.html\n\nOperating system(s): Platform independent\n\nProgramming language: R, C++\n\nOther requirements; R, SQLite library\n\nArchived source code (v1.0.4) at the time of publication: https://doi.org/10.5281/zenodo.134358729\n\nLicense: GNU GPL 2",
"appendix": "Grant information\n\nThis work was supported by a National Institutes of Health grant [R01 GM076667] to DEW and the University of Pittsburgh.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1: The Mega2R package: R tools for accessing and processing genetic data in common formats. PDF contains detailed descriptions of the data frames that Mega2R makes available.\n\nClick here to access the data.\n\n\nReferences\n\nMukhopadhyay N, Almasy L, Schroeder M, et al.: Mega2: data-handling for facilitating genetic linkage and association analyses. Bioinformatics. 2005; 21(10): 2556–7. PubMed Abstract | Publisher Full Text\n\nBaron RV, Kollar C, Mukhopadhyay N, et al.: Mega2: validated data-reformatting for linkage and association analyses. Source Code Biol Med. 2014; 9(1): 26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLathrop GM, Lalouel JM: Easy calculations of lod scores and genetic risks on small computers. Am J Hum Genet. 1984; 36(2): 460–5. PubMed Abstract | Free Full Text\n\nLathrop GM, Lalouel JM, White RL: Construction of human linkage maps: likelihood calculations for multilocus linkage analysis. Genet Epidemiol. 1986; 3(1): 39–52. PubMed Abstract | Publisher Full Text\n\nLathrop GM, Lalouel JM: Efficient computations in multilocus linkage analysis. Am J Hum Genet. 1988; 42(3): 498–505. PubMed Abstract | Free Full Text\n\nPurcell S, Neale B, Todd-Brown K, et al.: PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007; 81(3): 559–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDanecek P, Auton A, Abecasis G, et al.: The variant call format and VCFtools. Bioinformatics. 2011; 27(15): 2156–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHowie BN, Donnelly P, Marchini J: A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. PLoS Genet. 2009; 5(6): e1000529. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHowie B, Marchini J, Stephens M: Genotype imputation with thousands of genomes. G3 (Bethesda). 2011; 1(6): 457–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHowie B, Fuchsberger C, Stephens M, et al.: Fast and accurate genotype imputation in genome-wide association studies through pre-phasing. Nat Genet. 2012; 44(8): 955–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarchini J, Howie B: Genotype imputation for genome-wide association studies. Nat Rev Genet. 2010; 11(7): 499–511. PubMed Abstract | Publisher Full Text\n\nR Core Team: R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. 2017. Reference Source\n\nSchaid DJ, McDonnell SK, Sinnwell JP, et al.: Multiple genetic variant association testing by collapsing and kernel methods with pedigree or population structured data. Genet Epidemiol. 2013; 37(5): 409–418. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee S, Emond MJ, Bamshad MJ, et al.: Optimal unified approach for rare-variant association testing with application to small-sample case-control whole-exome sequencing studies. Am J Hum Genet. 2012; 91(2): 224–237. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAulchenko YS, Ripke S, Isaacs A, et al.: GenABEL: an R library for genome-wide association analysis. Bioinformatics. 2007; 23(10): 1294–6. PubMed Abstract | Publisher Full Text\n\nWu MC, Lee S, Cai T, et al.: Rare-variant association testing for sequencing data with the sequence kernel association test. Am J Hum Genet. 2011; 89(1): 82–93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee S, Wu MC, Lin X: Optimal tests for rare variant effects in sequencing association studies. Biostatistics. 2012; 13(4): 762–775. PubMed Abstract | Publisher Full Text | Free Full Text\n\n1000 Genomes. Reference Source\n\nVCF Format (early spec). Reference Source\n\nVCF Format. Reference Source\n\nXie Y: knitr: A General-Purpose Package for Dynamic Report Generation in R. [R package version 1.17]. 2017. Reference Source\n\nXie Y: knitr: A Comprehensive Tool for Reproducible Research in R. In Implementing Reproducible Computational Research. Edited by Stodden V, Leisch F, Peng RD, Chapman and Hall/CRC. 2014; [ISBN 978-1466561595]. Reference Source\n\nXie Y: Dynamic Documents with R and knitr. Boca Raton, Florida: Chapman and Hall/CRC, 2nd edition, 2015; [ISBN 978-1498716963]. Reference Source\n\nChung RH, Tsai WY, Hsieh CH, et al.: SeqSIMLA2: simulating correlated quantitative traits accounting for shared environmental effects in user-specified pedigree structure. Genet Epidemiol. 2015; 39(1): 20–4. PubMed Abstract | Publisher Full Text\n\nSaad M, Wijsman EM: Combining family- and population-based imputation data for association analysis of rare and common variants in large pedigrees. Genet Epidemiol. 2014; 38(7): 579–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKunji KB, Saad M: famSKATRC: Family Sequence Kernel Association Test for Rare and Common Variants. R package version 1.1.0. 2017. Reference Source\n\nZheng X, Levine D, Shen J, et al.: A high-performance computing toolset for relatedness and principal component analysis of SNP data. Bioinformatics. 2012; 28(24): 3326–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZheng X, Gogarten SM, Lawrence M, et al.: SeqArray-a storage-efficient high-performance data format for WGS variant calls. Bioinformatics. 2017; 33(15): 2251–2257. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBaron RV, Weeks DE: The Mega2R R package, version 1.0.4. Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1343587"
}
|
[
{
"id": "37771",
"date": "11 Sep 2018",
"name": "Dajiang J. Lui",
"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\nWeeks and colleagues described a very useful R package to access multiple data formats and facilitate its integration with R packages. This package is based upon its predecessor MEGA2, which is highly cited and used in genetic data analysis.\n\nThis tools is long overdue. While numerous statistical genetics methods are being developed, only a minority of them have really a usable interface. Due to the metadata that modern genetic datasets needs= to store, many of the files are not stored in a rectangular format. Parsing these files and making them analyzable in R is often the most time-consuming part to code but is the key for the development. The package filled in this gap and make easy the development of useful softwares.\n\nI only have a few minor comments:\n\n1) The software is unique and useful, but the existence of other packages that can read and parse genetic datasets in R should be acknowledged. For example, VariantAnnotation package in bioconductor and seqminer in CRAN can both randomly access tabix indexed VCF/BCF files. 2) One of the major strength of the package is to facilitate statistical genetics method development. So it would be helpful if the authors could give a toy example on how to make use of applyFnToGenes to handle a user defined statistical tests.\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": "4376",
"date": "25 Feb 2019",
"name": "Daniel Weeks",
"role": "Author Response",
"response": "Thank you for your helpful suggestions for improving our paper. In Version 2, as suggested, we now cite, in the Introduction, other R and Bioconductor packages that can read and parse genetic datasets. As suggested, we have added, to the Use Case section, a toy example illustrating how to use applyFnToGenes to apply a user-defined statistical test."
}
]
},
{
"id": "37770",
"date": "13 Sep 2018",
"name": "Alejandro Q. Nato, Jr.",
"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\nBaron, Stickel, and Weeks present a well-written article describing the newly-developed R package (Mega2R) as an additional tool to the robust Mega2 data-reformatting software. Mega2R helps alleviate problems in handling huge genetic data by allowing the user to specify regions of the genome, genes, or markers that will be included in the analysis. It keeps the genotype data in a compressed format when reading from the Mega2 SQLite database to efficiently overcome the memory restrictions of R. It includes several functions that allow a user to carry out additional association tests, outputs Mega2R data as a VCF file with other related files, converts data frames into GenABEL-compatible format, and enhances GenABEL by supporting additional input data formats. Mega2R is indeed a useful R package for genetic analysis.\n\nI enjoyed reading the manuscript since the authors have clearly described what a user can do through Mega2R.\nI have a few minor comments and suggestions below:\nAdd a schematic diagram of Mega2R showing the different functions and how they interact with the SQLite database. This will visually help readers understand this article. For future enhancements, would it be possible to have a Mega2pedgene function that allows the user to choose between using reference/alternate alleles instead of major/minor alleles? This may allow Mega2R to facilitate analysis involving datasets from different populations. Is the chunk of 1,000 markers at a time in building the VCF file a fixed value? Or will you allow the user to select a particular number in the future? Indicate the format of the genotype matrix, e.g., normal format (markers in columns and individuals in rows), based on your examples. Will Mega2R be back compatible or will it only be compatible with R 3.5.0 onwards? Typographical error (retruns)\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": "4377",
"date": "25 Feb 2019",
"name": "Daniel Weeks",
"role": "Author Response",
"response": "Thank you for your thoughtful comments and suggestions for strengthening our paper. In Version 2, as suggested, we have added two Figures, in the 'Implementation' section, which now provide graphically illustrate how Mega2 and Mega2R work together (Figure 1) and how Mega2R provides an efficient and flexible wrapper for iterating through gene regions (Figure 2). However, this does not directly address your request for a diagram showing how the different Mega2R functions interact with the SQLite database. We did not add such a diagram because the interaction is minimal – at the beginning, we load the SQLite Mega2 databases directly into R as data frames for further analysis and manipulation. After this step, the functions use the loaded data frames. Regarding your numbered comments, we respond here:1. Using reference/alternate alleles within Mega2R instead of major/minor alleles is not yet possible, but this would definitely be a useful future enhancement. 2. Currently Mega2R's chunk size is defined by an internal parameter, so altering the chunk size could be accomplished by altering the source code. Making this a parameter would also be a useful future enhancement. 3. The genotype matrix has individuals in rows and markers in columns. This was indicated in the paragraph that first introduces the getgenotypes and getgenotypesraw functions: \"The functions getgenotypes and getgenotypesraw return a matrix of nucleotide pairs or a matrix of encoded integers with a column for each marker and containing a row per sample.\" As suggested, in this revision we have added a comment to the 'show' function clarifying this later in the manuscript. 4. Due to constraints encountered when submitted the Mega2R package to CRAN, Mega2R requires an R version of 3.5.0 or higher. As version 3.5.0 was released some time ago, in April 2018, hopefully this version requirement is not a burden. 5. The typographical error (retruns) has been corrected."
}
]
},
{
"id": "37773",
"date": "09 Oct 2018",
"name": "Derek Gordon",
"expertise": [
"Reviewer Expertise applied biostatistics",
"statistical genetics"
],
"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\nFor those who perform computational analysis in the fields of statistical genetics, bioinformatics, or population genetics, it is no exaggeration to say that the lion’s share of one’s time is spent formatting data so that it can be input into a specific analysis program. Having worked in statistical genetics for over two decades, this reviewer recalls a time when commonly used linkage programs such as LINKAGE, GENEHUNTER, S.A.G.E., LIPED, SIMWALK2, and others, each had different data input formats. For those format conversion programs that existed, they worked for one or a small number of other programs (e.g., conversion from LINKAGE format to S.A.G.E. format). Also, the programs were usually lab-specific, in that they were not publically available, or if they were, there was no “accountability”; that is, there was no error checking for the code, nor was there a way to send “bug reports” to the programmer.\nThat all changed when the first version of Mega2 was released. It was the “Rosetta Stone” for researchers who performed computational analyses. Not only did it allow for one to convert from one format to another seamlessly, it performed data error checking, created log files with each run of the program, and critically, it created Unix/Linux shell scripts that greatly simplified analysis runs. The importance of data storage and format conversion has only grown over time. Like the adage states, the only constant in life is change. The genomic data employed for linkage, association, and other computational analyses has grown in complexity and storage requirements. A list of genomic data employed in computational studies, from an approximate chronological standpoint, contains: di-allelic RFLPs, multi-allelic VNTRs, di-allelic SNPs, quantitative copy number variants (CNVs), next generation sequence, and RNA-seq data. As an example of the complexity of statistical methods to analyze these data, a PubMed search using the key terms “next generation sequencing”, “statistical method” and “software” yields over 350 hits.\nThe practice of program-specific formatting continues. Programs like PLINK, MACH, and METAL have their own input formats. While some of the input files (like the pedigree files) may be similar, other files like those that contain chromosome and base-pair position for a given item (e.g., SNP), may be different.\nGiven the number of statistical methods being employed, that they will most likely have different formats, that error checking is critical, and that data storage for the volumes of data now being generated must be efficient, it is critical that Mega2R be recognized. Mega2 has enjoyed a long-standing tradition as the “gold-standard” program for data format conversion. Because Dr. Weeks has directed the updating and stability of Mega2 in its various versions, and continues to do so with Mega2R, one can reasonably expect that Mega2R will be widely used, widely cited, and most especially will make the life of computational geneticists much easier.\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/7-1352
|
https://f1000research.com/articles/8-215/v1
|
25 Feb 19
|
{
"type": "Case Report",
"title": "Case Report: Sarcoidosis mimicking head and neck cancer progression",
"authors": [
"Edgar Pratas",
"João Carvalho",
"Isabel Domingues",
"Sara Pinheiro",
"Susana Amaral",
"Leila Khouri",
"Miguel Costa",
"José Eufrásio",
"Isonda Pires",
"Michael Davies",
"Rita Garcia",
"Margarida Teixeira",
"João Carvalho",
"Isabel Domingues",
"Sara Pinheiro",
"Susana Amaral",
"Leila Khouri",
"Miguel Costa",
"José Eufrásio",
"Isonda Pires",
"Michael Davies",
"Rita Garcia",
"Margarida Teixeira"
],
"abstract": "Several case reports have been published describing the coexistence of sarcoidosis and cancer. In the literature, simultaneous occurrence of head and neck cancer and sarcoidosis is rarely reported. In this paper we present a case of a 42-year-old man with squamous cell carcinoma of the oral cavity, locally advanced, which after surgery and adjuvant radiotherapy developed local persistence and progression in the mediastinal lymph nodes. The patient was submitted to chemotherapy and after a complete response, new suspicious mediastinal and hilar lymph nodes appeared in the thoracic computed tomography (CT) scan and 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) scan. To enroll the patient in a clinical trial, the patient underwent mediastinoscopy with mediastinal lymph node dissection. The histopathological findings were consistent with sarcoidosis and no metastatic disease was found. Since the patient had no symptoms and the levels of serum angiotensin converting enzyme were normal, no further pharmacological intervention was done. After 4 years of follow up the patient remains without evidence of cancer. This case shows that although imagological techniques (CT and FDG-PET scan) are extensively used to assess the tumor response, false-positive cases can occur. Whenever it is possible a biopsy of the suspected metastatic site should always be performed.",
"keywords": [
"Head Neck Cancer",
"Sarcoidosis",
"Lymph Nodes"
],
"content": "Introduction\n\nSarcoidosis is a systemic disease of unknown etiology that is characterized by the development of noncaseating epithelioid granulomas in various organs, mainly the lungs and lymphatic system1.\n\nSeveral cases have been published describing the coexistence of sarcoidosis and cancer. It has been reported at diagnosis, during treatment, and in the surveillance of cancer patients2,3. In head and neck cancer, there are few cases reporting the simultaneous occurrence of sarcoidosis4,5.\n\nWe present a case of a patient with a squamous cell carcinoma of the right alveolar ridge of the mandible, that during treatment developed mediastinal lymphadenopathies, causing a diagnostic and therapeutic dilemma between disease progression and benign lesions.\n\n\nCase report\n\nA 42-year-old male, Caucasian, presented to his stomatologist in May 2013 with a painless lump in the right jaw with 3 months of evolution. His Eastern Cooperative Oncology Group Performance Status was 0. The patient was an active smoker (24 pack-year) and denied drinking alcohol. He didn´t have other comorbidities. Oral cavity inspection revealed an ulcero-infiltrative lesion on the gingival margin of tooth 45 with extension to the tongue. In the lymph node assessment, an enlarged right submandibular node was palpable. The remainder of the physical examination was unremarkable. He had a normal complete blood count and biochemical profile. A biopsy of the lesion was performed and revealed a squamous cell carcinoma of the oral cavity. The patient was then referred to our institution.\n\nA cervical and thoracic computed tomography (CT) scan revealed a lesion in the right alveolar ridge of the mandible with bone reabsorption, and multiple bilateral enlarged cervical lymph nodes with high contrast enhancement (Figure 1). No other anatomical changes were reported. The tumor was clinically classified in T4 N2c M0.\n\nThe case was discussed at the Head and Neck multidisciplinary team meeting (MDT) and a surgical approach was decided.\n\nIn August 2013, the patient was submitted to a partial glossectomy, right hemimandibulectomy and bilateral neck dissection. Histological analysis confirmed a squamous cell carcinoma of the oral cavity with cervical lymph node metastasis in 2 of 18 lymph nodes of the right lymph node dissection; the 12 lymph nodes dissected from the left side of the neck were clear from metastasis. The surgical margins were negative although the tumor was 1mm from the deep margin (pT4 N2b M0).\n\nDue to the presence of positive lymph nodes, high T stage and close margins, the patient underwent adjuvant radiation therapy (60 Gy/30 fr over 6 weeks) from October to November. During the treatment he developed a grade 3 mucositis requiring a nasogastric feeding tube.\n\nIn January 2014, the cervical and thoracic CT scan revealed local and distant relapse with an area of heterogenous contrast enhancement in the right side of the floor of mouth and tongue as well as presence of mediastinal lymphadenopathies (Figure 2). Due to this fact the patient started palliative chemotherapy with the EXTREME regimen: Cisplatin (at a dose of 100 mg/m2) and cetuximab [at a dose of 400 mg/m2 initially (loading dose), then 250 mg/m2 on day 1 and 5-FU (at a dose of 1000 mg/m2 per day for 4 days)] every three weeks. As major toxicities he developed a grade 3 skin rash and grade 2 hypomagnesemia.\n\n(A) Area of heterogenous contrast enhancement in right side of the floor of mouth and tongue as well as area of peripheral contrast enhancement with central necrosis (21,7×8 mm); (B) and (C) presence of enlarged mediastinal lymph nodes (16,78mm, 10,34mm and 11,98mm).\n\nAfter 6 cycles with clinical and imagological complete response, Cetuximab was maintained in monotherapy.\n\nIn November 2014, the control CT scan revealed tumor progression in mediastinal lymph nodes without local tumor relapse. An 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) scan was obtained, showing hypermetabolic mediastinal and hilar lymph nodes (Figure 3).\n\n(A) Enlarged lymph node (12,8 mm) visible in the CT scan with correspondent abnormal in FDG uptake in the PET (B); (C) Focus of increased uptake in multiple mediastinal and hilar lymph nodes.\n\nThe Head and Neck MDT planned to enroll him in a clinical trial with immunotherapy. To meet the inclusion criteria, it was necessary to confirm histologically the mediastinal lymph node metastasis. Therefore, the patient underwent mediastinoscopy with mediastinal lymph node dissection in March 2015. Histopathological findings of the surgical specimens revealed granulomatous inflammation consistent with sarcoidosis, no metastatic disease was identified in the lymph nodes and special stains for fungi and acid-fast bacilli were negative.\n\nThe patient was referred to the Pneumology Department. Since he had no symptoms and the levels of serum angiotensin converting enzyme were normal no further pharmacological intervention was done.\n\nIn September 2017 he was submitted to a Facial Reconstructive Surgery and until now, the patient remains in follow-up without signs of local recurrence or metastasis.\n\n\nDiscussion\n\nClinical and imagiological evaluation (CT and FDG-PET scan) are widely used to evaluate tumor response. Positive CT and FDG-PET findings in patients with suspected cancer recurrence are often used to guide therapeutic approach6. However false-positives results can occur7.\n\nWe reported a rare case of a patient with a locally advanced squamous cell carcinoma of the right alveolar ridge of the mandible, that during the systemic treatment developed sarcoidosis of mediastinal lymph nodes. In our clinical case the diagnosis of sarcoidosis was incidental. The positivity in the CT scan and FDG-PET scan led us to the assumption of tumor progression. At that time, we considered the enrollment of the patient in a clinical trial because the therapeutic options were limited. The obligation of a histological confirmation of recurrent disease (inclusion criteria) led to the diagnosis of sarcoidosis and exclusion of cancer recurrence, providing a major shift in our therapeutic approach. Since then the patient remains exclusively in follow-up.\n\nIn conclusion sarcoidosis is an uncommon, but critical disease in the setting of Head and Neck cancer because it might mimic metastatic disease as the literature8 and our case shows. In order to avoid incorrect treatment decisions a biopsy of the suspected metastatic site should always be performed for confirmatory diagnosis.\n\n\nConsent\n\nInformed written consent for the publication of clinical details and 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": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nValeyre D, Prasse A, Nunes H, et al.: Sarcoidosis. Lancet. 2014; 383(9923): 1155–1167. PubMed Abstract | Publisher Full Text\n\nShu X, Ji J, Sundquist K, et al.: Survival in cancer patients with previous hospitalization for sarcoidosis: a Swedish population-based cohort study during 1964-2006. Ann Oncol. 2011; 22(6): 1427–34. PubMed Abstract | Publisher Full Text\n\nSpiekermann C, Kuhlencord M, Huss S, et al.: Coexistence of sarcoidosis and metastatic lesions: A diagnostic and therapeutic dilemma. Oncol Lett. 2017; 14(6): 7643–7652. PubMed Abstract | Publisher Full Text | Free Full Text\n\nArana Yi C, McCue P, Rosen M, et al.: Sarcoidosis mimicking metastatic bone disease in head and neck cancer. Semin Oncol. 2013; 40(5): 529–534. PubMed Abstract | Publisher Full Text\n\nYao M, Funk GF, Goldstein DP, et al.: Benign lesions in cancer patients: Case 1. Sarcoidosis after chemoradiation for head and neck cancer. J Clin Oncol. 2005; 23(3): 640–641. PubMed Abstract | Publisher Full Text\n\nJuweid ME, Cheson BD: Positron-emission tomography and assessment of cancer therapy. N Engl J Med. 2006; 354(5): 496–507. PubMed Abstract | Publisher Full Text\n\nCheung MK, Ong SY, Goyal U, et al.: False Positive Positron Emission Tomography / Computed Tomography Scans in Treated Head and Neck Cancers. Cureus. 2017; 9(4): e1146. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaruyama T, Saio M, Arasaki A, et al.: Sarcoidosis of mediastinal lymph nodes mimicking distant metastasis of oral squamous cell carcinoma: a case report and review of literature. Int J Clin Exp Med. 2018; 11(3): 2698–2708. Reference Source"
}
|
[
{
"id": "44879",
"date": "04 Mar 2019",
"name": "Robin J.D. Prestwich",
"expertise": [
"Reviewer Expertise Radiotherapy and chemotherapy",
"head and neck 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\nThis is a useful case report of a patient with locally advanced oral cavity carcinoma treated with surgery and adjuvant radiotherapy. CT soon after completion of treatment was suggestive of local recurrence in floor of mouth and development of new mediastinal LN. A complete radiological response was obtained following palliative chemotherapy. It would be useful to know what imaging modality this was and to explicitly state whether the local recurrence also showed a complete response.\nA subsequent PET-CT showed avid mediastinal LN strongly suggestive of distant metastases. Would be useful to comment on whether there was any uptake at primary site. Biopsy showed sarcoidosis and patient remains well several years later. It would be interesting to know the explanation for apparent local recurrence which resolved with chemotherapy - was this thought to be misinterpretation of radiology or also a sarcoid process?\nOverall a useful message regarding the importance of confirming recurrence by biopsy.\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": "54554",
"date": "03 Oct 2019",
"name": "Christoph Spiekermann",
"expertise": [
"Reviewer Expertise Immunology",
"Biomarker",
"Calprotectin",
"Tonsil",
"HNSCC",
"Sarcoidosis",
"Quality of Life",
"PROM 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\nPratas et al. describe the case of a patient with a HNSCC localized in the oral cavity. A CT scan after surgery and adjuvant radiation therapy reveals lesions with contrast enhancement in the oral cavity and mediastinal LN suspicious for tumor progression. Hence, a palliative chemotherapy was added resulting in complete response. It would be interesting to discuss whether these findings could be explained by sarcoid like lesions or sarcoidosis, as well.\nA control CT and PET-CT scan, then, revealed mediastinal LN suggestive for metastatic lesions. Biopsy of the LN leads to the diagnosis of sarcoidosis and no further therapy was conducted. Although histopathological findings show sarcoidosis, a residual risk of overlooking metastases in systemic inflammatory diseases still remains existent because it is not possible to take biopsies of every hypermetabolic lesion. This would be an interesting addition to the discussion section of the manuscript.\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": []
}
] | 1
|
https://f1000research.com/articles/8-215
|
https://f1000research.com/articles/7-1732/v1
|
01 Nov 18
|
{
"type": "Review",
"title": "Research and development of new tuberculosis vaccines: a review",
"authors": [
"Lewis K. Schrager",
"Rebecca C. Harris",
"Johan Vekemans",
"Lewis K. Schrager",
"Rebecca C. Harris"
],
"abstract": "Tuberculosis kills more people worldwide than any other single infectious disease agent, a threat made more dire by the spread of drug-resistant strains of Mycobacterium tuberculosis (Mtb). Development of new vaccines capable of preventing TB disease and new Mtb infection are an essential component of the strategy to combat the TB epidemic. Accordingly, the WHO considers the development of new TB vaccines a major public health priority. In October 2017, the WHO convened a consultation with global leaders in the TB vaccine development field to emphasize the WHO commitment to this effort and to facilitate creative approaches to the discovery and development of TB vaccine candidates. This review summarizes the presentations at this consultation, updated with scientific literature references, and includes discussions of the public health need for a TB vaccine; the status of efforts to develop vaccines to replace or potentiate BCG in infants and develop new TB vaccines for adolescents and adults; strategies being employed to diversify vaccine platforms; and new animal models being developed to facilitate TB vaccine development. A perspective on the status of these efforts from the major funders and organizational contributors also is included. This presentation highlights the extraordinary progress being made to develop new TB vaccines and provided a clear picture of the exciting development pathways that are being explored.",
"keywords": [
"Tuberculosis",
"Mycobacterium tuberculosis",
"vaccines",
"immunization"
],
"content": "Table of Contents\n\n1. Public health need for new tuberculosis vaccines and landscape analysis\n\na. Introduction\n\nb. The WHO Global TB Programme\n\nc. The Value Proposition for TB Vaccines\n\nd. Overview of the tuberculosis vaccine clinical pipeline\n\ne. Modeling impact according to vaccine profile\n\n2. Vaccines for BCG replacement\n\na. The BCG vaccine’s profile and role in public health\n\nb. VPM1002\n\nc. MTBVAC\n\n3. Vaccines for adolescents and adults (phase 2 and beyond)\n\na. VPM1002\n\nb. H56:IC31\n\nc. ID93/GLA-SE\n\nd. RUTI™ for Adjunctive Immunotherapy\n\ne. Vaccae™\n\n4. Diversifying TB vaccine platforms\n\na. Importance of immune characterization\n\nb. Cytomegalovirus (CMV) recombinant vaccines against TB\n\nc. Novel emerging vaccine technologies: mRNA-based vaccines\n\nd. New candidates for live attenuated TB vaccines\n\ne. A systematic antigen discovery approach in humans\n\nf. Antibody-generating vaccines for TB\n\ng. Diverse T-Cell responses against Mtb\n\nh. Immune correlates and signatures of protection\n\n5. Use of models in translational TB vaccine research\n\na. Small animal models\n\nb. The role of a continuous exposure natural transmission animal model for vaccine development\n\nc. Simian models of TB\n\nd. Achieving sterilizing immunity in animal challenge models\n\ne. Mucosal immunisation against TB\n\nf. Alternative routes of BCG immunization\n\ng. Progress in clinical use of a controlled human infection model (CHIM) for vaccine testing\n\n6. Perspectives from funders and global health organizations working on TB vaccines\n\na. Perspectives from the Bill and Melinda Gates Foundation\n\nb. The Global TB Vaccine Partnership\n\nc. Aeras\n\nd. European and Developing Countries Clinical Trials Partnership\n\ne. The Stop TB Partnership, working group on new TB vaccines\n\nf. The Tuberculosis Vaccine Initiative\n\ng. Proposed stage gate criteria for TB vaccines\n\n7. Conclusions\n\n\n1. Public health need for new tuberculosis vaccines and landscape analysis\n\nThe World Health Organization (WHO) provides guidance on priority targets and development pathways for vaccines against diseases of high public health interest. The involvement of the WHO in vaccine development is driven by assessments of medical need and technical feasibility, and an absence of market incentives and sufficient funding to adequately drive the development process. As tuberculosis (TB) kills more people globally than any other single infectious agent1, developing a TB vaccine ranks among the highest global health priorities from a medical needs perspective. A number of factors support the technical feasibility of a TB vaccine, including the fact that an estimated 90% of persons infected with Mycobacterium tuberculosis (Mtb) do not progress to active disease1, evidence suggesting that past Mtb infection provides some protection against new infections2,3, and the existence of a century-old vaccine, Bacillus Calmette-Guérin (BCG), that provides partial protection in children which may, under some circumstances, extend for decades4. The current imperative is to improve our understanding of the type of immunological responses needed to provide robust protection against Mtb infection or TB disease, and to use this information to efficiently develop new, safe and effective TB vaccines.\n\nIn response to the compelling public health need for a TB vaccine, in October 2017 the WHO convened experts and representatives of important stakeholder institutions involved in vaccine development to assist the WHO in a two-pronged effort to accelerate TB vaccine development. The first effort was dedicated to defining a preferred product characteristics (PPC) guidance document for new TB vaccines, now publicly available5. The second was to convene a meeting providing an opportunity for the exchange of cutting-edge information relevant to TB vaccine development, from basic science to later stage research. The proceedings of this meeting are summarized in this document.\n\nEach year, the WHO publishes the Global Tuberculosis Report, documenting the overall status of the global TB epidemic. The 2017 version of this report highlights the following statistics for the state of the TB epidemic in 20161:\n\n1,674,000 TB deaths, of which 374,000 were in HIV co-infected persons (“the ninth leading cause of death worldwide and the leading cause from a single infectious agent, ranking above HIV/AIDS”1); 374,000 of these deaths occurring in HIV-infected persons\n\n10.4 million cases of TB disease, 56% of which were in 5 countries: India, Indonesia, China, the Philippines and Pakistan;\n\n600,000 cases of rifampicin-resistant TB (RR-TB), of which 490,000 were multidrug-resistant TB (MDR-TB). Globally, an estimated 4.1% of new cases and 19% of previously treated cases had MDR or RR-TB, with 47% of the global total of drug-resistant cases reported from China, India and the Russian Federation. 250,000 of these 490,000 died.\n\nA U.S. $2.3 billion funding gap for financing TB care and prevention existed in 2017\n\nThe WHO strategy to tackle the global epidemic of TB was set forth in the WHO End TB Strategy, endorsed by the WHO General Assembly in 20146. The End TB Strategy relies on a number of priorities in the global response to the TB epidemic. These priorities include improving approaches to identifying cases of TB, addressing MDR-TB as a true public health crisis, and accelerating strategies to better prevent, identify and treat TB in HIV-infected individuals. A three-pillared approach to the TB epidemic is being implemented: 1) to advocate for integrated, patient-centered TB care and prevention; 2) to support the creation of bold policies and supportive systems to curb the spread of TB; and 3) to support intensive research and innovation to create the new tools needed to prevent and treat control TB, including better diagnostics and drugs, and a new vaccine capable of preventing TB disease.\n\nA key objective of the WHO End TB Strategy is a 90% reduction in TB incidence, from the 2015 annual incidence of greater than 100 cases per 100,000 population to less than 10 cases per 100,000 persons by 20356. Currently, the annual decline in TB incidence is not nearly fast enough to meet the expressed goals1. As evidenced by the successes of the social protections provided under the Marshall Plan in the aftermath of World War II, a 10% rate of decline could be considered feasible through optimization of currently available social and medical interventions. Though this is a major proposition in and of itself, which would require a heightened degree of global commitment to this effort. Without the development of new TB prevention modalities, however, this rate would be expected to flatten out to a 5% decline per year by 2025, again, not nearly sufficient to meet the End TB Strategy 2035 goals6. Only the introduction of new tools, including better point-of-care diagnostics capable of quickly and reliably diagnosing TB, including cases of drug-resistant (DR) TB; better drugs for treating drug-sensitive and DR-TB over a shorter period of time than is currently required; and, most importantly, a new vaccine capable of preventing TB disease, introduced by 2025, would result in an acceleration to the 17% per year decline of TB incidence necessary to meet the 2035 goals6.\n\nCurrently, a 2.3 billion U.S dollar (USD) annual research gap exists in efforts to develop new tools to control TB, including better diagnostics and drugs, as well as TB vaccines1. Given the ambitious targets set by the WHO End TB Strategy, the $2.3 billion gap represents a conservative estimate, particularly considering the absolute need to develop an effective vaccine by 2025 for the targets to be reached. The ongoing United Nations (UN)–led efforts to prioritize action against TB is illustrated by a series of high-level political meetings. Top-level political engagement was demonstrated in Russia in November 2017 at the gathering of 194 ministers of health, research and finance, alongside the WHO Director General and the President of the UN General Assembly. This was followed by an interactive civil society meeting, attended by more than 250 representatives from civil society, academia, NGOs, and the private sector, setting the stage for the first ever UN General Assembly high level meeting (HLM) on TB in New York in September 2018. The HLM aims to deliver an ambitious political declaration on TB endorsed by Heads of State, which will strengthen action and investments to end TB. Issues to be discussed include the increase in drug-resistant TB strains, the need to provide adequate resources to control the epidemic, and to create a universal monitoring program to track the epidemic and the world’s response to it, will be highlighted. Accelerating development of a vaccine for TB will be a critical component in the future of TB prevention and care, therefore urgently requires increased resource allocation and political support in upcoming years.\n\nA key question that arises when attempting to assess the value of TB vaccines is how to measure “value”. Possible measures of value include the payer perspective – savings in direct healthcare costs resulting from the protective effect of vaccines; a human perspective – assessing reductions in death and disability; and a societal perspective – determining the overall benefits to society that result from vaccine-induced protection. Different stakeholders will have different views on which of these approaches best defines the value of vaccines. Perspectives of vaccine value also will differ based on whether the value proposition involves a regional, countrywide, or global assessment.\n\nA model of the health impact of the introduction in 2024 in low and middle income countries of a 60% efficacious vaccine with 10 years protection delivered to adolescents and adults, with coverage equal to school attendance for routine vaccination of 10 year olds and 72–76% for periodic mass campaigns, suggests that approximately 17 million TB cases could be averted between 2024 and 20507. If such vaccine were to be administered to infants, with approximately 90% coverage rate, 890,000 TB cases would be cumulatively averted by 2050. These benefits need to be weighed against the total systems cost of vaccine administration, including the purchase price and the cost of administration, transportation and storage, national program management and the facilities and equipment necessary to deliver and administer a vaccine. In the modelled scenarios above, the median cost per disability-adjusted life year (DALY) averted in low or middle income settings ranged from cost saving to $720 per DALY averted in the adult/adolescent scenario, and $1,690–$18,110 per DALY averted in the infant scenario7. Issues faced in the early stages of vaccine development, such as efficacy and duration of protection, and issues arising later, such as vaccine price, regimen of administration, thermostability and availability (overall supply) all contribute to the cost of vaccine administration and should be taken into consideration at the very earliest stages of the development of a vaccine.\n\nCurrently, multiple vaccine candidates are being assessed in clinical trials. Vaccine strategies being assessed include live, attenuated mycobacteria (MTBVAC, VPM-1002); killed, whole cell mycobacteria (DAR-901, M. vaccae, MIP), mycobacterial extracts (RUTI); adjuvanted protein vaccines (M72/AS01E, H4:IC31 (discontinued), H56:IC31, ID93 + GLA-SE), and viral vectored vaccines (Ad5Ag85A, ChAdOx185A/MVA85A, TB/FLU-04L). Although the target populations in these trials include adolescents, adults and infants, it is generally recognized that vaccinating adolescents and adults represents the most efficient means of stopping the cycle of Mtb transmission and thereby offers the best opportunity to control the global spread of Mtb, even to infants and children7,8.\n\nFour potential indications have been identified as targets for TB vaccine development in adolescents and adults: 1) prevention of TB disease (PoD); 2) prevention of recurrent TB disease, which can be related to relapse or re-infection, in persons cured post-treatment for active disease (PoR); 3) prevention of established Mtb infection in persons uninfected with Mtb (PoI); and 4) as an immunotherapeutic adjunct to drug treatment to shorten curative regimens or increase the efficacy of treatment of drug-resistant strains. In infants, the major goals have been to develop a vaccine that is more efficacious and/or safer than BCG that could serve as a BCG replacement and to develop a booster to improve and extend the protection provided by BCG.\n\nPoD represents the highest priority indication, as modeling suggests that preventing TB disease in adolescents and adults would be the quickest way to control the global TB epidemic7,8. The major issue in developing a vaccine for a PoD indication, however, is the cost of conducting late-stage efficacy trials given the relatively large sample sizes and prolonged period of post-vaccination follow-up required to accrue adequate endpoints in the general, healthy population. In light of this, novel trial designs for phase 2 proof of concept studies have been developed to reduce the risk of failure in expensive late-stage trials. These novel designs have focused on assessing candidate TB vaccines for clinically meaningful biological effects in selected high-risk populations, to allow shorter, less expensive trials for assessment of efficacy.\n\nOne such novel design is based on prevention of recurrent TB disease in recently cured TB patients since this population is at an elevated risk of developing TB compared to individuals that have never had TB, due either to reinfection or to reactivation of existing infection. While a PoR endpoint could serve as a licensable indication in its own right, it currently is also being assessed as a possible means of de-risking a decision to move a vaccine candidate forward into late-stage PoD evaluations in the general population9,10.\n\nAssessing the ability of a candidate vaccine to prevent establishment of de novo Mtb infection represents another strategy to demonstrate a clinically meaningful biological effect. If populations experiencing a high force of Mtb infection were recruited, such as adolescents in the Western Cape of South Africa, a PoI endpoint trial could potentially be less expensive to conduct than a PoR trial11. Results from a PoI trial of the adjuvanted protein vaccine, H4:IC31 versus BCG revaccination, conducted in adolescents in the Western Cape were recently published (after the meeting reported here). This trial demonstrated that neither the H4:IC31 vaccine nor revaccination with the BCG vaccine prevented initial quantiferon (QFT) conversion (efficacy point estimates of 9.4%, P=0.63 for H4:IC31; 20.1%, P=0.29 for BCG). BCG revaccination, however, reduced the rate of sustained QFT conversion (efficacy 45.4%, P=0.03); efficacy of sustained QFT conversion for H4:IC31 was 30.5%, P=0.16)12. The implications of these results for BCG revaccination strategies are currently being discussed. A recent review of available data on BCG revaccination before these data were released led WHO to conclude that the existing evidence, prior to the completion of the BCG revaccination PoI trial did not support BCG revaccination policy13.\n\nUnlike a PoR endpoint, however, it is unclear whether a PoI endpoint could serve as an indication for licensing for two reasons: 1) only approximately 10% of Mtb-infected individuals ultimately develop active TB, so it is theoretically possible that infection prevention, should it be demonstrated, is only occurring in persons who would otherwise control their Mtb even without vaccination and not develop TB disease; and 2) the prevention of infection endpoint depends on assessments such as the tuberculin skin test (TST) or interferon-gamma release assay (IGRA) whose specificity and sensitivity are not 100%.\n\nBoth the PoI and PoR approaches pose a risk when used as experimental endpoints towards the development of a PoD vaccine, as it remains unknown whether these other endpoints actually predict efficacy for PoD in the general population. It is possible that preventing disease, preventing infection and preventing recurrence of disease in cured patients may require fundamentally different immunological responses, thereby leading to positive or negative results that would not be predictive of a positive PoD outcome.\n\nThe current moment represents a critical juncture for TB vaccine development. In 2018–2020, clinical efficacy data are scheduled to be released from five phase 2b proof of concept studies: Vaccae, H4/BCG revaccination (data subsequently released and described above), M72/AS01 (data subsequently released and described below), Dar901 and VPM1002. Results from these trials will paint a clearer picture as to the potential of the global vaccine candidate pipeline, optimal paths for future TB vaccine design and development strategies, and the predictive ability of current animal models. Ideally, candidates that demonstrate efficacy will also enable discovery of candidate correlates of protection, with potential to substantially streamline further TB vaccine development.\n\nMany challenges must be addressed to develop a successful TB vaccine. Mtb represents a complex pathogen that has evolved with humanity for tens of thousands of years, allowing it to evolve in such a way as to survive the normal range of immune responses directed against it. The absence of a known correlate of immune protection against TB in humans has hampered vaccine discovery and development, as has the inability to assess the predictive nature of animal challenge models given the lack of vaccine efficacy data in humans to date. Late stage efficacy trials, particularly phase 3 trials for efficacy and safety, are long and expensive, as are subsequent phase 4 programs, together representing a potential disincentive for biotechnology and pharmaceutical companies to expend resources given the relatively high risk of failure and expectations of limited commercial returns. Moreover, the TB vaccine field remains severely underfunded, particularly given the potential impact of a vaccine in curbing the TB epidemic and the global societal cost of TB, estimated at $19.2 billion in 2017, including a $7.2 billion cost for treatment and control activities and $12 billion indirect costs to the economies1.\n\nDespite these challenges, however, it is becoming clear that developing new TB vaccines is an achievable goal. The partially protective effect of BCG, capable of preventing up to 80% of cases of severe, disseminated TB in infants and young children, provides important supportive evidence of this point. Additional evidence comes from the 90% of Mtb-infected persons who are capable of controlling their infection throughout their lives without developing disease; and the intriguing examples of individuals at high risk of Mtb infection who, when assessed for Mtb infection by TST or IGRA, demonstrate transient positivity on repeated assessments before reverting to a consistently negative state, a situation with unclear clinical significance but which has been interpreted by some as evidence of early immunological containment of Mtb infection and a reduced risk of TB disease14–17. Examples of individuals with lifelong exposure to Mtb, through household or occupational contact, for example, who never convert their TSTs or IGRAs, may represent another example of natural immunity to Mtb that a vaccine could be able to replicate or improve upon18.\n\nAccordingly, there are reasons for optimism in TB vaccine development. New strategic directions include diversifying the pipeline to explore candidates that generate immunity beyond CD4+ Th1 T-cells, including non-conventional, cellular immunity and trained innate immunity19; assessing new routes of vaccine administration, including aerosol and, possibly, intravenous approaches; investigating the extent to which antibody generating vaccines may contribute to protection afforded by the current, cell-mediated immune generating candidates; and utilizing new tools for vaccine candidate R&D, including positron emission tomography-computerized tomography (PET-CT) scans20 and bar-coded Mtb strains21 permitting ultrasensitive assessments of animal challenge models, and a potential controlled human infection model (CHIM) for Mtb22,23. These new strategic directions in TB vaccine development add to the sense of optimism and excitement in the field.\n\nMathematical modeling techniques provide useful estimates of the future impact of vaccines in development. A review of twenty-four mathematical modeling studies applied to the case of TB vaccine development and implementation, provide insights relevant to decisions concerning the optimum age of vaccination, the most impactful vaccine indication, and the minimum duration of vaccine effect, efficacy levels and implementation strategy needed to have a meaningful impact on the TB epidemic8.\n\nMathematical models suggest that, in low- and middle-income countries overall (LMICs), targeting adolescents and adults for TB vaccination campaigns will result in a greater and faster impact on the TB epidemic before 2050, than would targeting infants7. In aging, reactivation driven, epidemics like China, a high proportion of TB disease occurs in older adults; accordingly, targeting adolescents and younger adults would not be expected to have an important effect in controlling TB before 2050, as compared to a vaccine with administration targeted to older adults.\n\nVaccinating infants with an improved TB vaccine would represent an important public health benefit over the long term, but would not have as great an impact before 2050 as would an adolescent/adult vaccine, mainly because infants and young children are not an important source of Mtb spread. Accordingly, the public health impact of an infant vaccine would be delayed until adolescence, when the burden of Mtb infection, and the possibility of spreading Mtb to others, increases. Moreover, an infant vaccine would need to demonstrate higher efficacy and/or longer durations or protection than a vaccine administered to adolescents and adults to have an important, and cost effective, effect on public health7. Mathematical modeling of the impact of vaccinating infants extends to the year 2050. Beyond that, there is too much uncertainty to draw conclusions with confidence.\n\nModeling suggests that, before 2050, a PoD vaccine would have a greater and faster effect on controlling Mtb spread than would a PoI vaccine8. Even if a 100% efficacious PoI vaccine were to be developed and made available by 2025, the target date for TB vaccine introduction set by the WHO End TB Strategy, a PoD vaccine also would need to be utilized if the 2035 End TB Strategy targets were to be met. The body of available mathematical modelling literature is at equipoise as to whether the effect of a PoD vaccine would be greater if administered to persons who are not yet infected with Mtb as compared to those who are latently infected, with three models predicting that a pre-infection vaccine would provide the greatest impact, while four suggest that a post-infection vaccine would have a greater effect on TB control. More recent modelling evidence, specifically exploring this issue suggests that in aging, reactivation driven epidemics like China, because over the modelled vaccination time frame most TB disease is predicted to occur among the elderly population, a PoD vaccine administered to persons latently infected with Mtb would have greatest and most rapid impact24.\n\nWhile a PoR vaccine would be expected to contribute to TB control efforts, no mathematical modeling of the public health effect specifically of a PoR vaccine as yet has been undertaken.\n\nMathematical modeling exercises reinforce the intuitive conclusion that a vaccine with greater efficacy and a longer duration of effect will have greater benefit than those with lesser degrees of efficacy and durations. In addition, modeling has suggested that a PoD vaccine with protective efficacy as low as 20%, and duration as short as 5 years, administered to adolescents and adults, could have an important and cost-effective effect in controlling the spread of Mtb in LMICs7. Vaccines with shorter duration of effectiveness could be compensated for by higher efficacy or performing more frequent mass vaccination campaigns, to the extent that resources and logistical realities might permit.\n\nIn summary, mathematical models of TB vaccine effect suggest that, for maximum impact before 2050, the optimal new TB vaccine would prevent Mtb disease (although in transmission-driven epidemics a vaccine preventing infection would be useful), would be targeted for use in adolescents and adults, and would have a duration of effect of at least five years. Modeling also emphasizes that optimal vaccine strategies to control TB may differ between countries based upon the segment of the population that is driving the epidemic and so should be tailored accordingly. Future modeling exercises are needed to explore the effect of a PoR vaccine, to get a clearer picture of optimum vaccine strategies in other high burden countries such as India and South Africa, to assess the impact and cost-effectiveness of vaccines on controlling the spread of multi-drug resistant TB (MDR-TB) and extensively drug resistant TB (XDR-TB), to gauge the impact of new vaccines in combination with the scale up of other interventions, and to assess the impact of TB vaccine administered as immunotherapeutic agents in persons being treated for active TB.\n\n\n2. Vaccines for BCG replacement\n\nOptimally, certain information should be known when attempting to develop a vaccine, such as the immune mechanisms necessary and sufficient to protect against infection and/or disease, the antigens that stimulate a protective immunological response, the antigen delivery platforms best capable of evoking a protective response in a safe and acceptable manner, and confirmation of the protective mechanisms in human clinical efficacy studies. Currently, none of this information is available for TB vaccines. Accordingly, additional information must be sought and utilized to justify and guide the TB vaccine development effort.\n\nThere is some skepticism about whether a truly effective vaccine can be developed against TB. We know that about 90% of people infected do not develop disease and that individuals who are immunocompromised have 10-fold greater risk, establishing the importance of protective immune responses. One interesting piece of information comes from an observational clinical study published in the mid-1930s2. In this study, nurses entering service in a TB hospital in Norway were assessed for their rate of TB based on their TST status prior to entering this service. This study revealed that nurses who were TST positive had a 96% reduced risk of developing clinical TB than did the nurses who were TST negative, with an annual TB rate of 4% to 5% observed in the TST-negative nurses, suggesting a meaningful degree of protection against TB disease stemming from latent TB infection (LTBI). A later analysis of 14 studies confirmed this result, with LTBI found to be 78% protective against TB disease, even though Mtb infection, in itself, poses a risk for developing TB3.\n\nThe bacille Calmette-Guérin (BCG) vaccine represents the only successful vaccine as yet developed for TB. BCG is derived from Mycobacterium bovis, a bovine mycobacterium passaged 230 times in the Pasteur laboratory in Lyon by Drs. Calmette and Guérin between 1908 and its first human use in 1921. It has since become the most widely used vaccine in human history, with well over 3 billion doses administered; currently, more than 100 million children annually are administered BCG within hours or days of birth, representing 89% of the annual global birth cohort. BCG provides excellent protection in infants against disseminated miliary and meningeal TB, particularly serious forms of the disease with high mortality rate (RR 0.1; 95% CI 0.01 - 0.77)25.\n\nWhile evidence of BCG-induced protection against miliary and meningeal TB in infants is widely accepted, the degree to which BCG prevents pulmonary TB in infants, children and adults remains controversial. In the past, some of the differing results has been attributed to the variety of BCG strains currently being used globally, yet in meta-analyses of clinical trials utilizing different BCG strains, no statistically significant difference in efficacy has been identified25. An observational study determined that LTBI provides better protection than BCG vaccination against developing pulmonary TB in adults2, while a meta-analysis of randomized clinical trials of BCG efficacy in children with known exposure to Mtb identified a protective effect against acquiring Mtb infection of 27% (RR 0.73, 95% CI 0.61 – 0.87) and a protective effect against developing active TB of 71% (RR 0.29; 95% CI 0.15 – 0.58)26.\n\nOverall, the results of studies of BCG effectiveness, beyond prevention of miliary and meningeal TB in infants, has been marked by extraordinary variability and inconsistency, with rates of BCG-induced protection varying from 80% to totally non-protective at any age group. A number of hypotheses have been advanced to explain this variation and relative lack of effectiveness. These include differences in key antigens between Mtb and BCG, including the absence in BCG of the important Mtb RD1 and ESX genetic regions; differences in the potency of the various BCG strains in use globally; and interference from exposure to atypical environmental mycobacteria. Mycobacterial exposure, generally in countries closer to the equator, has the potential to confound assessing the efficacy of BCG or any new vaccine candidate in populations whose immune systems have been engaged by cross-reacting mycobacterial antigens.\n\nAnother unknown about BCG is the extent to which the vaccine persists following infant vaccination. Anecdotal reports of two teenage boys in France who developed disseminated BCG following immune-suppressive treatment for cancer suggest the potential for long, latent duration, a factor that may influence the activity of new TB vaccines.\n\nBCG administration has changed since its first delivery to an infant whose mother had active TB in 1921. Initially, BCG was administered orally, where most organisms were presumed to be killed by stomach acid. Calmette, however, observed bacillemia following oral administration, leading later to a switch of BCG administration to the intradermal (ID) route. Experiments with intravenous (IV) administration of BCG to rhesus macaques in the early 1970s demonstrated significantly greater protection against pulmonary and hematogenous TB following aerosol Mtb challenge than in macaques receiving ID BCG27. Recently, there has been renewed interest in the possibility of greater effectiveness of BCG if administered in a way that permitted broader diffusion in the lungs – via inhalation of an aerosolized preparation of BCG28, or systemic dissemination such as would occur following IV administration29. Preclinical application of such approaches offers the opportunity to understand the immunological mechanisms that convey heightened protection in, for example, Mtb challenge models utilizing non-human primates such as rhesus macaques and to explore possible clinical applicability. While early, exploratory clinical trials of inhaled BCG have not identified safety concerns30, assuring that an IV route is sufficiently safe for humans is likely to prove challenging.\n\nWhile BCG generally is regarded as safe, hematogenous dissemination of 1% following ID administration has reported, mainly in infants with HIV infection and other forms of immunosuppression31. For this reason, BCG is contraindicated for use in immunosuppressed infants. The potential for developing serious complications, including death, following uncontrolled hematogenous dissemination when administered to immunosuppressed infants represents an important rationale driving the need for replacing BCG with a safer TB vaccine that can be administered to all infants, regardless of HIV infection status or immune competence.\n\nImproved understanding of Mtb and BCG pathophysiology has led to potential paths forward for improving BCG. As BCG is missing the RD1 and ESX genetic regions present in Mtb, one possible future approach is to replace them in genetically modified BCG strains. Additionally, it is known that Mtb is capable of escaping the macrophage endosome and thereby has its antigens processed in the cytosol, inducing cytotoxic T-lymphocyte responses, and that antimicrobial peptides derived from cytotoxic lymphocytes (CTL) are able to kill Mtb within infected macrophages in vitro. While BCG is incapable of escaping the endosomes or generating a significant CTL response, a BCG-derived TB vaccine, VPM1002, has been genetically manipulated to permit its escape from the endosome, resulting in cytosolic antigen processing32. The extent to which this alteration may improve BCG efficacy for use in both infants and adults is under active investigation.\n\nFinally, developers of vaccines targeted at replacing BCG in infants will need to take into consideration the effects of BCG unrelated to protection against TB. BCG vaccination has been demonstrated to provide protection against leprosy, and this constitutes one of the bases for BCG recommendation13. Studies of BCG-associated protection against Buruli ulcer disease, however, have demonstrated mixed results33,34. Improvement in all-cause mortality in LMICs has also been reported35. Additionally, BCG has been demonstrated to provide a 70% protection from superficial bladder cancer36. Ultimately, investments will need to be made to support clinical studies to fully evaluate all candidates intended to replace BCG, as well as any new TB vaccine.\n\nVPM1002 is a recombinant BCG (rBCG) vaccine being developed both as a replacement for BCG vaccination in infants and as a TB vaccine in adolescents and adults32. In VPM1002, a listeriolysin gene has been added to the BCG genome and a urease gene has been deleted, permitting the rBCG to escape the macrophage lysosome, which occurs in Mtb infection but not with BCG. Processing VPM1002 antigens in the cytosol results in activation of the AIM2 inflammasome, inducing autophagy of the VPM1002-infected macrophage and stimulating innate immunity in a manner not seen with BCG. The manufacturing process for VPM1002 also offers the prospect of avoiding the frequent, global shortages of BCG, as it is manufactured using fermentation media, with a 50-liter batch yielding approximately 5 million doses.\n\nThe safety and tolerability of VPM1002 has been assessed in a recently-concluded phase 2 trial of HIV-exposed and HIV-unexposed infants in South Africa; data from this trial is due for public release in Q3/2018. A phase 3 trial, comparing VPM1002 safety and efficacy to BCG in 10,000 South African infants is scheduled to begin imminently, with completion targeted for 2021.\n\nMTBVAC is the only attenuated vaccine derived from Mtb currently in clinical trials. To create this vaccine, two stable deletions were made in the genome of an Mtb clinical isolate, selected so as to avoid laboratory subculture: deletions of the phoP gene, needed to control the transcription of key Mtb virulence genes permitting its survival in host cells; and fadD26 gene, required for the synthesis of cell surface lipids that play a critical role in Mtb pathogenicity37. As MTBVAC is derived from a clinical Mtb strain, and only carries two specific deletions in its genome, the vaccine candidate possesses 519 more epitopes (1603 epitopes) than currently used BCG (1084 epitopes)38. Challenge experiments in mice demonstrated improved protection by MTBVAC as compared to BCG, with protection associated with a T-cell mediated response to two Mtb-specific antigens, CFP10 and ESAT6, present in MTBVAC but absent in BCG39.\n\nMTBVAC is primarily being developed to replace BCG as a priming immunization against TB. The rationale for targeting infants includes the need for an improved vaccine over BCG given the remaining burden of pediatric TB, and the opportunity to deliver vaccination to a population without prior sensitization to BCG, Mtb or environmental mycobacteria, thereby avoiding potential “masking” or “blocking” effects on MTBVAC-induced protection.\n\nA phase 1a study, conducted in BCG-unvaccinated adults living in an area not endemic for TB, demonstrated the safety and immunogenicity of MTBVAC40. MTBVAC currently is being further assessed for safety and immunogenicity in newborns in a phase 1b dose escalation study conducted in a TB-endemic region of South Africa, with a safety arm in adults. Vaccination of newborns enrolled in this study was completed in September 2016, with one-year follow-up in September 2017. A phase 2a dose-defining safety and immunogenicity study of MTBVAC in 99 HIV-unexposed newborns in TB-endemic regions of sub-Saharan Africa is being planned. This study is also intended to build capacity for future, later-stage evaluations of this vaccine starting in 2018.\n\nIn addition to use in infants, MTBVAC is being assessed for use in adolescents and adults previously vaccinated with BCG. Preclinical studies of revaccination of guinea pigs with MTBVAC have demonstrated improved protection compared to that conferred by BCG41. Initiation of a phase 1b-2a, double blind, randomized, BCG-controlled, dose-escalation safety and immunogenicity study in 120 healthy South African adults, ages 18–50 years, with and without LTBI is being planned for the latter half of 2018.\n\n\n3. Vaccines for adolescents and adults (phase 2 and beyond)\n\nIn addition to its development as a vaccine to replace BCG in infants, VPM1002 also is being developed as a TB vaccine for adolescents and adults. A phase 2b/phase 3, randomized, double-blind, placebo-controlled trial to assess VPM1002 vaccine efficacy in preventing recurrence of TB in adults recently treated and cured of active TB (PoR) is underway in India. Sample size calculations for this study assume a 5% rate of TB recurrence in the placebo arm and a 50% reduction of TB recurrence among subjects in the VPM1002 arm during a 12-month follow-up period. Two thousand persons (1,000 in each arm) will be enrolled to achieve 80% power at a 5% significance level, with a 10% estimated drop-out rate. A phase 3 study to protect against TB disease among household contacts of persons with active TB is planned to begin in late 2018.\n\nM72/AS01E, the GSK proprietary candidate vaccine, consists of a fusion protein expressing two immunogenic Mtb antigens: Mtb39A, a membrane-associated protein expressed early in the Mtb life cycle, putatively identified as an immune evasion factor; and Mtb32A, a constitutively expressed secreted protein and a putative serine protease, combined with the adjuvant system AS01E containing monophosphoryl lipid A (MPL) and QS21 in a liposomal suspension, geared to promote a Th1 immune response42. Key components of the target product profile (TPP) for M72/AS01E include preventing TB in adults and adolescents; an efficacy level of 70% or greater (no minimally acceptable level of efficacy has yet been set by the developers); a clinically acceptable safety profile, including safety for use in HIV-infected individuals; acceptability for co-administration with widely recommended adolescent vaccines when administered in a 2-dose regimen 1 to 6 months apart; and a duration of protection of 10 or more years. The possible need for boosters will be explored.\n\nThe M72/AS01E candidate vaccine was first introduced into humans in 2004; clinical development has been conducted in collaboration with Aeras. M72/AS01E has been tested for safety and immunogenicity in 12 completed phase 1 and phase 2 studies. The vaccine has been tested in adults who were PPD negative, PPD positive43, HIV negative44, HIV positive on antiretroviral therapy (ART)45, and HIV positive not receiving ART44. Adults during or after TB treatment were the focus of one phase 2 trial. Adult trials have been conducted in non-endemic and endemic TB settings. The vaccine has also been assessed in adolescents in South Africa and in infants in The Gambia46. Overall, M72/AS01E was found to be safe and well tolerated in the recruited populations, with more reactogenicity noticed in subjects with active TB disease. The vaccine induced a high-magnitude, M72-specific CD4+ T-cell polyfunctional response (IFN-gamma, IL-2, TNF-alpha and CD40L) that persisted for more than 1 year.\n\nA phase 2b proof of concept efficacy study in approximately 3,500 IGRA positive, HIV-negative adults in clinics in South Africa, Kenya and Zambia, in which enrollees were followed up for three years for the occurrence of TB disease, has been concluded. Subsequent to the meeting reported here, results have been released47. The study demonstrated 54% (90%CI 14%–75%) vaccine efficacy against pulmonary tuberculosis, during a period of follow-up of approximately 2 years. Increased reactogenicity (local reactions and flu-like symptoms) were more common in M72/AS01E vaccine recipients relative to placebo recipients, but no unexpected safety concerns were raised. Secondary analyses suggest vaccine efficacy does not appear to wane over the follow-up time, and may be a function of age, but secondary signals should be interpreted with caution, as they are based on small numbers. The study is ongoing, with final results expected to be released in 2019.\n\nThe H56:IC31 vaccine is an adjuvanted fusion protein, consisting of three highly immunogenic Mtb antigens – Ag85B, ESAT-6 and Rv2660c – and adjuvanted with the Valneva IC31© adjuvant48. ESAT-6 is a premiere virulence-associated antigen highly expressed throughout all stages of infection while Rv2660c is a stress-induced antigen, strongly associated with latent TB infection. The IC31 adjuvant consists of ODN1a, a TLR9 ligand, and a stabilizing molecule that helps establish depot formation. The central immunologic strategy of this vaccine is to generate long-lived, Th1-type T-cell immunity. The vaccine is being developed by the Statens Serum Institute (Denmark), with multiple partners including Valneva, GmBH (Austria), Aeras, the South African TB Vaccine Initiative (SATVI), The European and Developing Countries Clinical Trials Partnership (EDCTP) and the Research Council of Norway.\n\nH56:IC31 is being developed to prevent TB disease in adolescents, adults and the elderly, as well as in children and infants, to prevent recurrent TB, and to shorten drug treatment for active TB and serve as adjunctive immunotherapy to patients with active TB, including those with drug-resistant TB. The targeted efficacy is ≥50%, with a duration of protection of ≥10 years. The vaccine is administered via intramuscular injection (IM) in two doses.\n\nPhase 1 and phase 2 studies of safety, immunogenicity and dose finding have been completed in IGRA+ and IGRA- adults, IGRA- adolescents, and in adults completing treatment for active TB. Thus far, the vaccine has demonstrated no safety concerns. Immunogenicity studies demonstrate a strong induction of ESAT-6 responses both in IGRA+ persons and in those completing treatment for active TB48. Antibody responses also were noted in a proportion of vaccinees. Along the therapeutic track, an ongoing, open label phase 2a trial in Norway is assessing the safety and immunogenicity of H56:IC31 given three months into active TB treatment as adjunctive immunotherapy with and without additional COX2-inhibition. A major phase 2 PoR trial of H56:IC31 is being planned and will enroll 900 persons being treated for pulmonary TB; 450 receiving vaccine and 450 administered a placebo. Vaccine will be administered within the last month of planned completion of 6 months of drug treatment, with follow-up for 12 months following vaccination. The trial is designed to have 80% power to detect 50% vaccine efficacy with a significance level of 0.05, given an expected TB recurrence rate of 4% per year in unvaccinated controls. Twenty-three endpoints (recurrent TB) are expected over the 1 year of follow-up. The primary objective of this trial is to accelerate the development of H56:IC31 toward a possible phase 3 PoR trial and licensure for this indication. Capacity building and studies of mechanisms of reactogenicity, immunogenicity and efficacy also represent important components of this trial. Enrollment in this trial is planned for Q3 2018.\n\nID93 is a fusion of four Mtb antigens with diverse roles that are recognized by T-cells isolated from Mtb-exposed individuals and lack human sequence homology49. The fused proteins in the ID93 vaccine include Rv1813, an antigen up-regulated under hypoxic conditions; Rv2608, the PPE protein, probably associated with the Mtb outer-membrane; and Rv3619 along with Rv3620, both included among the EsX protein family of secreted virulence factors. GLA-SE is a synthetic TLR-4 agonist adjuvant, formulated in a squalene oil in a water nano-emulsion. GLA-SE has been demonstrated to be safe in humans, with thousands of doses delivered, induces a TH1-biasing immunological response, and production is readily scalable.\n\nID93/GLA-SE is being developed for two indications: as an immunotherapeutic agent to improve the outcome of drug treatment for active TB, and as a prophylactic vaccine to prevent infection with TB. Proof of principle for the ability of ID93/GLA-SE to enhance TB treatment has been demonstrated in mice, guinea pigs and non-human primates, where immunotherapeutic vaccination with ID93/GLA-SE in combination with multidrug therapy was associated with a polyfunctional Th1 response, improved bacterial clearance and reduced pulmonary inflammation50. Protection against progression to pulmonary TB in a guinea pig Mtb challenge model was observed for more than 1 year51.\n\nTwo phase 1, and one phase 2a clinical trial of ID93/GLA-SE in healthy adults in the United States and South Africa have been completed. These trials, assessing safety, immunogenicity and dosing approaches, have included persons not vaccinated with BCG, BCG vaccinated individuals, and persons who are IGRA-, IGRA+ and those with active TB disease. Results from these studies demonstrated the induction of a broad, polyfunctional T-cell response, an increase in multi-functional antibodies, and robust CD4+ T-cell responses in IGRA+ adults, suggesting that the vaccine boosts the immune response to natural infection49. The phase 2a trial, in which vaccination occurred at the end of TB treatment, demonstrated encouraging CD4+ T-cell and antibody responses to vaccination. These studies also demonstrated an acceptable safety profile for ID93/GLA-SE in more than 200 research participants. Several new Phase 2 studies of ID93 / GLA-SE are being prepared. A Phase 2a study to evaluate the safety, immunogenicity and preliminary efficacy of ID93 / GLA-SE in preventing infection with TB among high-risk health care workers in Korea is being planned. Additionally, two Phase 2b clinical trials of ID93 / GLA-SE to evaluate the vaccine as an immunotherapeutic adjunct to multidrug therapy have received funding. In India, a collaboration with the All India Institute of Medical Sciences will administer the vaccine to patients undergoing multidrug therapy for both drug-sensitive and drug-resistant TB. In South Africa, plans are being developed for an expansion of the Phase 2a prevention of recurrence study, with the intention to advance administration of the vaccine proximally towards the initiation of TB therapy, beginning at the completion of therapy and moving as early as two months after chemotherapy has started.\n\nRUTI is being developed as an immunotherapeutic agent for adults, intended to improve the efficacy and shorten the duration of drug treatment for cases of active TB, including drug-resistant TB52. RUTI is made of cell wall fragments of Mtb formulated in a liposome suspension and is administered subcutaneously (SQ) in a single dose. Preclinical experiments demonstrate an added effect compared to Mtb drug treatment in reducing Mtb colony counts in mice, a reduction in regrowth rate as far out as week 28 following Mtb challenge in guinea pigs, and a diminished TB relapse rate in a paucibacilli murine model of TB.\n\nTwo clinical trials have been completed. The initial study, a phase 1 safety, immunogenicity and dose-ranging trial in 24 healthy adults in Spain, triggered a vaccine-specific immunological response against several Mtb antigens, without significant toxicity, supporting further development53. The second study was a phase 2 safety, immunogenicity and dose ranging trial in 48 HIV+ and 48 HIV- persons with LTBI in South Africa52. These studies demonstrated a good cellular polyantigenic response after a first injection of 25 mcg in both HIV- and HIV+ participants and elicited a long-term memory surrogate response compatible with a prophylactic potential already observed in animal models.\n\nFuture development will focus on utilizing RUTI as an adjunct to drug treatment in patients with MDR-TB, to prevent disease relapse after treatment for drug-sensitive TB, and to shorten the drug treatment regimen. A phase 2a trial to evaluate the safety and immunogenicity of RUTI in MDR-TB patients favorably responding to standard MDR-TB treatment is underway, with a plan for a subsequent phase 2b-phase 3 pivotal clinical trial in this population.\n\nThe Vaccae™ vaccine is a heat-killed preparation of Mycobacterium vaccae, an NTM closely related to M. obuense54. Vaccae™ has been licensed in China as adjunctive immunotherapy for drug treatment of active TB. A pooled analysis of data from trials of the M. vaccae vaccine suggested some potential utility when used as an immunotherapeutic adjunct to drug treatment of active pulmonary TB. In 2012, the Chinese government approved the initiation of a clinical trial to test the candidate vaccine for prevention of the occurrence of tuberculosis in subjects with LTBI. A 10,000-person, placebo-controlled phase 3 study involving a 6-dose vaccination regimen was initiated. Retention rates in the trial were high, and more than 100 cases of TB occurred during the study. The efficacy and safety data from this completed study have not yet been released.\n\n\n4. Diversifying TB vaccine platforms\n\nMost advanced vaccine candidates share one important commonality: they were selected for development on the basis of their ability to induce interferon-gamma producing T-cells obtained from vaccinated animals and individuals upon in vitro mycobacterial antigen stimulation. Very little evidence, however, is available to confirm the full scope of the critical immune responses necessary to control TB disease and Mtb infection, including the potential roles of CD8+ T cell responses, tissue resident memory T-cells, mucosal-associated invariant T (MAIT)-cells in the lungs and B cell activation. Studying these and other diverse immune responses generated by current and future vaccines likely will be critical to advancing TB vaccine development in a more efficient and effective manner55.\n\nMany of the more diverse immune responses likely to be important in controlling TB in humans may not occur in small animal models, or even in non-human primates. Accordingly, “experimental medicine” studies, involving 20–30 people and designed to gain a better understanding of a vaccine’s ability to generate the types of immune responses that may be critical in controlling TB, particularly those occurring in lung parenchyma, in bronchus-associated lymphoid tissue, and in the pulmonary mucosa, will be important to conduct. It also remains important to engage in parallel, more empirical vaccine efficacy and safety studies designed to inform immune mechanisms of protection. Collection of biospecimens in clinical studies in support of planned or future immunological studies, using the best technology available, is critical to permit retrospective analyses of correlates of protection once clinical protection is demonstrated by a vaccine under development.\n\nWhile it is important to develop and utilize a TB vaccine as soon as possible, it also is important to remember that evidence of TB disease has been discovered in humans occurring as far back as 9000 BCE56. Given this lengthy history between TB and humans, it is critical that we take the long view when developing vaccines for TB; whether we develop a vaccine by 2035 or by 2050 represents a rounding error from the perspective of human history.\n\nMtb represents a particularly difficult challenge for vaccine developers as it appears to both evade host immunity and modulate it in its favor. Accordingly, a successful TB vaccine likely will have to hit Mtb infection hard and early, in a manner that prevents Mtb from taking control of the host immune response directed against it. A CMV-vectored vaccine strategy offers the potential for accomplishing this by generating “effector-memory” T-cells prepositioned in the lungs, at sites of early pathogen colonization57.\n\nThe strategic rationale for selecting a CMV vector for a TB vaccine resides in its unique ability to elicit and maintain robust, effector-differentiated CD4+ and CD8+ T-cell responses in organs outside lymphoid structures, including the respiratory tract. Experiments with rhesus CMV (RhCMV) in rhesus macaques (RMs) demonstrate that RhCMV vectors expressing foreign antigens manifest the critical biological properties needed to generate effector-memory T-cells in a safe manner. These properties include the ability to super-infect rhesus already infected with CMV, and to persist indefinitely despite robust anti-CMV immunity in naturally CMV seropositive animals; the capacity to elicit and indefinitely maintain high frequency, “effector differentiated” T-cell (T effector and T tissue-resident effector) responses in mucosal sites, lymphoid tissues and parenchymal organs; and the ability to maintain immunogenicity despite profound attenuation. In particular, genetically attenuated (ULb’ mutated) RhCMV vectors (68-1 RhCMV) demonstrate an unexpected and unusual capability to elicit CD8+ T-cell responses that entirely, and broadly, target novel epitopes restricted by MHC-E and MHC-II58.\n\nThe primary rationale for a CMV-vectored TB vaccine is to generate tissue effector T-cells capable of immediate interception of Mtb-infected macrophages at sites of initial lung infection prior to the initiation of the substantial immune system modulation caused by Mtb necessary to permit Mtb survival and replication. Two RhCMV-TB vaccines have been created thus far, one expressing nine Mtb proteins across four different RhCMV vectors, the other expressing six Mtb proteins as a single polyprotein from one vector. Two RhCMV-TB efficacy pre-clinical assessments studies of these vectors, utilizing a low-dose (25 colony forming unit (CFU)) challenge of Erdman strain Mtb, have been completed57. In the first study, the extent of lung, non-lung and overall TB disease at one year following Mtb challenge was significantly reduced in RhCMV-TB vaccinated NHPs as compared to unvaccinated and BCG vaccinated controls. BCG vaccination prior to RhCMV-TB vaccination appeared to reduce the efficacy of the experimental vaccine. In the second study, no TB disease was detected via lung CT scans in 13 of the 27 CMV-TB vaccinated animals, as compared to unvaccinated controls where all demonstrated TB involvement of lungs and lung-draining lymph nodes. While all unvaccinated RMs demonstrated both pulmonary and extra-pulmonary TB at necropsy, no gross or microscopic TB was found in the 13 RMs with CT scans that did not demonstrate TB disease. Moreover, 10 of these 13 were Mtb culture negative. Taken together, these studies demonstrated unequivocal protection—defined as a finding of no granulomatous disease at necropsy (41%) and a significant reduction in TB disease (32%)—in 73% of vaccinated RMs. Comparisons between different CMV-TB constructs also determined that efficacy is not dependent on unconventional MHC-II and MHC-E restricted CD8+ T-cell responses.\n\nFuture steps in the development of these promising vaccine constructs will include optimization of the vector backbone and the gene inserts and validation of the final CMV-TB vector design in NHP challenge experiments. The first phase 1 clinical trial of a CMV-TB vaccine is targeted for 2020.\n\nMessenger RNA (mRNA) vaccines represent a non-viral delivery system of vaccine antigens that offers the potential of stimulating both the innate and adaptive immune systems in a manner that provides a balanced cellular and humoral antigen-specific response59. Factors spurring interest in mRNA vaccines include their ability to express complex membrane proteins, induce both CD4+ and CD8+ T-cell responses (unlike many protein antigens), permit efficient re-dosing given the absence of anti-vector immunity and their non-cell based manufacturing platform, allowing more rapid production than for vaccines requiring cell culture-based manufacturing processes. Potential cons to utilizing mRNA vaccines include the reactogenicity of the delivery systems currently in use which limits the mRNA dose and antigen multiplicity, and cost-of-goods considerations given target price points for a TB vaccine.\n\nNo mRNA vaccine as yet has been developed for TB. mRNA technology, however, is a potentially attractive vaccine strategy against TB in light of the platform’s ability to induce potent T- and B-cell responses, including polyfunctional T-cell responses that home to the lungs, the ability of these vaccines to co-express multiple antigens, and the potential to deliver these vaccines via ID, IM, subcutaneous (SC), intranasal (IN), aerosol (AE) and intravenous (IV) routes, based on the formulations. Limitations to their use for TB include the non-persistent nature of antigen expression and the limited number of antigens that can be included in each vaccine. Ultimately, the identification of a set of Mtb antigens that offer the possibility of effective protection would be needed to stimulate greater interest in applying mRNA vaccine technology to TB prevention efforts.\n\nLive, attenuated vaccines (LAVs) are pathogens missing genes responsible for producing virulence factors or key metabolic enzymes, with the absence of these genes resulting either from rational design or randomly through passage. When creating LAVs as vaccine candidates, developers must maintain a balance between attenuation and virulence, as too much attenuation may impair the ability of the candidate to generate a sufficiently protective immune response, while insufficient attenuation would raise safety concerns.\n\nMajor advantages to using LAVs include their ease and low cost of production, their longer persistence and the ability to stimulate both adaptive and innate immune responses, thereby obviating the need for adjuvants with the establishment of a long-term, comprehensive immunity. This stands in contrast to inactivated whole cell vaccines that primarily induce neutralizing (humoral, antibody-based) immunity with minimal mucosal and innate immune stimulation and little generation of cell-mediated immunity (CMI); and recombinant vector vaccines that primarily generate CMI with minimal innate and mucosal immunity, and with little to no humoral immunity60.\n\nThe primary disadvantage to LAVs is a concern about their safety. The potential for reversion to wild-type virulence remains a possibility for organisms attenuated through the deletion of one or a few virulence genes. Additionally, LAVs that behave in an attenuated fashion in persons with normally functioning immune systems may prove dangerous and even deadly if administered to persons with compromised immune function.\n\nRational deletion of Mtb virulence genes represents an exciting approach to creating a new generation of TB vaccines. One attenuated Mtb vaccine concept results from the deletion of MosR (Regulator of Mycobacterial Operons of Survival, Rv0348) that plays an important role in Mtb survival during infection61. C57BL/6 mice vaccinated subcutaneously with the ΔMosR vaccine demonstrated a low level of vaccine persistence in lung and spleen for at least 16 weeks following vaccination. After challenge with the highly virulent Mtb Beijing strain, improved protection over BCG-vaccinated mice was observed; in some animals Mtb could not be detected at 60 days after challenge, with benign-appearing lungs on histopathological examination62.\n\nIn order to reduce the chance of reversion to a virulent, wild-type state, another gene deletion, involving the echA7 gene, has been introduced. A plan is underway to combine these two deletions into the same Mtb strain and assess the potential utility of this construct as a TB vaccine.\n\nThe current approach to TB vaccine development suffers from three shortcomings: 1) it has focused mostly on a narrow set of candidate TB antigens which may have suboptimal activity in protecting against TB; 2) it has focused on generating classical, CD4+ T(h1) cells, which may be essential but not sufficient to generate an optimally protective response; and 3) has not taken into account recently emerging evidence suggesting a role for traditionally “ignored” cells, such as B-cells, in generating immunological protection against TB. Accordingly, exploring the role of unconventional T-cells and T-cell responses, such as donor unrestricted T-cells (DURTS), mostly CD8+ T-cells restricted by CD1, MR1, HLA-E, TCR-gamma-delta; non-IFN-γ producing T-cells; and the role of non-T-cells will be critical to future TB vaccine development efforts55,63.\n\nThe current global clinical pipeline of TB vaccine candidates utilizes a limited number of the approximately 4,000 Mtb antigens. A critical question is whether the optimal antigens are being selected for inclusion in protein subunit and recombinant viral vaccines. In a large phase 2b trial of the MVA85A vaccine in South African infants, the vaccine did not provide significant additional protection above BCG vaccination64. A possible explanation for this outcome is that Ag85A, although highly immunogenic, may not have been the right antigen given that its expression is downregulated later in infection and may not be presented sufficiently by infected cells.\n\nAn emerging hypothesis in defining optimal antigens to include in a TB vaccine is to identify Mtb antigens expressed during infection in the lungs of susceptible individuals65–67. To identify such antigens, an unbiased, genome-wide antigen discovery approach was taken, utilizing Mtb RNA isolated from the lungs of four different mouse strains, ranging from hyper-susceptible to TB (C3H/FeJ mice, the sst1 strain) to genetically resistant (C57Bl6 mice). Utilizing a genome wide qRT-PCR platform developed at Stanford University, Mtb genes were selected that were persistently and highly expressed in vivo (in vivo expressed – IVE-TB – genes) at multiple time points following AE Mtb infection. 194 IVE-TB highly expressed genes were identified and 50 further selected based on ranking in the top 15% during infection; hyperconservation with wide HLA coverage and/or homology with Mycobacterium leprae. Many of these IVE-TB antigens were found to induce strong CD4+ T central memory and CD8+ T-cell responses in PBMCs from long-term, latently Mtb infected individuals, and were recognized by both T- and B- cells67.\n\nAlmost all Mtb antigen discovery approaches have mainly relied on IFN-γ measurements. Recent evidence, however, suggests that IFN-γ only contributes marginally to overall protection against TB in the lungs of mice68. An investigation of other responses, including proliferative and cytokine profiles besides IFN-γ, in a cohort of healthy Dutch individuals with PPD and/or ESAT6+CFP10 in vitro responses, and in a cohort of Norwegian individuals with LTBI, demonstrated that many IVE-TB antigens induce cytokines other than IFN-γ69.\n\nThese new approaches may lead to novel classes of antigens with promising vaccine potential. Additionally, B-cells play a critical, underappreciated role in conferring immunity to human TB70,71. B-cells may provide an important, novel target for TB vaccination, but Mtb antigens that activate B-cells have not been extensively studied as yet.\n\nIt has long been appreciated that T-cell responses are critical in controlling TB. T-cells do more than direct CMI responses, however, as they also provide a crucial link to humoral immunity based on B-cell responses.\n\nExperiments in mouse models of Mtb infection demonstrate that depleting B-cells decreases mouse survival following Mtb exposure, while B-cell depletion followed by B-cell restoration returns survival to the degree observed in control mice, suggesting an important role for B-cell responses in controlling TB71,72. Moreover, histopathological studies in Mtb-infected humans and NHPs demonstrate aggregates of B-cells proximal to pulmonary granulomas, with B-cell laden tertiary germinal centers surrounding the granulomas73,74. Antibody-deficient mice also demonstrate a greater degree of viable Mtb in both lungs and spleen, and a marked decrease survival after infection with Mtb75. Passive transfer experiments with the monoclonal antibodies anti-lipoarabinomannan (LAM) and anti-HspX (directed against the 16 kD protein on Mtb outer membranes expressed during times of mycobacterial stress, including within granulomas) resulted in improved survival (with anti-LAM antibodies)76 and decreased lung Mtb CFUs (with HspX antibodies)77 in mice following Mtb challenge. These experiments provide additional evidence that antibodies limit Mtb infection. Improved mouse survival also was seen following administration of monoclonal antibodies directed against the mycobacterial heparin-binding hemagglutinin (HBHA), with no effect on diminishing lung Mtb CFUs but with a marked decrease in spleen Mtb CFUs, suggesting an effect on decreasing Mtb dissemination78. Additionally, the particular isotype of antibody makes a major difference in the ability to control Mtb, with IgA monoclonal antibodies having a larger effect than IgG1 antibodies79.\n\nMost clinical studies of TB vaccines have not studied the role that antibodies may play in contributing to prevention of TB. Accordingly, little thought has gone into defining vaccine-induced antibody target product profiles (TPPs). In creating an antibody TPP for TB vaccines, it is important to recall that antibodies do much more than recognizing and blocking infection via pathogen neutralization. By binding to antigens on the surface of infected cells, antibodies play crucial role in stimulating a broader and more effective immune response, including the inducement of cytokine secretion to recruit cells such as dendritic cells and T-cells involved in mediating CMI, inducing autophagy in infected cells, and prompting innate immune responses including the recruitment of neutrophils, natural killer (NK) cells, monocytes and phagocytes.\n\nAdvances in systems serology assays now permit qualified and validated approaches to assessing antibody qualities and functions. Studies of TB patients utilizing systems serology have discovered that qualitatively different TB-specific antibodies are induced and distinct innate immune recruiting profiles are manifest in persons with LTBI as compared to those with active TB disease80. Antibodies play a key role in creating these differences, as antibodies expressed in persons with latent infection help stimulate innate immune responses by attracting dendritic cells to macrophages infected with Mtb, resulting in dendritic cell-mediated activation of innate immune responses. In particular, antibodies expressed in LTBI prepare the immune system to kill Mtb-infected cells, a signal sent via glycosylation patterns on the Fc domain of antibody molecules that were naturally modified during Mtb infection to access different innate immune functions81. Identifying the Fc glycosylation patterns that result in enhanced innate immune responses to TB provide a potential target for vaccine developers attempting to selectively induce such antibodies via vaccination.\n\nAntibodies also play a unique antimycobacteriological role by identifying cells infected with Mtb and restricting Mtb survival. Non-classical killing mechanisms triggered by antibodies include enhancement of opsinophagocytosis, lysosomal maturation, and inflammasome activation via a metabolic rewiring of macrophages, resulting in enhanced restriction of Mtb growth.\n\nClinical trials of vaccine candidates provide important opportunities to profile the quality of antibodies generated by these vaccines. Additionally, utilizing adjuvants offers an opportunity to generate many distinct antibody responses, resulting in the potential of rationally choosing adjuvants to more specifically stimulate the types of antibody responses that may be optimal in contributing to the control of TB. Accordingly, it will be critical to continue the clinical assessment of TB candidates and to build into these trials the opportunity to assess antibody responses as well as CMI.\n\nFour hypotheses have emerged to explain the possible role of antibodies in contributing to the control of Mtb infection and TB disease. The first hypothesis involves restricting the extent of initial Mtb infection in and around the alveolus by blocking Mtb physically, by inducing complement-mediated Mtb killing, and by stimulating enhanced phagocytic activity against Mtb. The second proposed mechanism involves balancing inflammation resulting from Mtb infection, mainly by redirecting pathologic innate immune activity in a way that reduces corresponding immunopathology. The third proposed mechanism involves regulation of granulomas, including prevention of granuloma formation (in light of the emerging theory that granulomas are used by Mtb to more efficiently spread infection via attraction of susceptible macrophages) and the clearance of Mtb-infected cells. Finally, antibodies may enhance dendritic cell and T-cell antigen presentation and regulation, through signals sent by specific Fc glycosylation patterns.\n\nStudies of antibody responses in LTBI populations, individuals who appear to be resistant to Mtb infection despite ongoing exposure, and persons participating in TB vaccine trials or in experimental medicine studies, utilizing more aggressive serological sampling techniques such as plasmapheresis, will be critical to advance the understanding of the potential roles played by antibodies in controlling TB. Using this information to design vaccine strategies that stimulate both humoral and CMI responses effective in controlling TB represents an important new strategy in TB vaccine development.\n\nAn important characteristic of a vaccine-stimulated immune response to Mtb is the ability of T-cells to recognize an Mtb-infected cell via receptors that are highly conserved in humans. Most vaccines currently being developed do this through “classical” HLA-I or HLA-II restricted antigen presentation. Donor unrestricted T-cells (DURTS) represent different types of T-cells that interact with human antigen-presenting cells (APCs), including dendritic cells or macrophages, through mechanisms other than classical HLA-I or HLA-II restricted antigen presentation. Examples of restriction molecule ligands associated with DURTS include: CD1a, b, c molecules, which present lipid and glycolipid moieties and bind CD1-restricted T-cells, including Type I and Type II natural killer T-cells (NKT cells); MR1, which presents riboflavin metabolites and interacts with mucosal-associated invariant T-cells (MAITS); and HLA-E, a highly conserved receptor that recognizes a diverse array of peptide ligands, binds to HLA-E-restricted T-cells and helps to mediate natural killer-cell killing82. Although our understanding of DURTS ligands is still relatively superficial given that this is a young field of investigation, the potential applicability to TB vaccine development is great given that virtually all of the types of antigens presented by these ligands can be found in Mtb.\n\nStudies of DURTS suggest that they contribute to protective immunity, with MR1 restriction involved in responses against Mycobacterium bovis via BCG vaccination83, and Klebsiella pneumoniae84,85, and both CD186 and HLA-E87,88 mediated responses contributing to Mtb control. It is not yet clear, however, if these are memory responses that can be generated through vaccination. Additionally, the duration of such responses is not known.\n\nMR1-restricted responses are highly conserved in mammals and do not diverge across mammal species. MR-1 interacts with MAITS as sensors of the metabolome via riboflavin metabolites, but also through other, as-yet-unidentified molecules89. In studies of patients with TB, a dramatic enrichment of MAITS was found in samples obtained from bronchoscopies as compared to peripheral blood, with selective enrichment of certain T-cell receptors (TCRs) identified in the bronchoscopically-obtained samples, suggesting antigenic discrimination. This finding led to the discovery of a novel class of MR1 ligands called photolumazines. Experiments with synthetic photolumazines suggest that there is immunological memory underlying MR1-mediated antigenic recognition that could be harnessed in a TB vaccine90.\n\nCD1-restricted T-cells recognize glycolipid antigens91,92. They represent an intriguing area of study given the presence of known antigenic glycolipids in Mtb, although the absence of group 1 CD1 molecules in mice increases the difficulty of assessing the importance of CD1-restricted immune responses in protecting against Mtb. Additionally, it is not yet known whether immunologic memory can be imparted to CD1-restricted T-cells.\n\nHLA-E restriction encompasses presentation of both peptidic and non-peptidic antigens, and initially was identified as an inhibitor of recognition between NK cells and T-cells. Subsequently, however, HLA-E-restricted T-cell clones have been found to manifest antimicrobial properties, including the inhibition of intracellular Mtb growth93. This inhibition appears to be mediated both by generating cytolytic properties in the T-cells, and by inducing a cytokine profile from these T-cells that resemble a Th2 phenotype. Additionally, gamma-delta cells, specifically γ9δ2 cells, have been found to be expanded following vaccination with BCG94 with further investigations suggesting a role in controlling intracellular growth of Mtb95.\n\nFurther elucidation of the role of DURT-mediated immune control of Mtb will most likely depend on the development of fit-for-purpose animal models, particularly mouse models, engineered to assure the relevance of DURT response assessments. An example of this approach is the development of a humanized transgenic mouse model which expresses group 1 CD196.\n\nA critical question regarding the potential for targeting DURTs through vaccination is whether DURTs have memory. If they do, and if durable expansion and preferential recall responses to these novel ligands can be elicited through vaccination, then DURTs could be harnessed directly as a potentially powerful vaccination strategy. If they are not capable of developing an anamnestic response but, instead, function more as adjuvants, then DURTs could be manipulated to help facilitate the acquisition of adaptive immune responses to Mtb. Ultimately, it will be necessary to develop fit-for-purpose animal models, identify the appropriate immunogens, develop techniques to formulate the ligands and create relevant animal challenge models to explore the potential of harnessing DURTs as a vaccination strategy against TB.\n\nBiomarkers that can reliably and accurately predict clinical outcomes of diseases are invaluable to vaccine development efforts. Biosignatures, custom-made combinations of biomarkers, offer the potential to enhance the predictive power of a single biomarker. Principally, biosignatures can be useful both in selecting promising vaccine candidates for more advanced development, while rejecting those with a reduced probability of ultimate success. They also can be used to more finely tailor the selection of individuals being enrolled into clinical trials, thereby reducing the cost and increasing the efficiency of such trials. Thus far, studies exploring blood, urine or breath of persons with asymptomatic Mtb infection have not identified a biosignature predictive of the development of active TB disease.\n\nThe development of advanced, data-dense assessment techniques, including transcriptomics, proteomics, metabolomics has enhanced the opportunity to identify biosignatures of the progression of Mtb infection to active TB disease. Transcriptomes, which require samples of whole blood, combined with computational modelling, currently represent the most widely studied biomarkers in TB. Taken together, small signatures, utilizing three to four transcripts and assessed via decision trees or other computational methods, have resulted in a biosignature which discriminates active TB disease from LTBI with 85–90% sensitivity and specificity97,98. Signatures comprising 16 transcripts now can predict the development of active TB in individuals with LTBI 6–12 months prior to a clinical diagnosis99. Further refinement of algorithms led to a predictive signature composed of only two transcript pairs which detect risk of active TB up to one year prior to clinical diagnosis of TB in different regions of the African continent100. Metabolomics represents an under-researched area, particularly given that such studies can be run utilizing serum or plasma rather than whole blood, reducing the complexity and cost of sample collection and storage. Initial studies applying metabolomics to diagnosing active TB101 have been promising, and are being investigated to provide a 6- to 12-month prognosis among individuals with LTBI.\n\nIt now appears clear that both metabolomic and transcriptomic signatures can be harnessed for diagnosing incipient TB, and for providing a 6- to 12-month prognostic window regarding the likelihood of developing TB among those with LTBI. While this has immediate relevance to decision-making around initiating preventive drug therapy before the development of clinical disease, it also provides the opportunity to stratify individuals in clinical trials of vaccines. Using these techniques to identify and selectively enroll LTBI individuals who are at risk of developing TB in an efficacy trial of a vaccine to prevent TB disease could considerably reduce the numbers of study participants required, shorten the clinical trial duration and greatly reduce cost by several orders of magnitude, assuming that these signatures could be utilized at reasonable cost and speed.\n\nImportant future directions in biosignature research relevant to TB vaccine development will be to identify signatures of vaccine efficacy and signatures of vaccine safety. While studies of BCG responses can lay the groundwork for this102,103, ultimately this effort will depend upon identifying an initial, partially efficacious TB vaccine through ongoing clinical trials, and having had the foresight to bank sufficient blood and serum specimens from trial participants to permit biomarker studies104. This further emphasizes the need to continue to conduct clinical efficacy trials of TB vaccines, and to invest in the collection of biological specimens from the participants. Only through efforts such as this will unexpected findings be identified that potentially could result in a major acceleration of novel TB vaccine development.\n\n\n5. Use of models in translational TB vaccine research\n\nEffective natural immunity against developing TB exists, given that 90–95% of Mtb-infected individuals fail to develop active TB disease1. While this rate of natural immune protection is difficult to improve upon, it will be necessary to do so to control the global TB epidemic. In order to effectively manipulate the human immune response to TB we require new and diverse vaccine candidates and we need to correlate the immune responses induced by these vaccines with the degree of protection conveyed. Small animal models of Mtb infection can assist in this effort105.\n\nSmall animals can be experimentally manipulated to provide a reproducible challenge model capable of screening vaccine modulated vertebrate immune responses to limit Mtb growth in the lung. Mice and guinea pigs (GPs) have been the main small animal species utilized for assessing vaccine candidates105. Mice are inbred, easily manipulated, inexpensive, and there are extensive reagents for immunological studies. Both mice and GPs provide the opportunity to compare vaccine efficacy by assessing bacterial load and survival following Mtb challenge. GPs are susceptible to very low dose Mtb challenge and can be used for natural transmission or repeated low- and ultra-low-exposure models. It may be feasible to utilize GPs in a post-exposure model as well. The pulmonary pathology of Mtb infection in GPs is similar to that of human primary TB106.\n\nPotential indications for TB vaccines in humans include prevention of TB disease (PoD), prevention of relapse or reinfection following treatment (PoR), and prevention of Mtb infection (PoI). To discriminate between potential PoD vaccines, TB challenge in vaccinated small animals provides a direct measure of an anti-mycobacterial host response as well as the opportunity to detect pathological outcomes. For meaningful PoD efficacy, the optimal outcome in a small animal would be to reduce bacterial burden below the level of detection. Assessment of PoR vaccines can also be performed in small animals, but due to the variability in outcomes, large numbers of mice are required107. A PoI vaccine model is not currently optimized for small animals, although studies of GPs in a natural exposure environment could provide proof of concept17.\n\nOther factors, such as the nutritional status of the animals, corticosteroid stress, the animal microbiome and the genetic background of the animal can influence the outcomes of challenge experiments. The development of “collaborative cross” (CC) mice, a multiparent panel of recombinant inbred mouse strains derived from eight founder laboratory strains108 provides a striking demonstration of the effect of genetics on response to vaccination and Mtb challenge, as shown by the differential effect of BCG vaccination on the genetically diverse CC mouse strains109. Additional factors that can affect outcomes are the bacterial strain selected for the challenge110 and the bacterial dose. While these external factors should be controlled for during vaccine screening, they also provide the potential for testing vaccine efficacy under a variety of conditions.\n\nThe revised TBVI/Aeras stage gate criteria make experiments in small animals an explicit part of TB vaccine candidate development110. Stage gates B and C include calls for exploration of safety, immunogenicity and protection in small animal species. In this context, protection is defined as being reproducibly and statistically better at preventing TB disease than BCG or a relevant benchmark. Ultimately, it will be important to determine the extent to which vaccine protection demonstrated in small animal Mtb challenge experiments is corroborated by data from experiments in NHPs and in human clinical trials.\n\nAssessment of TB vaccine candidates in small animal models can be compromised by suboptimal experimentation as well as inappropriate or over-interpretation of experimental outcomes. It is important therefore to state that there are no ‘bad’ models, only inappropriate interpretation of the results.\n\nStandard animal challenge models of Mtb infection differ from natural infection in a number of ways, including challenge with a higher Mtb exposure than the very low exposures that characterize natural infection; a single challenge as compared to the repeated exposures associated with natural infection; and exposure to a naturally occurring form of Mtb rather than laboratory-grown strains. Laboratory-grown strains are usually treated with mild surfactants such as Tween™ that prevent clumping and permit accurate quantification of dose, but strip the outer lipid capsule off of Mtb and thereby potentially alter its immunogenicity111. To overcome these issues, a human to guinea pig (GP) continuous natural exposure model was established17. In this model ambient air from a specially built patient unit caring for individuals with active TB is removed via negative pressure and exhausted into a neighboring facility harboring GPs, with each GP receiving an average 4-month exposure to the stream of air laden with naturally occurring, uncultured Mtb. A total of 362 GPs have been exposed over 12 years: 91 (25%) remained free of infection while 271 (75%) became infected as determined by tuberculin skin test (TST) conversion. Of the 271 TST converters, 53 (15%) experienced a reversion of their skin test, with 33 (9%) of the reverters subsequently reconverting, presumably due to re-infection. Only 54 (15%) of the infected or re-infected GPs went on to develop TB disease.\n\nTST reversions following initial conversion occurred in 15% of convertors, more commonly than had been anticipated. Repeat TST conversion following initial reversion was completely eliminated by irradiating the air from the patient unit, suggesting that the reversion and repeat conversion events were due to subsequent re-infection rather than non-specific reactions to non-viable Mtb. Examining the time course of TST reactions in these GPs over time revealed a pattern suggesting repeated Mtb exposures and subclinical infections, many which initially resolve but which ultimately result in established infection and disease.\n\nThese findings illustrate the critical role that repeated reinfection with Mtb appears to play in the overall course of TB pathogenesis. A correlative finding was described in a study of TB patients in a Siberian hospital, where admission into the hospital with drug-sensitive TB (DS-TB) translated into a six-fold increased risk of developing drug-resistant TB (DR-TB), as compared to patients with DS-TB treated as outpatients, due to re-infection with DR-TB while in hospital112.\n\nPreventing disease due to re-infection with new Mtb strains represents an important target for TB vaccines but a difficult one to assess given the absence of a biomarker for Mtb reinfection. A continuous exposure, natural transmission animal model that mimics human exposure in many high-burden settings, such as the human to GP exposure model described here, can serve to address this need. Vaccine suppression of transient TST or IGRA conversions seen in this model may represent an important target. A clinical study also is being implemented to assess the role that BCG vaccination may play in preventing IGRA-negative, BCG-naïve individuals who will be working in TB-endemic areas from acquiring Mtb infection. This investigation, called the Tuberculosis Immunization to Prevent Infection (TIPI) study, will enroll 2,000 USA-based healthcare and humanitarian workers (e.g., Peace Corps workers, persons entering service with Doctors without Borders or other NGOs). Half will be will vaccinated with BCG, and IGRA conversion rates will be compared against the unvaccinated arm over the period of exposure. Blood samples also will be collected to permit an assessment for correlates of immune protection if a difference in IGRA conversion rates is found between vaccinated and unvaccinated cohorts.\n\nNon-human primates (NHPs) represent the closest model of human Mtb infection113. NHPs demonstrate the full spectrum of TB, including active disease, latent infection and reactivation114,115. TB-induced pathology in NHPs is reflective of human pathology, including caseating granulomas and other granuloma types, as well as cavitary lung disease115. Additionally, NHPs are immunologically closer to humans than other animal models as they possess unconventional T-cell subsets, including CD1a-d and delayed type hypersensitivity (DTH) responses to BCG and TB113. Reagents and laboratory technologies are available for the detailed analysis of the immune system of macaques, a NHP species frequently used to study TB pathogenesis and in Mtb challenge experiments. Sequential sample collections from blood and mucosal sites also are possible, permitting time course studies of systemic and mucosal immunity that enable studies of biomarkers and immune correlates of protection.\n\nRhesus macaques, mainly of Indian origin, and cynomolgous macaques currently are the most common NHP species used to model Mtb infection and TB disease116. One major advance in improving NHP models of Mtb challenge applicable to vaccine candidate assessment has been the development and application of advanced, non-invasive imaging techniques to serially track disease progression in a sensitive and quantifiable manner. The simultaneous application of computerized tomography (CT), which provides detailed images of internal structures and permits the quantification of pulmonary disease burden, in combination with positron electron tomography (PET), which demonstrates the spatial distribution of metabolic activity, thereby providing fundamental information relevant to TB pathogenesis, have proven valuable in assessing the degree of protection provided by newly-developed vaccine candidates20.\n\nAnother major improvement in NHP models of TB has been the development of low-dose Mtb challenge (<25 CFU) and very-low-dose Mtb challenge (<10 CFU), doses that more closely approximate natural human exposure to Mtb than did the 500–3,000 CFU exposures previously used. Additionally, the development of advanced tools for studying immune responses, permitting detailed studies of adaptive immune responses as well as assessments of diversified components of the rhesus immune response, have increased our understanding of NHP immune responses to vaccination and its effect on controlling an Mtb challenge before the candidate under study is selected for clinical evaluation.\n\nAs macaque populations demonstrate different responses to Mtb infection, it is important to select the rhesus species most relevant to the target product profile of the candidate vaccine when designing challenge experiments. It is also important to consider the species in which the experiments have been done when attempting to draw comparisons between challenge studies. For example, cynomolgus macaques demonstrate a greater degree of disease control following Mtb challenge than do rhesus macaques, making the former potentially more representative of Mtb infection in adults, and the latter modeling Mtb infection in infants and young children, where de novo infection tends to have more serious consequences117. The ability to control Mtb infection also varies within species; among cynomolgus macaques, the Asian sub-species controls Mtb infection better than those of the Mauritian sub-species. Among rhesus macaques, Chinese rhesus control Mtb better than do those of the Indian subspecies. To further facilitate inter-study comparisons, a greater degree of harmonization and standardization of experimental methods must be achieved, including challenge strain selection, use of imaging, scoring gross pathology specimens and identifying priorities as to future experimental directions, such as assessing AE delivery of vaccines. This harmonization and standardization effort is underway within the Collaboration for TB Vaccine Development (CTVD), led by the Bill and Melinda Gates Foundation (BMGF). CTVD recommendations regarding NHP-based study design for TB vaccine development have recently been published118.\n\nMost TB vaccine candidates reduce Mtb CFU by ~1–1.5 logs in mice challenge experiments and are usually considered sufficiently protective to merit further development if they result in a ~1 log improvement over the reduction seen by BCG. As even this degree of CFU reduction results in ~1×104 Mtb CFU in mouse lungs following a standard Mtb challenge experiment, the question is whether this is a sufficient degree of CFU reduction to suggest the potential success of the vaccine candidate. The acceptance of this degree of post-vaccination pulmonary Mtb CFU has been largely based on the idea that this may be the full extent of suppression that TB vaccines can produce in mice, given the “limited dynamic range” of the murine immune system. It is unclear, however, whether this assumption is correct, or whether sterilizing immunity, or something approximating this, can, in fact, be generated in mice through manipulations of murine immunity that might be achievable through vaccination.\n\nThe standard mouse model of vaccination and Mtb challenge calls for vaccination at time 0, permitting 30 days to pass to allow the generation of a memory response, and then administering an AE Mtb challenge. Subsequently, the degree of protection afforded by the vaccination, and the Mtb-specific T-cell responses are assessed.\n\nChallenge studies in vaccine-naïve mice reveal that the pulmonary accumulation of Mtb-specific T-cells during primary Mtb infection is delayed by approximately 21 days, with peri-granulomar B-cell follicles – inducible bronchus-associated lymphoid tissue (iBALT) – appearing at 24 days119. Vaccinating the mice with BCG accelerates the Mtb-specific T-cell response by one week, with these cells appearing in the pulmonary parenchyma between 12–14 days following Mtb challenge and with iBALT appearing at day 17. In either case, Mtb is capable of establishing viable and progressive pulmonary infection, albeit with fewer organisms following BCG vaccination.\n\nIt is possible that the delay in appearance of Mtb-specific T-cell responses in the lungs may represent a critical bottleneck to developing effective TB vaccine-induced immunity, a bottleneck that could be overcome if Mtb-specific T-cells could arrive at the initial sites of infection earlier. To test this hypothesis, dendritic cells (DCs) pulsed with Mtb antigens were transferred intra-tracheally one day before and then four days after AE Mtb challenge to mice that had received BCG vaccination, and some of which had also received a mucosal boost with Ag85B administered intranasally. This DC transfer overcame the TB vaccine bottleneck, resulting in the presence of Mtb-specific, cytokine-producing T-cells on the third day following challenge, inducing iBALT 8 days following Mtb infection, and resulting in rapid lung macrophage induction in the vaccinated mice120. The rapid vaccine-induced T-cell responses limited early Mtb growth, reducing Mtb CFU by 2 logs or more over vaccinated animals that had not received the DC transfer, with Mtb undetectable in approximately 50% of the mice upon sacrifice on day 20.\n\nThe DC transfer proved to be more effective in inducing sterilizing or near-sterilizing immunity in mice that had received the mucosal Ag85A vaccine in addition to BCG, as compared to mice that only had received BCG. Gene expressions studies revealed that the gene expression signature of T-cells induced by the vaccines and antigen-pulsed DCs were enriched for the mucosal/airway phenotype. When the experiment was repeated using a multi-drug resistant Mtb strain as the challenge strain, however, sterilizing immunity was not achieved. Additionally, endogenous, host DCs could be induced to generate near-sterilizing immunity through the administration of a select activation of innate mechanisms including CD103 and CD40 pathways. Delivery of amph-CpG adjuvant and CD40 agonist at the time of challenge, rather than through the transfer of pre-stimulated DCs also provided complete Mtb control, providing additional support for the possibility of stimulating this type of response through vaccination120.\n\nHints at the possibility of achieving near-sterilizing immunity against Mtb through vaccination have occurred with Mtb challenge experiments in rhesus macaques following vaccination with a CMV-TB vaccine construct48 and in mice following vaccination with the ΔMosR and ΔechA7 vaccine concepts62 or the ΔSigH vaccine concept121, all live, attenuated Mtb vaccines. If ongoing experiments continue to demonstrate sterilizing or near-sterilizing immunity with these or other vaccine concepts, this would suggest that a “limited dynamic range” of murine immunity was not the explanation for the limited degree of Mtb suppression noted with other vaccine candidates. The location and timing of the vaccine-induced immune response to Mtb (i.e. developing an Mtb-specific immune response in the lungs within days of initial infection, rather than weeks) may be the keys to a successful vaccination strategy against TB.\n\nAn inhaled TB vaccine makes intuitive and logistical sense for a number of reasons122. Primarily, an inhaled vaccine mimics the route of Mtb infection, and offers the potential for generating potent and durable mucosal immune responses and stimulating specialized, peri-bronchial lymphoid tissue that may be critical in controlling TB. Additionally, BCG, administered through the universally used ID route does not reliably protect against pulmonary TB. From a logistical perspective, inhalation represents a common route of drug delivery which is safe, feasible, needle-free and pain-free.\n\nIn a phase 1 clinical trial assessing the safety and immunogenicity of 1×107 plaque-forming units (pfu) of the MVA85A TB vaccine administered via inhalation or ID in 22 BCG-vaccinated adults, Ag85A-specific CD4+ T-cell responses in fluid obtained via bronchioalveolar lavage (BAL) were found to be stronger after AE than ID administration30. Systemic Ag85A CD4+ T-cell responses were at least as strong after AE vaccination compared to ID administration. Additionally, no systemic anti-MVA vector IgG, IgA or IgM responses were detected following aerosol, but did develop following ID vaccination.\n\nAn important question to address when contemplating AE administration of TB vaccines is the safety of this administration strategy in persons with LTBI. This issue is being addressed in a phase 1 study of 15 persons with LTBI. Initial results from the first two patients enrolled have not raised safety concerns thus far.\n\nAdditional studies of TB vaccines administered via the AE route are due to begin soon, including a first-in-human study comparing AE and ID administered ChAdOx1.85 (University of Lausanne, Switzerland) and aerosolized Ad585B (McMaster University, Canada). A clinical trial assessing the safety and immunogenicity of increasing doses of BCG administered via the AE route based on encouraging protection and immunogenicity data in rhesus macaques, is ongoing, with no apparent safety issues raised and with encouraging, early immunogenicity data. It will be important to expand AE vaccination studies to sites endemic for TB to fully assess the safety, immunogenicity and efficacy of this vaccine administration strategy.\n\nAn effective vaccine against TB will likely need to 1) stimulate an effective and persistent T-cell response against multiple antigens; and 2) generate this response in the lung, the site of initial Mtb infection and TB-associated pathogenesis.\n\nLive, attenuated vaccines are an attractive vaccine strategy when targeting infections that are comprised of thousands of potential antigens and whose control depends on an effective and persistent T-cell response at tissue sites, such as malaria, leishmania and tuberculosis. Accordingly, live attenuated vaccines are an effective approach for generating T-cell responses of high magnitude and antigenic breadth compared to other vaccine platforms. Additionally, live or live-attenuated vaccines can stimulate T-cell responses in locations critical to inducing protection depending on the route of administration, thereby providing a rapid immune response at the time of infection. For malaria, intravenous but not subcutaneous immunization of a live, attenuated whole sporozoite vaccine was shown to induce a high frequency of antigen specific T-cells in the livers of NHPs. Also, attenuated sporozoites administered IV but not SC could protect humans against experimental controlled challenge123. These data highlight the importance of the route of immunization on an infection requiring tissue resident T cells to multiple antigens.\n\nFor TB, the route of vaccine administration also influences the sites of T-cell priming, and the location and strength of where tissue resident T-cells develop. ID or IM vaccination is the route by which BCG and all other TB vaccine candidates in phase 2 or phase 3 clinical trials are administered respectively. These routes primarily result in T-cell priming in local draining lymph nodes, a moderate circulating T memory (Tmem) response and a weak lung-resident effector (Teff) memory response, not an optimal situation when a rapid, lung-based T-cell response is needed for protection. Vaccines administered via the AE route result in local T-cell priming in the lung with a weak peripheral response121,124,125. Combining ID or IM administration with AE administration theoretically represents a strategy by which both lung Teff memory cells, in combination with circulating Tmem, may be generated. With regard to AE delivery of vaccines, a series of NHP experiments with recombinant adenoviral-vectored TB vaccines encoding 2–4 antigens delivered via AE, generated robust T cell immunity in the lung, but did not result in greater protection against TB disease than with BCG given ID126. These data suggest that there was either lack of antigenic breadth presented by the vectors, poor quality of the virally induced responses, or unfavorable innate imprinting by the viral vectors associated with increased risk of Mtb infection. Based on the results of clinical trials with subunit vaccines given by IM route and the NHP data by the AE route, data suggest that vaccines with greater antigenic breadth and an alternative route of immunization may be optimal.\n\nVaccines administered IV result in T-cell priming in the spleen, and strong lung-resident Teff memory and peripheral Tmem responses. Previous Mtb challenge experiments in NHPs receiving IV BCG administration resulted in greater protection than NHPs receiving either ID or a combination of ID and intra-tracheal BCG29,127,128. To directly assess the effect of various vaccine administration routes of a vaccine presenting a wide breadth of antigens, rhesus challenge experiments were conducted with BCG administered by ID, AE, IV and ID + AE, with a low-dose (15 CFU) challenge 6 months after immunization (BCG doses: ID – 5×105; AE – 5×107; IV – 5×107). Initial immunogenicity results from BAL samples demonstrated that IV BCG resulted in a significant increase in the frequency of TB-specific CD4+ T-cells, and CD8+ T-cells compared to the other administration routes. Additionally, IV BCG strikingly altered the proportion of T-cells and macrophages in BAL samples. At 16 weeks following vaccination, BAL samples after IV BCG demonstrated approximately 75% T-cells and 25% macrophages, while BAL obtained following the other routes of vaccine administration demonstrated greater than 75% macrophages and less than 25% T-cells at this time. Intravascular staining of NHP lungs 10 months following vaccination demonstrated that 5–10% of cells in the lung parenchyma were T-cells, showing that IV BCG resulted in a long-lasting increase in pulmonary T-cells. The extent to which IV BCG actually persists in the lung, and the degree of protection against Mtb mediated by administration of BCG via the IV route, is being assessed.\n\nControlled human infection models (CHIMs) have proven useful in vaccine development efforts, including vaccines for malaria, influenza and Salmonella typhi. Development of a CHIM for TB vaccine development would represent a major step forward, given the uncertain predictive value of preclinical animal models, the lack of a validated immune correlate of protection and the value of detecting a signal of vaccine protection in the host species. A CHIM to support TB vaccine development could be used for vaccine selection as well as for immunobiology studies to inform basic knowledge gaps.\n\nDeveloping a TB vaccine CHIM is challenging task given that virulent Mtb cannot be used in the challenge model for obvious ethical reasons. Accordingly, a major question facing TB vaccine CHIM developers is whether Mtb can be manipulated to be safe enough to administer to volunteers. An additional question is whether BCG could be used either as the challenge organism or at least as an agent that would permit further clinical development of the clinical challenge model.\n\nThere are two key elements to developing a human challenge strain of Mtb for a CHIM: 1) developing a control system to elicit bacterial death; and 2) developing a system to detect viable Mtb in the days and weeks following challenge. When developing a system to control Mtb death, consideration must be given to the need to permit Mtb survival for multiple generations to permit an assessment of the effect of the vaccine on Mtb survival. At the end of the experiment, the killing control system must be able to eliminate all bacteria without relying on a lengthy course of antimycobacterial agents, and without the possibility of clinical relapse. Currently, three potential viability control systems are being assessed. The first strategy relies on the degradation of non-canonical amino acids upon which Mtb auxotrophs are dependent for survival. Another approach is the insertion of “kill switches” into the Mtb genome, which are inducible through the administration of exogenous molecules, such as tetracycline, that induce a gene that codes for a mycobacterium-directed toxin. Thirdly, a strategy involving protein degradation targeting is under investigation. It is possible that more than one of these approaches could be used in combination to improve safety.\n\nDetection systems for viable Mtb will need to measure the levels of bacterial load in the lungs without relying on a CFU count. Moreover, the detection strategy must be non- or minimally invasive. Detection systems being assessed for a TB CHIM include skin detection of fluorescent proteins, serum detection of metabolites engineered to be expressed by the Mtb challenge strain, and breath detection of engineered volatile organic compounds.\n\nMany questions regarding a CHIM for TB vaccine development remain to be answered, questions that will confront both CHIM developers and regulators tasked with assuring the safety of the mycobacterial control and detection strategies ultimately selected for the initial TB CHIM. Safety is paramount; questions regarding the confidence in the level of safeguards built into the Mtb challenge strain, and the potential for reversion to a virulent state, will need to be addressed. The duration of survival of a CHIM Mtb strain necessary to permit an adequate assessment of potential vaccine protection, and the level of sensitivity of detection that would need to be built into the system to permit these assessments remains unclear. Whether or not a vaccine effect can actually be detected via a CHIM strategy, and, if so, the types of TB vaccine for which this approach would be relevant, also remains unclear. Finally, the likelihood of regulatory acceptance of any Mtb CHIM is not yet known.\n\nGiven these uncertainties, a different CHIM strategy, based on a human intradermal (ID) BCG challenge, is being developed in parallel to CHIMs based on intrapulmonary Mtb administration. The rationale behind the ID BCG CHIM is based on the theory that an effective vaccine against BCG also should protect against Mtb. BCG vaccination has been found to suppress growth of an ID BCG challenge in mice, cattle and NHPs22,129,130. A pilot BCG challenge study has been conducted in which volunteers freshly vaccinated with BCG, and those who were BCG naïve, were administered BCG ID, with punch biopsies obtained at 1, 2 or 4 weeks following ID BCG administration23. PCR and culture assessments of the biopsy specimens demonstrated a small but significant degree of protection. A similar study, in which MVA85A and BCG + MVA85A was assessed in addition to persons vaccinated with BCG alone, also revealed a statistically significant reduction of ID BCG as determined by culture and PCR of skin punch biopsies in persons receiving BCG or BCG + MVA85A as compared to placebo or MVA85A alone131.\n\nEfforts are currently underway to develop a human AE BCG challenge model. Key issues in this effort are the safety and tolerability of BCG administered via AE and the ability to recover BCG in the bronchoalveolar lavage fluid (BALF). No safety concerns have been reported from the study to date. Early assessments suggest that the extent to which BCG can be recovered in BALF appears to be low, which may indicate the need to administer higher BCG doses.\n\nIn summary, work is ongoing to develop an attenuated, labelled Mtb strain for use in a human challenge study. An Mtb CHIM is approximately 3–5 years away from clinical use, and regulatory discussions regarding its development are ongoing. Work is progressing on ID and AE BCG challenge models. Overall, a CHIM to assess TB vaccine candidates is considered feasible. Ultimately, however, any CHIM developed for this purpose will require validation against field efficacy trials.\n\n\n6. Perspectives from funders and global health organizations working on TB vaccines\n\nThe BMGF represents the largest global funder of research and development efforts specifically directed at TB vaccine development. The BMGF approaches TB vaccine development from the perspective of four strategic objectives:\n\n1. To understand the natural immune response associated with protection against infection and disease: Efforts are underway to better understand the workings of the granuloma, the histopathologic signature of Mtb infection and disease132. Unexpected diversity of granulomas has been shown in infected NHPs, even within the same animal: some granulomas are sterile while others appear permissive for Mtb survival21,133,134. Ongoing research seeks to understand the immunological mechanisms underlying permissive and non-permissive granulomas, and correlating results with those found in human granulomas obtained from surgical resections. Prospective cohort studies focus on outlier populations, such as persons who do not develop Mtb infection despite presumed prolonged exposure to household contacts with active TB disease. Differences in host control between these persons and individuals who become infected are being investigated. Smaller experimental medicine studies will address specific immunological hypotheses. Knowledge generated from these efforts should inform rational vaccine design.\n\n2. To develop new vaccine concepts that exploit immunological diversity: The focus is on vaccination approaches that target unconventional (non-Th1) immunity82 and /or target the lung directly (e.g., mucosal vaccination)122. In addition to delineating immunity that occurs in the natural infection setting, induction of “unnatural immunity,” will also be explored, i.e., immune responses not induced during natural infection and disease. An example of a vaccine that fits this strategy is the CMV-vectored TB vaccine, being developed by Louis Picker and colleagues at the Oregon Health Sciences University (OHSU), with major support from the BMGF57. Other examples of unconventional immunity targeted include those mediated by antibodies or B cells70, CD-1 restricted T cells135, and MAIT cells136.\n\n3. To develop improved tools and infrastructure to support an efficient, iterative process to test vaccine concepts: The BMGF continues to refine the paradigm for testing vaccine candidates, to more efficiently test candidates in human studies, to use improved animal models for up selection of candidates, and to improve learning along the experimental pathway. The BMGF supports development of improved NHP models, primarily, in support of this goal.\n\n4. To foster greater innovation, collaboration and coordination within the TB vaccine landscape: In support of this the BMGF has created the Collaboration for TB Vaccine Discovery (CTVD), an international network of scientists and experts dedicated to fostering innovation, cooperation, and collaboration in the up-stream TB vaccine discovery space. Further activities include improved alignment with other funders and exploring global portfolio management for clinical testing of vaccines through the use of stage gating.\n\nTB vaccine discovery is in an early stage, lagging drug discovery efforts by at least 10 years. Accordingly, the BMGF is putting increased focused on the upstream space with the intent of learning more about Mtb pathogenesis, and the immune responses necessary to control TB disease and prevent established Mtb infection. This information will help guide more rational and efficient TB vaccine development in the future.\n\nThe Global TB Vaccine Partnership (GTBVP) is a five-year-old initiative which is comprised of the leading organizations conducting and supporting TB vaccine research and development. The goals of the GTBVP are to enhance communication between key members of the TB vaccine R&D community, as well as between members of the community and the public at large—particularly public and private decision-makers responsible for allocating resources in support of critical global health issues – in an effort to attract new funding to this under-resourced initiative. The GTBVP is made up of a leadership forum, a technical advisory group and a communications group. Under the direction of the GTBVP, a document describing the current state of the TB vaccine development effort, as well as future directions in TB vaccine R&D, is being drafted. Current plans call for this document to be revised every two years to reflect the rapid changes occurring in this dynamic field.\n\nTB vaccine R&D efforts are extremely resource constrained, with the burden falling on a few donors. The human impact of the disease in terms of suffering and death, along with the $19.2 billion annual cost in treatment, care and loss of productivity, is far out of balance with the small and inadequate investment currently being made in TB vaccine development and testing.\n\nA lack of balance also exists in the nature of the partners involved in TB vaccine R&D, as support from non-profit organizations, such as the BMGF, and governments far outweigh the involvement of commercial entities such as established pharmaceutical companies and smaller biotech companies. Additionally, the shortage of funding has resulted in a lack of academic partners and new researchers entering the field. Moving forward, it will be important to create incentives for industry to increase activity in TB vaccine development initiatives.\n\nBalance also needs to be achieved between support of “upstream” (basic research and discovery) and “downstream” (assessments of vaccine concepts and candidates in animal challenge models and human clinical trials) efforts. Increasingly, the scarce research dollars that exist are flowing upstream at the expense of support for downstream initiatives, particularly clinical trials. It will be critical to continue to move forward with animal challenge studies and clinical trials utilizing a diversity of vaccine candidates and vaccine strategies, with a commitment to challenging dogma. It will be important to recognize, however, that TB vaccine development is challenging, and many efforts will lead to failure. Failure must be accepted if success is to be achieved.\n\nThe European and Developing Countries Clinical Trials Partnership (EDCTP) applauds the collaborative spirit that infuses the TB vaccine development field. An example of this is the collaboration between TuBerculosis Vaccine Initiative (TBVI), Aeras, the European Commission (EC), BMGF and EDCTP in developing stage gate criteria, designed to promote an efficient and coordinated approach to managing the TB vaccine portfolio by selecting the best TB vaccine candidates and advancing them through development in an optimized and non-redundant fashion.\n\nIt is imperative to continue conducting clinical trials of TB vaccine concepts and candidates. To further this effort, the EDCTP will be supporting a number of clinical trials of TB vaccines in the upcoming years. Additionally, the EDCTP will help support development of a human challenge model for TB, given the potential utility of such a model in driving the TB vaccine development effort forward.\n\nThe Stop TB Partnership exists to support the WHO in its efforts to eradicate TB. Key goals of the Partnership are to strengthen the capacity and research literacy of advocates, and to strengthen the advocacy literacy of researchers. Developing advocates for TB vaccine development must represent an essential component of any strategy geared to increasing funding for TB vaccine R&D efforts.\n\nThe TuBerculosis Vaccine Initiative (TBVI) is a non-profit foundation that facilitates the discovery and development of new, safe and effective TB vaccines that are accessible and affordable for all people. As a Product Development Partnership (PDP), TBVI integrates, translates and prioritises R&D efforts to discover and develop new TB vaccines and biomarkers for global use. TBVI provides essential services that support the R&D efforts of its consortium partners—50 partners from academia, research institutes and private industry in the TB vaccine field. A realistic strategy for TB vaccine development includes support for both upstream research and downstream clinical trials. It is imperative that a robust, healthy pipeline of TB vaccine concepts and candidates be maintained if the field is to move forward. The stage gating initiative, designed to encourage optimal management of the global TB vaccine portfolio, represents an important initiative in furthering this goal.\n\nProjects stages, stage gates, and stage-gate criteria are a suite of tools that allow effective management of large R&D programs by sponsors and funders. An R&D program can be divided into segments referred to as stages, wherein multiple mostly related activities are conducted in parallel. Stage gates are defined as check points that separate the stages and the stage-gate criteria constitute a set of objective, pre-defined data targets upon which continuation of the project into the next stage is decided. A stage-gating tool has been in use by TBVI and AERAS since 2010137.\n\nThe BMGF is now funding a new joint TBVI-Aeras effort to revise the general TB vaccine stage-gate criteria that will aid vaccine researchers and developers as well as develop more specific stage- gating criteria based on vaccine target populations and/or vaccine indications.\n\nThe stage-gate criteria will be revisited biannually to ensure that the latest knowledge and developments relevant to the development of TB vaccines are included. The development process will include consultation with stakeholders participating in TB vaccine development.\n\nFour different indications for TB vaccines will be addressed under the stage-gate criteria: 1) prevention of TB disease in pediatric, adolescent and adult/elderly populations; 2) BCG replacement in infants; 3) prevention of TB disease recurrence; and 4) immunotherapy of active TB disease. Stage gating criteria for these indications may differ in pre-clinical immunogenicity, efficacy and safety requirements; in clinical development strategy, including safety, immunogenicity and efficacy targets, and target populations; and in regulatory and marketing targets. Commonalities between indications will be found in process and manufacturing criteria, and well as in the overall precut characteristics. These differences and commonalities ultimately will be reflected in the stage gating criteria. Efforts towards the finalization of the stage gating document are underway.\n\n\n7. Conclusions\n\nAs the TB vaccine development field moves forward, it will be important to educate potential funders as to the barriers faced in conducting clinical trials of TB vaccines and to clearly frame the specific goals of these trials. The WHO holds an important responsibility for bringing researchers and funders together in support of needed efforts.\n\nThe WHO has provided excellent leadership through its introduction of preferred product criteria (PPC) for TB vaccine development5. Importantly, the criteria appear to be flexible enough to keep open many directions in TB vaccine research and avoiding the trap of focusing on the wrong strategic priorities. While the WHO is presenting different PPCs for TB vaccines targeting infants and adolescents/adults, respectively, it will be important to model the effect of using adult and infant vaccines together to end the TB epidemic.\n\nThis is a critically important juncture for TB vaccine development. Results from the recently concluded BCG revaccination study in South African adolescents at high risk of acquiring Mtb infection have reemphasized the potential promise of looking at this old vaccine in new ways12. M72/AS01E candidate vaccine results, published after the meeting reported here, constitute a major encouragement for the field, and argue in favor of strengthening global health stakeholder engagement and funding for further TB vaccine R&D efforts138. Ongoing, robust support for TB vaccine development will be critical to ensuring that a TB vaccine is available to help end the global scourge of tuberculosis.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "Author contributions\n\n\n\nJ.V. conceived the work. L.K.S. and J.V. organized the WHO consultation. L.K.S. prepared the first draft of the manuscript, based on the meeting presentations (see Acknowledgement section) and an extended literature review. L.K.S., R.C.H. and J.V. were involved in the revision and finalization of the manuscript and have agreed to the final content.\n\n\nGrant information\n\nThis work was supported in part by a grant to WHO from the Bill & Melinda Gates Foundation: OPP1160685.\n\nThe funders had no role in study design, data collection and analysis and decision to publish.\n\n\nAcknowledgements\n\nThe various sections were generated through a review of the scientific literature extended from presentations provided respectively by the following individuals, as follows. Introduction: Paul Henri Lambert (Geneva University, Geneva, Switzerland), Barry Bloom (Harvard TH Chan School of Public Health, Boston, USA), Martin Friede (WHO, Geneva, Switzerland); The WHO Global TB Programme: Christian Lienhardt (WHO, Geneva, Switzerland), Mario Raviglione (WHO, Geneva, Switzerland); The Value Proposition for TB Vaccines: Jacqueline Shea (Aeras, Maryland, USA); Overview of the Tuberculosis Vaccine Clinical Pipeline: Ann Ginsberg (Aeras, Maryland, USA); Modeling impact according to vaccine profile: Rebecca Harris (London School of Hygiene and Tropical Medicine, London, UK); The BCG Vaccine’s Profile and Role in Public Health: Barry Bloom; VPM1002: Umesh Shaligram (Serum Institute of India Limited, Pune, India); MTBVAC: Carlos Martin (University of Zaragoza, Zaragoza, Spain); VPM1002: Umesh Shaligram; M72/AS01E: Oliver Van Der Meeren (GSK, Brussels, Belgium); H56/IC31: Morten Ruhwald (Statens Serum Institute, Copenhagen, Denmark); ID93/GLA-SE: Corey Casper (Infectious Disease Research Institute, Seattle, USA); RUTI ™ for Adjunctive Immunotherapy: Pere Joan Cardona (Institut Germans Trias i Pujol, Barcelona, Spain); Vaccae™: Lu Jinbiao (National Institutes for Food and Drug Control, Beijing, China); Importance of immune characterization: Barry Bloom and Paul Henri Lambert; Cytomegalovirus Recombinant Vaccines Against TB: Louis Picker (Oregon Sciences University, Portland, USA); mRNA-Based Vaccines: Danilo Casimiro (Aeras, Maryland, USA); New Candidates for Live Attenuated Tuberculosis Vaccines: Adel Talaat (University of Wisconsin-Madison, Madison, USA); A Systematic Antigen Discovery Approach in Humans: Tom H.M. Ottenhoff (University of Leiden, Leiden, The Netherlands); Antibody Generating Vaccines for TB: Galit Alter (Ragon Institute, Cambridge, USA); Diverse T-Cell Responses Against MTB: David Lewinsohn (Oregon Health and Science University, Portland, USA); Immune Correlates and Signatures of Protection: Stefan Kaufmann (Max Planck Institute for Infection Biology, Berlin, Germany); Small Animal Models: Andrea Cooper University of Leicester, Leicester, UK); The Role of a Continuous Exposure Natural Transmission Animal Model for Vaccine Development: Edward Nardell (Harvard TH Chan Public School of Health, Boston, USA); Simian Models of Tuberculosis: Sally Sharpe (Public Health England, London, UK); Achieving Sterilizing Immunity in Animal Challenge Models: Shabaana Khader (Washington University, Saint Louis, USA); Mucosal immunization against TB: Helen McShane (Oxford University, Oxford, UK); Alternative Routes of BCG Immunization: Robert Seder (National Institute of Health, Maryland, USA); Progress in clinical use of a controlled human infection model for vaccine testing: Helen McShane; Perspectives from the Bill and Melinda Gates Foundation: Willem Hanekom (Bill and Melinda Gates Foundation, Seattle, USA); The Global TB Vaccine Partnership: Nick Drager (Tuberculosis Vaccine Initiative, Lelystad, The Netherlands); Aeras: Jacqueline Shea; European and developing countries clinical trials partnership (EDCTP): Ole Olesen (EDCTP, The Hague, The Netherlands); The Stop TB Partnership, Working Group on New TB Vaccines: David Lewinsohn; The Tuberculosis Vaccine initiative: Gerald Voss (Tuberculosis Vaccine Initiative, Lelystad, The Netherlands); Proposed Stage Gate Criteria for TB Vaccines: Danilo Casimiro, Georges Thiry (Tuberculosis Vaccine Initiative, Lelystad, The Netherlands); Conclusions: Barry Bloom, Paul Henri Lambert.\n\nWe are grateful for the contribution provided by all the individuals who attended the consultation and contributed to the discussions. The contents of this report are those of the authors and do not necessarily represent the decisions or the policies of the World Health Organization.\n\n\nReferences\n\nOrganisation mondiale de la santé: Global tuberculosis report 2017. 2017. Reference Source\n\nHeimbeck JJ: Incidence of tuberculosis in young adult women, with special reference to employment. British Journal of Tuberculosis. 1938; 32(3): 154–6. Publisher Full Text\n\nAndrews JR, Noubary F, Walensky RP, et al.: Risk of progression to active tuberculosis following reinfection with Mycobacterium tuberculosis. Clin Infect Dis. 2012 ; 54(6): 784–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVoss G, Casimiro D, Neyrolles O, et al.: Progress and challenges in TB vaccine development [version 1; referees: 2 approved]. F1000Res. 2018; 7: 199. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchrager LK, Chandrasekaran P, Fritzell BH, et al.: WHO preferred product characteristics for new vaccines against tuberculosis. Lancet Infect Dis. 2018; 18(8): 828–9. PubMed Abstract | Publisher Full Text\n\nWHO End TB Strategy. 2015. Reference Source\n\nKnight GM, Griffiths UK, Sumner T, et al.: Impact and cost-effectiveness of new tuberculosis vaccines in low- and middle-income countries. Proc Natl Acad Sci U S A. 2014; 111(43): 15520–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarris RC, Sumner T, Knight GM, et al.: Systematic review of mathematical models exploring the epidemiological impact of future TB vaccines. Hum Vaccin Immunother. 2016; 12(11): 2813–32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMillet JP, Shaw E, Orcau A, et al.: Tuberculosis recurrence after completion treatment in a European city: reinfection or relapse? PLoS One. 2013; 8(6): e64898. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNunn AJ, Phillips PP, Mitchison DA: Timing of relapse in short-course chemotherapy trials for tuberculosis. Int J Tuberc Lung Dis. 2010; 14(2): 241–2. PubMed Abstract\n\nMahomed H, Ehrlich R, Hawkridge T, et al.: TB incidence in an adolescent cohort in South Africa. PLoS One. 2013; 8(3): e59652. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNemes E, Geldenhuys H, Rozot V, et al.: Prevention of M. tuberculosis Infection with H4:IC31 Vaccine or BCG Revaccination. N Engl J Med. 2018; 379(2): 138–49. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorld Health Organization: BCG vaccine: WHO position paper, February 2018 - Recommendations. Vaccine. 2018; 36(24): 3408–10. PubMed Abstract | Publisher Full Text\n\nAndrews JR, Hatherill M, Mahomed H, et al.: The dynamics of QuantiFERON-TB gold in-tube conversion and reversion in a cohort of South African adolescents. Am J Respir Crit Care Med. 2015; 191(5): 584–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHawn TR, Day TA, Scriba TJ, et al.: Tuberculosis vaccines and prevention of infection. Microbiol Mol Biol Rev. 2014; 78(4): 650–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRiley RL, Mills CC, Nyka W, et al.: Aerial dissemination of pulmonary tuberculosis. A two-year study of contagion in a tuberculosis ward. 1959. Am J Epidemiol. 1995; 142(1): 3–14. PubMed Abstract | Publisher Full Text\n\nDharmadhikari AS, Basaraba RJ, Van Der Walt ML, et al.: Natural infection of guinea pigs exposed to patients with highly drug-resistant tuberculosis. Tuberculosis (Edinb). 2011; 91(4): 329–38. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHall NB, Igo RP Jr, Malone LL, et al.: Polymorphisms in TICAM2 and IL1B are associated with TB. Genes Immun. 2015; 16(2): 127–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLerm M, Netea MG: Trained immunity: a new avenue for tuberculosis vaccine development. J Intern Med. 2016; 279(4): 337–46. PubMed Abstract | Publisher Full Text\n\nColeman MT, Maiello P, Tomko J, et al.: Early Changes by 18Fluorodeoxyglucose positron emission tomography coregistered with computed tomography predict outcome after Mycobacterium tuberculosis infection in cynomolgus macaques. Infect Immun. 2014; 82(6): 2400–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartin CJ, Cadena AM, Leung VW, et al.: Digitally Barcoding Mycobacterium tuberculosis Reveals In Vivo Infection Dynamics in the Macaque Model of Tuberculosis. mBio. 2017; 8(3): pii: e00312-17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMinassian AM, Ronan EO, Poyntz H, et al.: Preclinical development of an in vivo BCG challenge model for testing candidate TB vaccine efficacy. PLoS One. 2011; 6(5): e19840. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMinassian AM, Satti I, Poulton ID, et al.: A human challenge model for Mycobacterium tuberculosis using Mycobacterium bovis bacille Calmette-Guerin. J Infect Dis. 2012; 205(7): 1035–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarris RC, Sumner T, Knight GM, et al.: New tuberculosis vaccines – modelling the impact of age targeted vaccination in China and implications for vaccine development. Lancet Glob Health. (In Press).\n\nMangtani P, Abubakar I, Ariti C, et al.: Protection by BCG vaccine against tuberculosis: a systematic review of randomized controlled trials. Clin Infect Dis. 2014; 58(4): 470–80. PubMed Abstract | Publisher Full Text\n\nRoy A, Eisenhut M, Harris RJ, et al.: Effect of BCG vaccination against Mycobacterium tuberculosis infection in children: systematic review and meta-analysis. BMJ. 2014; 349: g4643. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAnacker RL, Brehmer W, Barclay WR, et al.: Superiority of intravenously administered BCG and BCG cell walls in protecting rhesus monkeys (Macaca mulatta) against airborne tuberculosis. Z Immunitatsforsch Exp Klin Immunol. 1972; 143(4): 363–76. PubMed Abstract\n\nWhite AD, Sarfas C, West K, et al.: Evaluation of the Immunogenicity of Mycobacterium bovis BCG Delivered by Aerosol to the Lungs of Macaques. Clin Vaccine Immunol. 2015; 22(9): 992–1003. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSharpe S, White A, Sarfas C, et al.: Alternative BCG delivery strategies improve protection against Mycobacterium tuberculosis in non-human primates: Protection associated with mycobacterial antigen-specific CD4 effector memory T-cell populations. Tuberculosis (Edinb). 2016; 101: 174–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSatti I, Meyer J, Harris SA, et al.: Safety and immunogenicity of a candidate tuberculosis vaccine MVA85A delivered by aerosol in BCG-vaccinated healthy adults: a phase 1, double-blind, randomised controlled trial. Lancet Infect Dis. 2014; 14(10): 939–46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHesseling AC, Johnson LF, Jaspan H, et al.: Disseminated bacille Calmette-Guérin disease in HIV-infected South African infants. Bull World Health Organ. 2009; 87(7): 505–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNieuwenhuizen NE, Kulkarni PS, Shaligram U, et al.: The Recombinant Bacille Calmette-Guérin Vaccine VPM1002: Ready for Clinical Efficacy Testing. Front Immunol. 2017; 8: 1147. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPortaels F, Aguiar J, Debacker M, et al.: Mycobacterium bovis BCG vaccination as prophylaxis against Mycobacterium ulcerans osteomyelitis in Buruli ulcer disease. Infect Immun. 2004; 72(1): 62–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPhillips RO, Phanzu DM, Beissner M, et al.: Effectiveness of routine BCG vaccination on buruli ulcer disease: a case-control study in the Democratic Republic of Congo, Ghana and Togo. PLoS Negl Trop Dis. 2015; 9(1): e3457. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHiggins JP, Soares-Weiser K, López-López JA, et al.: Association of BCG, DTP, and measles containing vaccines with childhood mortality: systematic review. BMJ. 2016; 355: i5170. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoehm BE, Cornell JE, Wang H, et al.: Efficacy of bacillus Calmette-Guérin Strains for Treatment of Nonmuscle Invasive Bladder Cancer: A Systematic Review and Network Meta-Analysis. J Urol. 2017; 198(3): 503–10. PubMed Abstract | Publisher Full Text\n\nArbues A, Aguilo JI, Gonzalo-Asensio J, et al.: Construction, characterization and preclinical evaluation of MTBVAC, the first live-attenuated M. tuberculosis-based vaccine to enter clinical trials. Vaccine. 2013; 31(42): 4867–73. PubMed Abstract | Publisher Full Text\n\nGonzalo-Asensio J, Marinova D, Martin C, et al.: MTBVAC: Attenuating the Human Pathogen of Tuberculosis (TB) Toward a Promising Vaccine against the TB Epidemic. Front Immunol. 2017; 8: 1803. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAguilo N, Gonzalo-Asensio J, Alvarez-Arguedas S, et al.: Reactogenicity to major tuberculosis antigens absent in BCG is linked to improved protection against Mycobacterium tuberculosis. Nat Commun. 2017; 8: 16085. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpertini F, Audran R, Chakour R, et al.: Safety of human immunisation with a live-attenuated Mycobacterium tuberculosis vaccine: a randomised, double-blind, controlled phase I trial. Lancet Respir Med. 2015; 3(12): 953–62. PubMed Abstract | Publisher Full Text\n\nClark S, Lanni F, Marinova D, et al.: Revaccination of Guinea Pigs With the Live Attenuated Mycobacterium tuberculosis Vaccine MTBVAC Improves BCG’s Protection Against Tuberculosis. J Infect Dis. 2017; 216(5): 525–33. PubMed Abstract | Publisher Full Text\n\nDay CL, Tameris M, Mansoor N, et al.: Induction and regulation of T-cell immunity by the novel tuberculosis vaccine M72/AS01 in South African adults. Am J Respir Crit Care Med. 2013; 188(4): 492–502. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMontoya J, Solon JA, Cunanan SR, et al.: A randomized, controlled dose-finding Phase II study of the M72/AS01 candidate tuberculosis vaccine in healthy PPD-positive adults. J Clin Immunol. 2013; 33(8): 1360–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKumarasamy N, Poongulali S, Bollaerts A, et al.: A Randomized, Controlled Safety, and Immunogenicity Trial of the M72/AS01 Candidate Tuberculosis Vaccine in HIV-Positive Indian Adults. Medicine (Baltimore). 2016; 95(3): e2459. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThacher EG, Cavassini M, Audran R, et al.: Safety and immunogenicity of the M72/AS01 candidate tuberculosis vaccine in HIV-infected adults on combination antiretroviral therapy: a phase I/II, randomized trial. AIDS. 2014; 28(12): 1769–81. PubMed Abstract | Publisher Full Text\n\nIdoko OT, Owolabi OA, Owiafe PK, et al.: Safety and immunogenicity of the M72/AS01 candidate tuberculosis vaccine when given as a booster to BCG in Gambian infants: an open-label randomized controlled trial. Tuberculosis (Edinb). 2014; 94(6): 564–78. PubMed Abstract | Publisher Full Text\n\nVan Der Meeren O, Hatherill M, Nduba V, et al.: Phase 2b Controlled Trial of M72/AS01E Vaccine to Prevent Tuberculosis. N Engl J Med. 2018. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLuabeya AK, Kagina BM, Tameris MD, et al.: First-in-human trial of the post-exposure tuberculosis vaccine H56:IC31 in Mycobacterium tuberculosis infected and non-infected healthy adults. Vaccine. 2015; 33(33): 4130–40. PubMed Abstract | Publisher Full Text\n\nPenn-Nicholson A, Tameris M, Smit E, et al.: Safety and immunogenicity of the novel tuberculosis vaccine ID93 + GLA-SE in BCG-vaccinated healthy adults in South Africa: a randomised, double-blind, placebo-controlled phase 1 trial. Lancet Respir Med. 2018; 6(4): 287–98. PubMed Abstract | Publisher Full Text\n\nColer RN, Bertholet S, Pine SO, et al.: Therapeutic immunization against Mycobacterium tuberculosis is an effective adjunct to antibiotic treatment. J Infect Dis. 2013; 207(8): 1242–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBertholet S, Ireton GC, Ordway DJ, et al.: A defined tuberculosis vaccine candidate boosts BCG and protects against multidrug-resistant Mycobacterium tuberculosis. Sci Transl Med. 2010; 2(53): 53ra74. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNell AS, D’lom E, Bouic P, et al.: Safety, tolerability, and immunogenicity of the novel antituberculous vaccine RUTI: randomized, placebo-controlled phase II clinical trial in patients with latent tuberculosis infection. PLoS One. 2014; 9(2): e89612. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVilaplana C, Montané E, Pinto S, et al.: Double-blind, randomized, placebo-controlled Phase I Clinical Trial of the therapeutical antituberculous vaccine RUTI. Vaccine. 2010; 28(4): 1106–16. PubMed Abstract | Publisher Full Text\n\nYang XY, Chen QF, Li YP, et al.: Mycobacterium vaccae as adjuvant therapy to anti-tuberculosis chemotherapy in never-treated tuberculosis patients: a meta-analysis. PLoS One. 2011; 6(9): e23826. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLewinsohn DA, Lewinsohn DM, Scriba TJ: Polyfunctional CD4+ T Cells As Targets for Tuberculosis Vaccination. Front Immunol. 2017; 8: 1262. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHershkovitz I, Donoghue HD, Minnikin DE, et al.: Detection and molecular characterization of 9,000-year-old Mycobacterium tuberculosis from a Neolithic settlement in the Eastern Mediterranean. PLoS One. 2008; 3(10): e3426. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHansen SG, Zak DE, Xu G, et al.: Prevention of tuberculosis in rhesus macaques by a cytomegalovirus-based vaccine. Nat Med. 2018; 24(2): 130–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHansen SG, Sacha JB, Hughes CM, et al.: Cytomegalovirus vectors violate CD8+ T cell epitope recognition paradigms. Science. 2013; 340(6135): 1237874. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrito LA, Kommareddy S, Maione D, et al.: Self-amplifying mRNA vaccines. In: Adv Genet. Elsevier; 2015; 89: 179–233. PubMed Abstract | Publisher Full Text\n\nPlotkin SA: Vaccines: past, present and future. Nat Med. 2005; 11(4 Suppl): S5–11. PubMed Abstract | Publisher Full Text\n\nAbomoelak B, Hoye EA, Chi J, et al.: mosR, a novel transcriptional regulator of hypoxia and virulence in Mycobacterium tuberculosis. J Bacteriol. 2009; 191(19): 5941–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarcus SA, Steinberg H, Talaat AM: Protection by novel vaccine candidates, Mycobacterium tuberculosis ΔmosR and ΔechA7, against challenge with a Mycobacterium tuberculosis Beijing strain. Vaccine. 2015; 33(42): 5633–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJoosten SA, Sullivan LC, Ottenhoff TH: Characteristics of HLA-E Restricted T-Cell Responses and Their Role in Infectious Diseases. J Immunol Res. 2016 ; 2016: 2695396. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTameris MD, Hatherill M, Landry BS, et al.: Safety and efficacy of MVA85A, a new tuberculosis vaccine, in infants previously vaccinated with BCG: a randomised, placebo-controlled phase 2b trial. Lancet. 2013; 381(9871): 1021–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCommandeur S, van Meijgaarden KE, Prins C, et al.: An unbiased genome-wide Mycobacterium tuberculosis gene expression approach to discover antigens targeted by human T cells expressed during pulmonary infection. J Immunol. 2013; 190(4): 1659–71. PubMed Abstract | Publisher Full Text\n\nCommandeur S, van Meijgaarden KE, Lin MY, et al.: Identification of human T-cell responses to Mycobacterium tuberculosis resuscitation-promoting factors in long-term latently infected individuals. Clin Vaccine Immunol. 2011; 18(4): 676–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCoppola M, van Meijgaarden KE, Franken KL, et al.: New Genome-Wide Algorithm Identifies Novel In-Vivo Expressed Mycobacterium Tuberculosis Antigens Inducing Human T-Cell Responses with Classical and Unconventional Cytokine Profiles. Sci Rep. 2016; 6(1): 37793. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSakai S, Kauffman KD, Sallin MA, et al.: CD4 T Cell-Derived IFN-γ Plays a Minimal Role in Control of Pulmonary Mycobacterium tuberculosis Infection and Must Be Actively Repressed by PD-1 to Prevent Lethal Disease. PLoS Pathog. 2016; 12(5): e1005667. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOrlando V, La Manna MP, Goletti D, et al.: Human CD4 T-Cells With a Naive Phenotype Produce Multiple Cytokines During Mycobacterium Tuberculosis Infection and Correlate With Active Disease. Front Immunol. 2018; 9: 1119. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAchkar JM, Chan J, Casadevall A: B cells and antibodies in the defense against Mycobacterium tuberculosis infection. Immunol Rev. 2015; 264(1): 167–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaglione PJ, Xu J, Chan J: B cells moderate inflammatory progression and enhance bacterial containment upon pulmonary challenge with Mycobacterium tuberculosis. J Immunol. 2007; 178(11): 7222–34. PubMed Abstract | Publisher Full Text\n\nMaglione PJ, Xu J, Casadevall A, et al.: Fc gamma receptors regulate immune activation and susceptibility during Mycobacterium tuberculosis infection. J Immunol. 2008; 180(5): 3329–38. PubMed Abstract | Publisher Full Text\n\nPhuah JY, Mattila JT, Lin PL, et al.: Activated B cells in the granulomas of nonhuman primates infected with Mycobacterium tuberculosis. Am J Pathol. 2012; 181(2): 508–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJoosten SA, van Meijgaarden KE, Del Nonno F, et al.: Patients with Tuberculosis Have a Dysfunctional Circulating B-Cell Compartment, Which Normalizes following Successful Treatment. PLoS Pathog. 2016; 12(6): e1005687. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTorrado E, Fountain JJ, Robinson RT, et al.: Differential and site specific impact of B cells in the protective immune response to Mycobacterium tuberculosis in the mouse. PLoS One. 2013; 8(4): e61681. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHamasur B, Haile M, Pawlowski A, et al.: A mycobacterial lipoarabinomannan specific monoclonal antibody and its F(ab') fragment prolong survival of mice infected with Mycobacterium tuberculosis. Clin Exp Immunol. 2004; 138(1): 30–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWilliams A, Reljic R, Naylor I, et al.: Passive protection with immunoglobulin A antibodies against tuberculous early infection of the lungs. Immunology. 2004; 111(3): 328–33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPethe K, Alonso S, Biet F, et al.: The heparin-binding haemagglutinin of M. tuberculosis is required for extrapulmonary dissemination. Nature. 2001; 412(6843): 190–4. PubMed Abstract | Publisher Full Text\n\nZimmermann N, Thormann V, Hu B, et al.: Human isotype-dependent inhibitory antibody responses against Mycobacterium tuberculosis. EMBO Mol Med. 2016; 8(11): 1325–39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLawn SD, Zumla AI: Tuberculosis. Lancet. 2011; 378(9785): 57–72. PubMed Abstract | Publisher Full Text\n\nGuilliams M, Bruhns P, Saeys Y, et al.: The function of Fcγ receptors in dendritic cells and macrophages. Nat Rev Immunol. 2014; 14(2): 94–108. PubMed Abstract | Publisher Full Text\n\nGodfrey DI, Uldrich AP, McCluskey J, et al.: The burgeoning family of unconventional T cells. Nat Immunol. 2015; 16(11): 1114–23. PubMed Abstract | Publisher Full Text\n\nChua WJ, Truscott SM, Eickhoff CS, et al.: Polyclonal mucosa-associated invariant T cells have unique innate functions in bacterial infection. Infect Immun. 2012; 80(9): 3256–67. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUssher JE, Klenerman P, Willberg CB: Mucosal-associated invariant T-cells: new players in anti-bacterial immunity. Front Immunol. 2014; 5: 450. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGeorgel P, Radosavljevic M, Macquin C, et al.: The non-conventional MHC class I MR1 molecule controls infection by Klebsiella pneumoniae in mice. Mol Immunol. 2011; 48(5): 769–75. PubMed Abstract | Publisher Full Text\n\nMontamat-Sicotte DJ, Millington KA, Willcox CR, et al.: A mycolic acid-specific CD1-restricted T cell population contributes to acute and memory immune responses in human tuberculosis infection. J Clin Invest. 2011; 121(6): 2493–503. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrezzemolo T, van Meijgaarden KE, Franken KLMC, et al.: Detailed characterization of human Mycobacterium tuberculosis specific HLA-E restricted CD8+ T cells. Eur J Immunol. 2018; 48(2): 293–305. PubMed Abstract | Publisher Full Text\n\nJoosten SA, van Meijgaarden KE, van Weeren PC, et al.: Mycobacterium tuberculosis peptides presented by HLA-E molecules are targets for human CD8 T-cells with cytotoxic as well as regulatory activity. PLoS Pathog. 2010; 6(2): e1000782. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatel O, Kjer-Nielsen L, Le Nours J, et al.: Recognition of vitamin B metabolites by mucosal-associated invariant T cells. Nat Commun. 2013; 4: 2142. PubMed Abstract | Publisher Full Text\n\nHarriff MJ, McMurtrey C, Froyd CA, et al.: MR1 displays the microbial metabolome driving selective MR1-restricted T cell receptor usage. Sci Immunol. 2018; 3(25): pii: eaao2556. PubMed Abstract | Publisher Full Text\n\nYoung DC, Moody DB: T-cell recognition of glycolipids presented by CD1 proteins. Glycobiology. 2006; 16(7): 103R–112R. PubMed Abstract | Publisher Full Text\n\nLy D, Kasmar AG, Cheng TY, et al.: CD1c tetramers detect ex vivo T cell responses to processed phosphomycoketide antigens. J Exp Med. 2013; 210(4): 729–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Meijgaarden KE, Haks MC, Caccamo N, et al.: Human CD8+ T-cells recognizing peptides from Mycobacterium tuberculosis (Mtb) presented by HLA-E have an unorthodox Th2-like, multifunctional, Mtb inhibitory phenotype and represent a novel human T-cell subset. PLoS Pathog. 2015; 11(3): e1004671. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpencer CT, Abate G, Blazevic A, et al.: Only a subset of phosphoantigen-responsive gamma9delta2 T cells mediate protective tuberculosis immunity. J Immunol. 2008; 181(7): 4471–84. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpencer CT, Abate G, Sakala IG, et al.: Granzyme A produced by γ9δ2 T cells induces human macrophages to inhibit growth of an intracellular pathogen. PLoS Pathog. 2013; 9(1): e1003119. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhao J, Siddiqui S, Shang S, et al.: Mycolic acid-specific T cells protect against Mycobacterium tuberculosis infection in a humanized transgenic mouse model. eLife. 2015; 4: pii: e08525. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaertzdorf J, McEwen G, Weiner J 3rd, et al.: Concise gene signature for point-of-care classification of tuberculosis. EMBO Mol Med. 2016; 8(2): 86–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSweeney TE, Braviak L, Tato CM, et al.: Genome-wide expression for diagnosis of pulmonary tuberculosis: a multicohort analysis. Lancet Respir Med. 2016; 4(3): 213–24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZak DE, Penn-Nicholson A, Scriba TJ, et al.: A blood RNA signature for tuberculosis disease risk: a prospective cohort study. Lancet. 2016; 387(10035): 2312–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSuliman S, Thompson EG, Sutherland J, et al.: Four-gene Pan-African Blood Signature Predicts Progression to Tuberculosis. Am J Respir Crit Care Med. 2018; 197(9): 1198–208. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWeiner J 3rd, Parida SK, Maertzdorf J, et al.: Biomarkers of inflammation, immunosuppression and stress with active disease are revealed by metabolomic profiling of tuberculosis patients. PLoS One. 2012; 7(7): e40221. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFletcher HA, Snowden MA, Landry B, et al.: T-cell activation is an immune correlate of risk in BCG vaccinated infants. Nat Commun. 2016; 7: 11290. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFletcher HA, Filali-Mouhim A, Nemes E, et al.: Human newborn bacille Calmette-Guérin vaccination and risk of tuberculosis disease: a case-control study. BMC Med. 2016; 14: 76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaufmann SH, Evans TG, Hanekom WA: Tuberculosis vaccines: time for a global strategy. Sci Transl Med. 2015; 7(276): 276fs8. PubMed Abstract | Publisher Full Text\n\nJacobs WR Jr, McShane H, Mizrahi V, et al.editors: Mouse and Guinea Pig Models of Tuberculosis. In Tuberculosis and the Tubercle Bacillus, Second Edition. American Society of Microbiology. 2017; 143–62. Publisher Full Text\n\nWilliams A, Orme IM: Animal Models of Tuberculosis: An Overview. Microbiol Spectr. 2016; 4(4). PubMed Abstract | Publisher Full Text\n\nHoang T, Aagaard C, Dietrich J, et al.: ESAT-6 (EsxA) and TB10.4 (EsxH) based vaccines for pre- and post-exposure tuberculosis vaccination. PLoS One. 2013; 8(12): e80579. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSrivastava A, Morgan AP, Najarian ML, et al.: Genomes of the Mouse Collaborative Cross. Genetics. 2017; 206(2): 537–56. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith CM, Proulx MK, Olive AJ, et al.: Tuberculosis Susceptibility and Vaccine Protection Are Independently Controlled by Host Genotype. mBio. 2016; 7(5): pii: e01516-16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOrdway DJ, Shang S, Henao-Tamayo M, et al.: Mycobacterium bovis BCG-mediated protection against W-Beijing strains of Mycobacterium tuberculosis is diminished concomitant with the emergence of regulatory T cells. Clin Vaccine Immunol. 2011; 18(9): 1527–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeisching G, Pietersen RD, Wiid I, et al.: Virulence, biochemistry, morphology and host-interacting properties of detergent-free cultured mycobacteria: An update. Tuberculosis (Edinb). 2016; 100: 53–60. PubMed Abstract | Publisher Full Text\n\nGelmanova IY, Keshavjee S, Golubchikova VT, et al.: Barriers to successful tuberculosis treatment in Tomsk, Russian Federation: non-adherence, default and the acquisition of multidrug resistance. Bull World Health Organ. 2007; 85(9): 703–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFlynn JL, Gideon HP, Mattila JT, et al.: Immunology studies in non-human primate models of tuberculosis. Immunol Rev. 2015; 264(1): 60–73. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarry CE, Boshoff HI, Dartois V, et al.: The spectrum of latent tuberculosis: rethinking the biology and intervention strategies. Nat Rev Microbiol. 2009; 7(12): 845–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCapuano SV 3rd, Croix DA, Pawar S, et al.: Experimental Mycobacterium tuberculosis infection of cynomolgus macaques closely resembles the various manifestations of human M. tuberculosis infection. Infect Immun. 2003; 71(10): 5831–44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScanga CA, Flynn JL: Modeling tuberculosis in nonhuman primates. Cold Spring Harb Perspect Med. 2014; 4(12): a018564. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaiello P, DiFazio RM, Cadena AM, et al.: Rhesus Macaques Are More Susceptible to Progressive Tuberculosis than Cynomolgus Macaques: a Quantitative Comparison. Infect Immun. 2018; 86(2): pii: e00505-17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaddy DJ, Bonavia A, Hanekom WA, et al.: Toward Tuberculosis Vaccine Development: Recommendations for Nonhuman Primate Study Design. Infect Immun. 2017; 86(2): pii: e00776-17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKhader SA, Bell GK, Pearl JE, et al.: IL-23 and IL-17 in the establishment of protective pulmonary CD4+ T cell responses after vaccination and during Mycobacterium tuberculosis challenge. Nat Immunol. 2007; 8(4): 369–77. PubMed Abstract | Publisher Full Text\n\nGriffiths KL, Ahmed M, Das S, et al.: Targeting dendritic cells to accelerate T-cell activation overcomes a bottleneck in tuberculosis vaccine efficacy. Nat Commun. 2016; 7: 13894. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaushal D, Foreman TW, Gautam US, et al.: Mucosal vaccination with attenuated Mycobacterium tuberculosis induces strong central memory responses and protects against tuberculosis. Nat Commun. 2015; 6(1): 8533. PubMed Abstract | Publisher Full Text | Free Full Text\n\nManjaly Thomas ZR, McShane H: Aerosol immunisation for TB: matching route of vaccination to route of infection. Trans R Soc Trop Med Hyg. 2015; 109(3): 175–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSeder RA, Chang LJ, Enama ME, et al.: Protection against malaria by intravenous immunization with a nonreplicating sporozoite vaccine. Science. 2013; 341(6152): 1359–65. PubMed Abstract | Publisher Full Text\n\nVerreck FAW, Tchilian EZ, Vervenne RAW, et al.: Variable BCG efficacy in rhesus populations: Pulmonary BCG provides protection where standard intra-dermal vaccination fails. Tuberculosis (Edinb). 2017; 104: 46–57. PubMed Abstract | Publisher Full Text\n\nPerdomo C, Zedler U, Kühl AA, et al.: Mucosal BCG Vaccination Induces Protective Lung-Resident Memory T Cell Populations against Tuberculosis. mBio. 2016; 7(6): pii: e01686-16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHokey DA, Wachholder R, Darrah PA, et al.: A nonhuman primate toxicology and immunogenicity study evaluating aerosol delivery of AERAS-402/Ad35 vaccine: Evidence for transient t cell responses in peripheral blood and robust sustained responses in the lungs. Hum Vaccin Immunother. 2014; 10(8): 2199–210. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarclay WR, Anacker RL, Brehmer W, et al.: Aerosol-Induced Tuberculosis in Subhuman Primates and the Course of the Disease After Intravenous BCG Vaccination. Infect Immun. 1970; 2(5): 574–82. PubMed Abstract | Free Full Text\n\nBarclay WR, Busey WM, Dalgard DW, et al.: Protection of monkeys against airborne tuberculosis by aerosol vaccination with bacillus Calmette-Guerin. Am Rev Respir Dis. 1973; 107(3): 351–8. PubMed Abstract\n\nVillarreal-Ramos B, Berg S, Chamberlain L, et al.: Development of a BCG challenge model for the testing of vaccine candidates against tuberculosis in cattle. Vaccine. 2014; 32(43): 5645–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarris SA, White A, Stockdale L, et al.: Development of a non-human primate BCG infection model for the evaluation of candidate tuberculosis vaccines. Tuberculosis (Edinb). 2018; 108: 99–105. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarris SA, Meyer J, Satti I, et al.: Evaluation of a human BCG challenge model to assess antimycobacterial immunity induced by BCG and a candidate tuberculosis vaccine, MVA85A, alone and in combination. J Infect Dis. 2014; 209(8): 1259–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarakalala MJ, Raju RM, Sharma K, et al.: Inflammatory signaling in human tuberculosis granulomas is spatially organized. Nat Med. 2016; 22(5): 531–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLin PL, Ford CB, Coleman MT, et al.: Sterilization of granulomas is common in active and latent tuberculosis despite within-host variability in bacterial killing. Nat Med. 2014; 20(1): 75–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGideon HP, Phuah J, Myers AJ, et al.: Variability in tuberculosis granuloma T cell responses exists, but a balance of pro- and anti-inflammatory cytokines is associated with sterilization. PLoS Pathog. 2015; 11(1): e1004603. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVan Rhijn I, Moody DB: CD1 and mycobacterial lipids activate human T cells. Immunol Rev. 2015; 264(1): 138–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGold MC, Napier RJ, Lewinsohn DM: MR1-restricted mucosal associated invariant T (MAIT) cells in the immune response to Mycobacterium tuberculosis. Immunol Rev. 2015; 264(1): 154–66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrennan PJ: A new TB vaccine blueprint. Tuberculosis (Edinb). 2012; 92 Suppl 1: S1. PubMed Abstract | Publisher Full Text\n\nBloom BR: New Promise for Vaccines against Tuberculosis. N Engl J Med. 2018. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "40107",
"date": "26 Nov 2018",
"name": "Gerald H. Voss",
"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\nThis review article provides a comprehensive and detailed summary of a WHO consultation on TB vaccine development. It is well structured, clearly written and based on the presentations during the consultation.\n\nThe article describes the need for new TB vaccines and provides an overview of the current clinical pipeline as well as considerations and modeling for anticipated impact. Subsequently, the different indications and populations that would benefit from new TB vaccines are described, notably BCG replacement and vaccines for adolescent and adult populations. An overview of new vaccine platforms, for example CMV, mRNA and whole cell vaccines, is provided. Then, the importance and usefulness of models for TB vaccine development, from rodents to controlled human challenge studies, is laid out. The final section provides perspectives from the main funders and other stakeholders in the field.\n\nImportantly, new developments, including clinical data on BCG revaccination and the M72/AS01E TB vaccine candidate have been incorporated to reflect the major recent advances in the field. This review is a comprehensive and useful reference for the current state of TB vaccine development.\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": "41391",
"date": "17 Dec 2018",
"name": "Stefan H.E. Kaufmann",
"expertise": [
"Reviewer Expertise Tuberculosis",
"immunology",
"vaccinology"
],
"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 provides a comprehensive summary of a WHO meeting on vaccine development and also includes reference to some relevant publications published afterwards. Unfortunately, the review is somehow unbalanced in several instances (see below) and in several cases uses specific immunologic details which are insufficiently explained and therefore may be hard to understand for a non-expert reader (see below). Moreover, in several cases, the article was written in a sloppy way (see below).\n\nSome comments (not at all comprehensive):\nP. 4, left column: Compare writing of US dollar here and p. 4, right column, 2nd paragraph. Be consistent. P. 4, right column, 1st paragraph, 2nd sentence: It doesn’t make sense to me the way it is written. P. 7, 2nd paragraph: What is a truly effective vaccine? A vaccine which protects everybody lifelong? Or a vaccine which protects everybody for a certain time? Or some for a certain time? P. 8, left column, 1st and 2nd paragraph: Need references P. 8, right column, 2nd paragraph: It is likely that BCG does not protect against bladder cancer, but is a therapeutic immunomodulator. P. 8, right column, 2nd last paragraph: VPM is being tested at different sites in Sub-Saharan Africa, not just South Africa. P. 9, left column, 1st paragraph: What does the discussion about epitopes really mean for vaccine efficacy? Does this assume that every epitope is equally important (dominant vs. subdominant)? Discussion of different vaccines: a) This whole section from page 8 to 12 to this reviewer appears quite imbalanced both in length and details of individual candidates. B) Wherever possible, registration numbers of trials should be included. C) For some candidates a very detailed description is being given which may be understandable for those in the field, but may be less so for outsiders. P. 30, 3rd paragraph: Another example for highly specific details which are not explained and therefore may be difficult to understand since text mentions all the different presentation molecules for T cells (CD1, MR1 etc.) without further explanation about their relative importance (e.g. compared to canonical MHC molecules). P. 13, right column, 2nd last paragraph: Role of antibodies in animal models: To this reviewer, several experimental animal models and experiments argue against a role for B cells. This topic would need to be discussed more carefully (this reviewer agrees that antibodies likely play a role in human TB). P. 15, right column, end of 2nd paragraph: A kind hint: The paper is now out (Weiner et al., Nat Commun, 2018) P. 16, right column, 2nd paragraph, Ref. 110: incorrect. Best to provide relevant website? P. 19: Discussion about memory T cells could benefit from the short description of the different subtypes and their biological function. P. 19, right column, 2nd paragraph: Is there a reference or is this a personal communication (from whom)? P. 20, left column, 2nd paragraph: Is there a reference or is this a personal communication (from whom)? P. 21, left column, last paragraph: Is there a website for GTBVP? P. 21, right column: paragraphs about Aeras: It might be worth to mention that this entity does not exist anymore (reviewer is aware that this happened after the meeting). Text has no direct relationship with Aeras, but rather describes strategies in general. P. 21, right column: EDCTP paragraphs: It would be helpful to mention TB vaccine trials supported by EDCTP (several!).\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? Partly\n\nAre the conclusions drawn appropriate in the context of the current research literature? Yes",
"responses": []
},
{
"id": "41392",
"date": "19 Dec 2018",
"name": "Roberto Zenteno-Cuevas",
"expertise": [
"Reviewer Expertise microbiology",
"molecular and genomic epidemiology of tuberculosis"
],
"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 aim of this review was to show the state of the art tuberculosis vaccine development, also describing the major problems related with this issue viewed from the public health, the financial, and governance. The major difficulties and pitfalls related within these aspects are properly addressed, and the final conclusions are clear and well presented.\nAlso, it has to be recognized the value of the inclusion of new elements in this kind of review, such as the short review of the WHO global TB program, the value proposition for TB vaccines, the mathematical modelling techniques and its impact in the evaluation of new vaccines, bio-signatures, biomarkers and “omic” sciences, and the perspectives of the funders for the development of vaccines.\nI am certain that this review will be of great interest for those researchers related to tuberculosis vaccines and all people interested in this thematic.\n\nIn conclusion, this review could be suitable for indexing in the actual condition, no modifications or corrections are suggested.\nGreat job, 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": []
},
{
"id": "41389",
"date": "14 Jan 2019",
"name": "Salil Bhargava",
"expertise": [
"Reviewer Expertise Tuberculosis and Chest 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 review article has covered the problem statement, the need for the development of new vaccines and the newer vaccines in phase 2 and 3 very well. This article covers the newer platform i.e. mRNA and new viral vector hCMV for vaccine delivery. The authors have also elaborated on immune mechanisms stimulated in response to Mtb including HLA E, CD1a, antibody response in addition to MHC I and II responses. These mechanisms are targeted by newer vaccine candidates to generate an immune response in adults and adolescents.\nThe review also covers the use of advanced assessment techniques, including transcriptomics, proteomics and metabolomics for identifying newer biomarkers for identifying active disease and can further be used to assess vaccine efficacy and safety. The article also highlights alternative routes of immunization and discusses BCG revaccination status.\n\nHowever, it would be helpful if the following points are discussed in the review:\nRole of DNA vaccines (Tian et al., 20181 and Bruffaerts et al., 20142). 231 subcultures in place of 230 subcultures over 13 years to develop BCG on page 7 (Norouzi et al., 20123).\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/7-1732
|
https://f1000research.com/articles/8-213/v1
|
24 Feb 19
|
{
"type": "Method Article",
"title": "Relative Abundance of Transcripts (RATs): Identifying differential isoform abundance from RNA-seq",
"authors": [
"Kimon Froussios",
"Kira Mourão",
"Gordon Simpson",
"Geoffrey J. Barton",
"Nicholas Schurch",
"Kimon Froussios",
"Kira Mourão",
"Gordon Simpson",
"Geoffrey J. Barton"
],
"abstract": "The biological importance of changes in RNA expression is reflected by the wide variety of tools available to characterise these changes from RNA-seq data. Several tools exist for detecting differential transcript isoform usage (DTU) from aligned or assembled RNA-seq data, but few exist for DTU detection from alignment-free RNA-seq quantifications. We present the RATs, an R package that identifies DTU transcriptome-wide directly from transcript abundance estimates. RATs is unique in applying bootstrapping to estimate the reliability of detected DTU events and shows good performance at all replication levels (median false positive fraction < 0.05). We compare RATs to two existing DTU tools, DRIM-Seq & SUPPA2, using two publicly available simulated RNA-seq datasets and a published human RNA-seq dataset, in which 248 genes have been previously identified as displaying significant DTU. RATs with default threshold values on the simulated Human data has a sensitivity of 0.55, a Matthews correlation coefficient of 0.71 and a false discovery rate (FDR) of 0.04, outperforming both other tools. Applying the same thresholds for SUPPA2 results in a higher sensitivity (0.61) but poorer FDR performance (0.33). RATs and DRIM-seq use different methods for measuring DTU effect-sizes complicating the comparison of results between these tools, however, for a likelihood-ratio threshold of 30, DRIM-Seq has similar FDR performance to RATs (0.06), but worse sensitivity (0.47). These differences persist for the simulated drosophila dataset. On the published human RNA-seq dataset the greatest agreement between the tools tested is 53%, observed between RATs and SUPPA2. The bootstrapping quality filter in RATs is responsible for removing the majority of DTU events called by SUPPA2 that are not reported by RATs. All methods, including the previously published qRT-PCR of three of the 248 detected DTU events, were found to be sensitive to annotation differences between Ensembl v60 and v87.",
"keywords": [
"Transcriptomics",
"Differential Isoform Usage",
"Transcriptional regulation",
"Gene regulation",
"Feature selection",
"Algorithms",
"Visualization"
],
"content": "Introduction\n\nHigh-throughput gene regulation studies have focused primarily on quantifying gene expression and calculating differential gene expression (DGE) between samples in different groups, conditions, treatments, or time-points. However, in higher eukaryotes, alternative splicing of multi-exon genes and/or alternative transcript start and end sites leads to multiple transcript isoforms originating from each gene. Since transcripts represent the executive form of genetic information, analysis of differential transcript expression (DTE) is preferable to DGE. Unfortunately, isoform-level transcriptome analysis is more complex and expensive since, in order to achieve similar statistical power in a DTE study, higher sequencing depth is required to compensate for the expression of each gene being split among its component isoforms. In addition, isoforms of a gene share high sequence similarity and this complicates the attribution of reads among them. Despite these challenges, several studies have shown that isoforms have distinct functions1–3 and that shifts in individual isoform expression represent a real level of gene regulation4–7, suggesting there is little justification for choosing DGE over DTE in the study of complex transcriptomes.\n\nIt is possible to find significant DTE among the isoforms of a gene, even when the gene shows no significant DGE. This introduces the concept of differential transcript usage (DTU), where the abundances of individual isoforms of a gene can change relative to one another, with the most pronounced examples resulting in a change of the dominant isoform (isoform switching). The definitions of DGE, DTE and DTU are illustrated in Figure 1.\n\nThe expression of two genes (Gene A and Gene B), with 3 and 2 isoforms respectively, is compared across two conditions (Condition 1 and Condition 2). The horizontal width of each coloured box represents the abundance of the relevant gene or transcript. A negative differential expression result (red cross-mark) for a given entity in any one of the three analysis types does not exclude that same entity from having a positive result (green tick-mark) in one of the other two analysis types. The relative isoform abundances in [iii] are scaled to the absolute isoform abundances in [ii], which in turn are scaled to the gene expressions in [i]. Gene A is differentially expressed, but only two of its three isoforms are differentially expressed (A.2 and A.3). Proportionally, Gene A’s primary isoform (A.3) remains the same, but the ratios of the two less abundant isoforms change. Gene B is not differentially expressed, but both its isoforms are differentially expressed, and demonstrate an example of isoform switching. DGE: Differential gene expression, DTE: Differential transcript expression, DTU: Differential transcript usage.\n\nTo quantify the isoforms and assess changes in their abundance, most existing tools for DTE and DTU analysis (e.g. Cufflinks8, DEXSeq9, LeafCutter10) rely on reads that either span splice-junctions or align to unique exons. However, with the newest generation of transcript quantification tools (Kallisto11,12, Sailfish13, Salmon14), reads are aligned to neither the transcriptome nor the genome. Instead, these tools combine a pseudo-mapping of the k-mers present within each read to the k-mer distributions from the transcriptome annotation with an expectation maximization algorithm, to infer the expression of each transcript model directly. Such alignment-free methods are much faster than the traditional alignment-based methods (RSEM15, TopHat216, STAR17) or assembly-based methods (Cufflinks8, Trinity18), making it feasible to repeat the process many times on iterative subsets of the read data and, thus, quantify the technical variance in the transcript abundance estimates. However, the lack of alignments prevents these new methods from being compatible with differential expression methods such as Cufflinks, DEXSeq and Leafcutter. Instead, Sleuth19 is a tool that handles DTE analysis from alignment-free transcript quantifications. DTU analysis is currently less straight-forward. SwitchSeq20 focuses on a particular subset of DTU analysis from alignment-free data, namely isoform switching, whereas iso-kTSP6 identifies both DTU and isoform switching, but focuses on the highest-ranking pair of change-exhibiting isoforms per gene. SUPPA21,22, on the other hand, primarily identifies differential splicing events at the junction level, with recent developmental versions having added isoform-level capability. Finally, DRIM-Seq23 identifies DTU directly from quantification data, but defines the effect size as a fold change which may not be the most appropriate way to compare proportions.\n\nIn this paper, we present RATs (Relative Abundance of Transcripts), an R package for identifying DTU directly from isoform quantifications. It is designed to use alignment-free abundance data and is the only tool that exploits bootstrapping to assess the robustness of the DTU calls. RATs provides raw, summary and graphical results, allowing for ease of use as well as for advanced custom queries, and the R language is the environment of choice for many widely-used DGE and DTE tools, allowing for easy integration of RATs in existing workflows. We assess the accuracy of RATs in comparison to SUPPA2 and DRIM-Seq and find RATs to perform at as well as or better than its competitors. Finally, we demonstrate that the results of both RNA-seq based and qRT-PCR based analyses are sensitive to the annotation used for transcript quantification and primer design, respectively.\n\n\nMethods\n\nRATs identifies DTU independently at both the gene and transcript levels using an efficient implementation of the G-test of independence24, without continuity corrections. The criteria RATs uses to identify DTU are described in detail below.\n\nPrior to statistical testing by either method, RATs first filters the input isoform abundance data to reduce both the number of low quality calls and the number of tests carried out. Specifically: (i) isoform ratio changes can only be defined for genes that are expressed in both conditions, with at least two isoforms detected, and (ii) transcript abundances must exceed an optional minimum abundance threshold. Transcripts with abundances below the threshold are considered as not detected.\n\nSignificant changes in relative transcript abundance are detected using two separate approaches: one at the gene level and the other at the transcript level. At the gene level, RATs compares the set of each gene’s isoform abundances between the two conditions to identify if the abundance ratios have changed. At the transcript level, RATs compares the abundance of each individual transcript against the pooled abundance of its sibling isoforms to identify changes in the proportion of the gene’s expression attributable to that specific transcript. Both methods include the Benjamini-Hochberg false discovery rate correction for multiple testing25. These tests are performed on the summed abundance of each isoform across the replicates.\n\nTranscripts whose absolute difference in isoform proportion is below a set threshold are rejected, even if the difference is statistically significant.\n\nRATs provides the option to use the bootstrapped abundance estimates obtainable from alignment-free quantification tools to apply a reproducibility constraint on the DTU calls, by randomly selecting individual quantification iterations from each replicate and measuring the fraction of these iterations that result in a positive DTU classification. Typically, each sample is represented by the mean abundance of each transcript, calculated across the quantification iterations. However, this loses the variance information of the quantification. By referring back to the quantification iterations, RATs highlights cases where the quantification was unreliable due to high variability and therefore the DTU result should also be considered unreliable. Similarly, RATs optionally also measures the reproducibility of the DTU results relative to the inter-replicate variation by iteratively sub-setting the samples pool.\n\nRATs is implemented in R26 and has been freely distributed through Github as an R source package since August 2016. RATs accepts as input either a set of R tables with abundances (with or without bootstrap information), or a set of Salmon14 or Kallisto11 output files. An annotation table mapping the correspondence between transcript and gene identifiers is also required, either provided directly or inferred from a GTF file. Results are returned in the form of R data.table objects27. Along with the DTU calls per transcript and gene, the tables record the full provenance of the results. Convenience functions are provided for summary tallies of DTU and isoform-switching results, for ID retrieval, and for visualization of the results via ggplot2 (v2.2.1)28. Details on these are available through the user manual of the package. Once created, all plots produced by RATs remain customisable via standard ggplot2 operations.\n\nThe performance was assessed in two ways. Firstly, the false positives (FP) performance of RATs (v0.6.2) for detection of DTU between two groups relative to the level of experimental replication was measured on groups generated by random selection without replacement from a pool of 16 high-quality wild-type Colombia-0 Arabidopsis thaliana replicates29 1. This was iterated 100 times for each replication level in the range 3 ≤ n ≤ 8. As the two groups are drawn from the same condition, any positive DTU calls must be considered to be false positives. For each iteration, we recorded the fraction of genes and transcripts that were reported as DTU, relative to the total number of genes or transcripts tested in that iteration. The commands and scripts used are from the RATs Github repository.\n\nSecondly, two simulated datasets30 were used to benchmark the sensitivity (s, the fraction of the 1000 DTU events actually detected), false discovery rate (FDR, the fraction of reported DTU events that is not part of the 1000 “real” events) and Matthews correlation coefficient (MCC) of RATs, SUPPA2 and DRIM-Seq. The datasets were made of simulated RNA-seq reads based on the transcriptome annotation and to match realistic RNA-seq transcript expression values. To create the second condition, the abundance values of the two most abundant transcript isoforms originating from a gene locus were swapped for 1000 well-expressed coding gene loci. The transcriptome annotation used for both Human and fly comprised only annotated protein coding genes (13937 in the Drosophila, 20410 in the human) leaving a number of other classifications of gene unaccounted for (1745 in the Drosophila, 41483 in the human). These genes constitute a convenient negative set for simulation and should have no expression, save for any reads misallocated to them by the quantification tools. The simulated datasets were obtained from ArrayExpress2 and quantified with both Kallisto (v0.44;11 and Salmon (v0.9.1;14 using the respective complete annotations that match the simulation of the datasets (Ensembl v70 for the Drosophila and Ensembl v71 for the human;30). The sensitivity, FDR and MCC were measured for a range of comparable parameters between RATs (v0.6.4), SUPPA2 (v2.3) and DRIM-Seq (v1.6, Bioconductor v3.6, R v3.4). No transcript abundance pre-filter was imposed on any of the three DTU tools, and the significance level was set to 0.05 for all runs. For RATs and SUPPA2, three thresholds for the effect size (difference in proportion) were tested; the RATs’ current default of 0.2, and more permissive values 0.1 & 0.05. For DRIM-Seq, threshold values of the likelihood ratio were explored from 0-30. Finally, RATs reproducibility thresholds were explored in the range of 0.8-0.95 for the quantification reproducibility and 0.55-0.85 for the inter-replicate reproducibility. The tool performance was measured using annotations comprised of all annotated genes and only protein coding genes.\n\nTo test the ability of RATs to identify known instances of DTU, we compared it against validated instances of DTU from publicly available RNA-seq data. We took read data from Deng et al. (2013, 31), who identified non-DGE changes in the isoform levels of genes between three human patients with idiopathic pulmonary fibrosis (IPF) and three lung cancer patients used as controls. The dataset contains 25 million 54-base long single-end Illumina reads per lung tissue sample. As in the original at study, we used Ensembl v6032 as the source of the reference human genome and its annotation, in which each of the three discussed genes features two isoforms. Unlike the original study, we used Salmon (v0.7.1, with sequence bias correction enabled, 100 bootstrap iterations and default values for the remaining parameters, using k=21 for the index) to quantify the isoform abundances. DTU was identified by RATs v0.6.2. For comparison, we repeated the quantification and DTU analysis of the data with the same tool versions and parameters, but using the annotation and assembly from Ensembl v87, the current version at the time of this study.\n\nWe also submitted the quantification data to SUPPA2, in its psiPerIsoform mode, and to DRIM-Seq. For a fair comparison, we tried to minimize variability in the parameters and data type used by the three tools. As SUPPA2 offered no abundance pre-filtering, RATs and DRIM-Seq were run with abundance threshold values of 0. The p-value cut-off was set at 0.05 for all three tools, using the corrected p-values where available. For the difference in isoform proportion (SUPPA2 and RATs) the threshold was set at 0.20. No threshold was set for the fold-changes in DRIM-Seq. SUPPA2 required and was provided with TPM abundances. For consistency in the use of abundances normalised for transcript length, RATs and DRIM-Seq were also provided with TPM, but the values were scaled up to the average library size of 25M reads, as their testing methods expect counts and would be under-powered if used directly with TPMs. Again, the commands and scripts used are available from the RATs Github repository.\n\n\nResults\n\nBoth the gene-level and transcript-level approaches to identifying DTU implemented in RATs achieved a median FP fraction <0.05 on our A. thaliana dataset, even with only three replicates per condition (Figure 2A). Higher replication results in both a reduction in the number of false positives and restricts the false positives to smaller effect sizes (Figure 2B). The gene-level and transcript-level approaches, however, have different strengths and weaknesses. Simultaneously utilizing the expression information across all the isoforms in a gene makes the gene-level test sensitive to smaller changes in relative expression, compared to testing transcripts individually, but it also makes the gene-level test more prone to false positives. Figure 2 shows that the gene-level test has a higher FP fraction than the transcript-level test, irrespective of replication level or effect size, although the two methods converge for highly replicated experiments or large effect sizes. Furthermore, the gene-level test only identifies the presence of a shift in the ratios of the isoforms belonging to the gene, without identifying which specific isoforms are affected. The transcript-level test, in contrast, directly identifies the specific isoforms whose proportions are changing and has fewer false positives than the gene-level test. However, considering each isoform independently requires a larger number of tests to be performed, thus resulting in a greater multiple testing penalty.\n\nFalse positive fraction measured over 100 permutation iterations of randomly selected (without replacement) replicates from a pool of 16 high-quality wild-type Colombia-0 Arabidopsis thaliana replicates from Froussios et al. (2017,29). [A] FP fraction of each bootstrap iteration, for default values of all RATs parameters (v0.6.2), across a range of replication levels, separately for the gene-level test (red) and transcript level test (blue). [B] Mean FP fraction by replication level, as a function of the effect size threshold (effect size = difference between conditions of an isoform’s proportion). For a gene, the effect size is defined as the largest proportion difference observed among that gene’s isoforms. In every iteration, the FP fraction was calculated against the number of genes or transcripts that were eligible for testing each time (a number which remains very stable across iterations and replication levels – see Extended data 133).\n\nThe sensitivity, FDR and MCC performance of RATs, SUPPA2 and DRIM-Seq using Salmon transcript quantifications of annotated protein coding gene isoforms are summarised in Figure 3. Tested with the simulated Human dataset, the parameter defaults for RATs (quantification reproducibility >95%, inter-replicate reproducibility >85% & effect-size >0.2) result in a sensitivity of s = 0.55, MCC = 0.71 and FDR = 0.04, outperforming both other tools. With the same thresholds, SUPPA2 has a higher sensitivity (s = 0.61) but poorer FDR performance (FDR = 0.33). Direct comparison with DRIM-Seq is complicated by different methods for measuring DTU effect-sizes between the tools, however for a likelihood-ratio threshold of 30, DRIM-Seq has similar FDR performance to RATs (FDR = 0.06), but worse sensitivity (s = 0.47). These differences persist for the simulated drosophila dataset. DRIM-Seq consistently shows the lowest sensitivity (≤0.65), while maintaining a FDR ≤0.2 in any of the tried parameter sets. SUPPA2 is the most sensitive of the three tools (0.6 ≤ s ≤ 0.9), but also has the highest FDR (0.35 ≤ FDR ≤ 0.65 in human, 0.10 ≤ FDR ≤ 0.25 in Drosophila). RATs can match the sensitivity of SUPPA2 while maintaining a lower FDR than SUPPA2 by relaxing its quantification reproducibility (Qrep) and inter-replicate reproducibility (Rrep) thresholds. At the highest effect-size thresholds (DpropRATs = 0.2 and lrDRIM-Seq = 0.3) DRIM-Seq has a comparable FDR to that of RATs. Surprisingly, the sensitivity, MCC and FDR of DRIM-Seq is not strongly sensitive to variations in the likelihood ratio effect-size threshold. Consequentially, RATs has worse FDR performance, but better sensitivity than DRIM-Seq at lower effect-size thresholds. Across all the simulated dataset and parameter combinations the gene-level test implemented in RATs shows higher sensitivity and higher FDR compared with the results from the transcript-level test. Extending the test to isoforms from the full set of annotated genes, rather than only those from protein coding genes, adds a considerable number of additional true negatives (Drosophila: 1745, human: 4148, see Section: Performance) resulting in a small increase of FDR and slight reduction of MCC for all tools in both datasets (Extended data 233). Similarly, using Kallisto isoform expression quantifications in place of the quantifications from Salmon does not strongly affect the results (Extended data 233). The performance results of RATs on these simulated datasets are in good agreement with those presented in Love et al. (2018,34), which also demonstrates that the performance of RATs is similar to, or exceeds, the performace of other DTU tools , including DRIM-seq, SUPPA2 or DEX-Seq.\n\nThe performance was assessed on the human [A] and Drosophila [B] simulated datasets from ArrayExpress E-MTAB-376630, over a range of threshold values for the effect size (RATs - Dprop, SUPPA2 - dPSI, DRIM-Seq likelihood ratio - lr) and confidence in the result (RATs quantification reproducibility – Qrep, RATs inter-replicate reproducibility - Rrep). The statistical significance cut-off was at 0.05 for all cases. The measures of performance are the sensitivity, false discovery rate (FDR) and Matthews correlation coefficient (MCC). The datasets were quantified using Salmon 0.9.2 and the metrics were calculated accounting only for the genes strictly listed in the “truth” sets. The results using Kallisto for the quantification are practically identical (see Extended data 233).\n\nAfter pre-filtering, Deng et al. (2013, 31) tested 3098 Ensembl v60 genes for DTU by quantifying their isoform proportions with RAEM35 and using Pearsons Chi-squared test of independence with a FDR threshold of 5%. They identified 248 genes that were not differentially expressed but displayed significant DTU. Subsequently, they confirmed three of them with qRT-PCR: TOM1L1 (ENSG00000141198), CMTM4 (ENSG00000183723), and PEX11B (ENSG00000131779). Table 1 shows the fraction of the 248 DTU genes identified in this study that were also called by RATs, SUPPA2 and DRIM-Seq, as well as each tool’s verdict on each of the three validated genes. The genes reported as DTU by RATs are listed in Extended data 3 & 433 respectively, based on the Ensembl v60 and v87 human annotations.\n\nThe first column shows the fraction of the 248 genes that was recaptured by each method. For methods reporting at the transcript level, results were aggregated to the respective genes. The last three columns show whether the verdicts for each of the validated genes (DTU Yes/No). DTU: Differential transcript usage.\n\nNone of the three tools recapitulated the reported 248 genes well, with the highest fraction of 26% achieved by DRIM-Seq possibly due to a tendency to over-predict (see next section). Of the three validated genes, only CMTM4 is reported by all methods, and only SUPPA2 reports all three genes. Although the rejection of TOM1L1 and PEX11B by DRIM-Seq was due to poor statistical significance, RATs reported that the changes found were both statistically significant and of sufficient effect size. Instead, RATs rejected the genes on the grounds of poor reproducibility (see Section: DTU Calling).\n\nThere have been extensive changes in the human transcriptome annotation since Ensembl v60. We hypothesized that these changes could have a significant impact on the set of genes identified in Deng et al. (2013, 31). Table 2 shows that in addition to the new genome assembly, the human transcriptome complexity has increased significantly from Ensembl v60 to the more recent v87. Changing the version of the human annotation from Ensembl v60 to v87 removes 10,253 gene IDs and adds 15,839 new ones. Re-quantifying the RNA-seq data with the updated annotation and re-calling DTU resulted in similarly poor overlap between the tools’ results and the original report (see Extended data 533). Of the three validated genes, TOM1L1 was unanimously rejected by all methods, CMTM4 remained unanimously reported as DTU, and PEX11B was reported as DTU by RATs and SUPPA2, but not by DRIM-Seq.\n\nIn total, the later annotation contains 25% more transcript models. The three genes identified by Deng et al. (2013, 31), TOM1L1, CMTM4 and PEX11B, have all acquired additional isoform models.\n\nThe isoform abundances in Figure 4 reveal that all three genes showed plausible shifts in relative isoform abundance with the Ensembl v60 quantifications, but only PEX11B showed the same shift with Ensembl v87. Instead, TOM1L1 showed no significant changes in any of its 23 isoforms and the primary isoform in the Control samples changed from isoform 2 (ENST00000445275) to isoform 1 (ENST00000348161). CMTM4 shows a similar abundance shift with v87 as it did with v60, but the isoforms implicated changed from isoforms 1 (ENST00000330687) and 2 (ENST00000394106) to isoforms 1 and 5 (ENST00000581487). These changes of context raised questions about the qRT-PCR validation performed in the original analysis of the data31. Indeed, when the reported qRT-PCR primers were aligned to the Ensembl v87 sequence and annotation (see Extended data 633), only the primers for PEX11B yielded the same conclusion as with Ensembl v60. For TOM1L1, the primers intended for ENST00000445275 no longer matched that isoform, but matched two other isoforms instead (ENST00000570371 and ENST00000575882). Additionally, the primers intended to quantify the gene as a whole failed to match half of the gene’s new isoforms, and the two sets of captured isoforms did not overlap completely and were thus incomparable in any meaningful way. As a consequence, the qRT-PCR intensities measured in the original study are actually impossible to interpret in the context of the updated annotation and the originally reported conclusion is likely wrong. For CMTM4 the primers reported matched multiple but not all isoforms, casting doubt on the interpretation of the qRT-PCR measurements for this gene as well. Only for PEX11B did the primers target the isoforms in a way that would give interpretable results and indeed lead to the same conclusion as originally reported31.\n\nIsoform IDs on the x axis were replaced with simple numbers to minimize clutter, but the mapping of number to ID is maintained between the two annotations. The y axis represents the relative abundance of each isoform. In red are the quantifications from the three replicates of the Control condition, and in blue are those from the IPF condition. The full version of the plots by RATs, including the full isoform IDs, is available in Extended data 533.\n\nTable 3 summarises the results obtained by RATs, SUPPA2 and DRIM-Seq for the Deng et al. (2013, 31) dataset using Ensembl v60 (same as the original study) and Ensembl v87 (current version at time of the present work). With either annotation, DRIM-Seq reported the most DTU genes – almost 1000 with v60 and almost 1700 with v87. The RATs gene-level method reported fewer genes by a factor of 1.5 and 2 respectively compared to DRIM-Seq with each annotation. SUPPA2 reported several hundred transcripts more than RATs, but at the gene level the numbers were comparable. RATs and DRIM-Seq reported more genes and transcripts with v87 of the annotation than with v60, whereas SUPPA2 reported slightly fewer with v87. Despite overall similar volume of results between the two versions of the annotation, it is evident from Table 3 that the overlap of the results between annotations is poor for all methods. For RATs and SUPPA2, only 30–40% of the genes reported with Ensembl v60 were also reported with v87. For DRIM-Seq this overlap was 55% of its Ensembl v60 results.\n\nDRIM-Seq reports DTU only at the gene level. SUPPA2 reports DTU only at the individual transcript level. RATs reports at both the transcript and the gene levels, using its respective test implementations. For SUPPA2 and the transcript-level approach in RATs, gene-level results can be inferred from the reported transcripts; these are included in the table, enclosed in parentheses. The last two columns show the reproducibility of the results between annotation versions. DTU: Differential transcript usage.\n\nThe overlap of results between different methods is similar to the overlap of results between annotations, as shown in Table 4. 97% of the genes reported by gene-level method in RATs are also identified as DTU by the transcript-level method. Among all the pairwise comparisons of RATs, SUPPA2 and DRIM-Seq, however, the highest level of agreement at both transcript and gene level is between SUPPA2 and RATs. SUPPA2 identifies DTU in 53% of the transcripts that are called as DTU by the transcript-level method in RATs, however RATs calls DTU for only 35% of the transcripts identified as DTU by SUPPA2. DRIM-Seq consistently reports a higher number of DTU identifications than either RATs or SUPPA2, but still only manages at most 43% agreement with the other two tools.\n\nThe overlaps are shown as the proportion of the results from the methods on the columns captured by the methods on the rows.\n\nRATs and SUPPA2 are more similar than implied by the level of agreement presented in Table 4. Figure 5 shows that the novel reproducibility testing feature in RATs, which discounts DTU identification from highly variable quantifications (see Section: DTU Calling), is responsible for rejecting 43% of the SUPPA2 DTU transcripts and 28% of the DRIM-Seq DTU genes that pass the significance and effect size filtering criteria. 53% of the DRIM-Seq results and, perplexingly, 18% of the SUPPA2 results are rejected due to the effect size filter (after passing the significance testing, but prior to the reproducibility filter), despite all the tools operating on the same input isoform quantifications.\n\nThe colours represent the different criteria imposed by RATs. Since no abundance pre-filtering was enabled for any of the tools, there are no rejections caused by the transcript abundance and the effective number of expressed isoforms. DTU: Differential transcript usage.\n\nRATs’ runtime and memory consumption depend on the size of the annotation and the number of bootstraps iterations. Where multiple processing cores are available, RATs can be instructed to take advantage of them. The runtime and maximum memory usage for the two simulated datasets from our benchmarks, running on a high-specification laptop, are shown in Table 5.\n\nMeasured via the peakRAM package36. For the bootstrapped runs, 100 iterations were used for the quantification reproducibility and 9 for the cross-replicate reproducibility, representing all the pairwise combinations of the 3 replicates per condition.\n\n\nDiscussion\n\nReliable identification of differential isoform usage depends critically on i) the accuracy of the upstream isoform expression quantifications, and ii) on the accuracy of the annotation they use. RATs is the first differential isoform usage tool to include the reproducibility of the upstream isoform expression quantifications to refine its DTU identifications, directly addressing the accuracy of the upstream isoform expression quantifications. Leveraging the bootstrapped isoform expression quantifications from fast modern alignment-free isoform expression quantification tools (such as Kallisto and Salmon) allows RATs to reject those cases of DTU that are based on highly uncertain isoform quantifications. Existing tools rely on the mean isoform abundances, which can hide a large degree of variability, and are thus insensitive to this reproducibility criterion. We recommend running RATs, and the underlying alignment-free isoform expression quantification tools that generate the data it operates on, with at least 100 bootstrap iterations.\n\nWe evaluated RATs on both simulated data and on a high-quality experimental dataset from Deng et al. (2013, 31) and show that it outperforms both DRIM-Seq and SUPPA2. On the simulated data with stringent effect-size, reproducibility and statistical significance threshold, both the gene-level and transcript-level methods in RATs have a lower FDR than the other two tools, for a comparable sensitivity and comparable or superior Matthews correlation coefficient. This makes RATs particularly useful for data from organisms with large transcriptomes where the risk of false positives is higher. Relaxing these stringent thresholds increases the FDR for all the tools and for the lowest tested effect-size thresholds all the tools struggle to control their FDR adequately leaving little room for optimism regarding the identification of DTU with small effect sizes, particularly in low expression genes. The choice of alignment-free transcript quantification tool did not strongly affect the performance of the DTU tools within the examined parameter space, although in the simulated datasets Kallisto appears more prone to overestimating the expression of non-protein-coding genes that in the design of the simulation are not expressed (see Extended data 233). Comparing the DTU classifications of the three tools against the instances of DTU identified in the Deng et al. (2013, 31) dataset, we found pairwise overlaps between the tools of at most 53%. The low level of agreement between the three tools reflects their different methodological choices, such as the very different definitions of effect size. Both SUPPA2 and RATs use the difference in relative isoform abundance as their measure of the DTU effect size, however RATs tests this difference directly whereas SUPPA2 extrapolates it from the differential inclusion of splice sites. This comparison also highlights the dependence of DTU identification methods on the accuracy of the underlying transcriptome annotation, (a limitation common to all biological tools that use an annotation as guide37). Running RATs, SUPPA2, and DRIM-Seq on the Deng et al. (2013, 31) datasets with two different versions of the ensembl H. sapiens transcriptome annotation separated by six years produces dramatic differences in the DTU identification results. All three validated DTU genes from the original Deng et al. study contained additional isoforms in the newer annotation and only one of these genes displayed the same isoform abundance shifts using both annotations. With the newer annotation, the DTU of one validated gene was attributable to different isoforms depending on the annotation version, while another showed no significant DTU with the newer annotation. qRT-PCR has long considered the de facto standard for orthogonal confirmation of high-throughput transcriptomic results however it too is subject to the same limitation, illustrated by multiple matches of the specific primer sequences used for validation in the Deng et al. (2013, 31) study in the newer annotation. Annotation of the transcriptomes remains a work in progress even for model organisms and the extensive sequence overlap between isoforms together with the ongoing discovery of additional isoforms suggests that qRT-PCR may not be a suitable method for the validation of transcript abundance changes. For hybridization-based methods like qRT-PCR to serve as a reliable validation method for RNA quantification, the suitability of the primers should first be validated by sequencing the captured amplicons. Soneson et al. (2016,30) show that pre-filtering annotations can improve quantification performance and this approach may also be helpful in qRT-PCR primer design.\n\nIn the future, experiment-specific transcriptome annotations could be obtained by including a parallel set of full-length isoform RNA-seq data in the experimental design, such as via PacBio sequencing or Oxford Nanopore Direct RNA-seq. An advantage of this approach is that it would better define the transcriptome for the specific experiment38–41. This may be of importance for experiments focusing on specific tissues or developmental stages of an organism, where the active transcriptome is likely to be only a subset of the global reference transcriptome of the organism.\n\n\nData availability\n\nThe Arabidopsis thaliana RNA-sequencing data used in this study is available from ArrayExpress under the study E-MTAB-5446. The simulated Homo sapiens and Drosophila Melanogaster datasets are available from ArrayExpress under the study E-MTAB-3766. The Deng et al. (2013, 31) data are available from the European Nucleotide Archive, or the Short Read Archive, under the study SRA048904.\n\nExtended data are available along with the source code from GitHub and archived with Zenodo\n\nZenodo: Extended data. bartongroup/RATS: RATs 0.6.5 - R source package, http://doi.org/10.5281/zenodo.255656433\n\nLicence: MIT\n\nThe RATs R package is open source and available through Github\n\nSource code: https://github.com/bartongroup/RATs.\n\nArchived source code: http://doi.org/10.5281/zenodo.255656433\n\nLicence: MIT",
"appendix": "Grant information\n\nThis work has been supported by the Biotechnology and Biological Sciences Research Council grants [BB/H002286/1; BB/J00247X/1; BB/M010066/1; BB/M004155/1] and the Wellcome Trust Strategic Awards [098439/Z/12/Z and WT097945].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or prepa-ration of the manuscript.\n\n\nFootnotes\n\n1 https://www.ebi.ac.uk/arrayexpress/experiments/E-MTAB-5446/\n\n2 https://www.ebi.ac.uk/arrayexpress/experiments/E-MTAB-3766/\n\n\nReferences\n\nCarvalho RF, Feijão CV, Duque P: On the physiological significance of alternative splicing events in higher plants. Protoplasma. 2013; 250(3): 639–50. PubMed Abstract | Publisher Full Text\n\nLee Y, Rio DC: Mechanisms and Regulation of Alternative Pre-mRNA Splicing. Annu Rev Biochem. 2015; 84(1): 291–323. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBaralle FE, Giudice J: Alternative splicing as a regulator of development and tissue identity. Nat Rev Mol Cell Biol. 2017; 18(7): 437–451. PubMed Abstract | Publisher Full Text\n\nAanes H, Østrup O, Andersen IS, et al.: Differential transcript isoform usage pre- and post-zygotic genome activation in zebrafish. BMC Genomics. 2013; 14: 331. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHong SE, Song HK, Kim DH: Identification of tissue-enriched novel transcripts and novel exons in mice. BMC Genomics. 2014; 15(1): 592. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSebestyén E, Zawisza M, Eyras E: Detection of recurrent alternative splicing switches in tumor samples reveals novel signatures of cancer. Nucleic Acids Res. 2015; 43(3): 1345–56. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhao W, Hoadley KA, Parker JS, et al.: Identification of mRNA isoform switching in breast cancer. BMC Genomics. 2016; 17: 181. 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\nAnders S, Reyes A, Huber W: Detecting differential usage of exons from RNA-seq data. Genome Res. 2012; 22(10): 2008–17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi YI, Knowles DA, Humphrey J, et al.: Annotation-free quantification of RNA splicing using LeafCutter. Nat Genet. 2018; 50(1): 151–158. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBray NL, Pimentel H, Melsted P, et al.: Near-optimal probabilistic RNA-seq quantification. Nat Biotechnol. 2016; 34(5): 525–7. PubMed Abstract | Publisher Full Text\n\nBray NL, Pimentel H, Melsted P, et al.: Erratum: Near-optimal probabilistic RNA-seq quantification. Nat Biotechnol. 2016; 34(8): 888. PubMed Abstract | Publisher Full Text\n\nPatro R, Mount SM, Kingsford C: Sailfish enables alignment-free isoform quantification from RNA-seq reads using lightweight algorithms. Nat Biotechnol. 2014; 32(5): 462–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatro R, Duggal G, Love MI, et al.: Salmon provides fast and bias-aware quantification of transcript expression. Nat Methods. 2017; 14(4): 417–419. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi B, Dewey CN: RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics. 2011; 12(1): 323. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim D, Pertea G, Trapnell C, et al.: TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 2013; 14(4): R36. PubMed Abstract | Publisher Full Text | Free Full Text\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\nGrabherr MG, Haas BJ, Yassour M, et al.: Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nat Biotechnol. 2011; 29(7): 644–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPimentel H, Bray NL, Puente S, et al.: Differential analysis of RNA-seq incorporating quantification uncertainty. Nat Methods. 2017; 14(7): 687–690. PubMed Abstract | Publisher Full Text\n\nGonzàlez-Porta M, Brazma A: Identification, annotation and visualisation of extreme changes in splicing from RNA-seq experiments with SwitchSeq. bioRxiv. 2014; 005967. Publisher Full Text\n\nAlamancos GP, Pagès A, Trincado JL, et al.: Leveraging transcript quantification for fast computation of alternative splicing profiles. RNA. 2015; 21(9): 1521–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTrincado JL, Entizne JC, Hysenaj G, et al.: SUPPA2: fast, accurate, and uncertainty-aware differential splicing analysis across multiple conditions. Genome Biol. 2018; 19(1): 40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNowicka M, Robinson M: DRIMSeq: a Dirichlet-multinomial framework for multivariate count outcomes in genomics [version 2; referees: 2 approved]. F1000Res. 2016; 5: 1356. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSokal RR, Rohlf FJ: Biometry: The Principles and Practice of Statistics in Biological Research. W. H. Freeman. 1995. Reference Source\n\nBenjamini Y, Hochberg Y: Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. J Roy Stat Soc B Met. 1995; 57(1): 289–300. Reference Source\n\nR Development Core Team: R: A Language and Environment for Statistical Computing. Vienna, Austria: The R Foundation for Statistical Computing. 2011. Reference Source\n\nDowle MSA, Short T, Lianoglou S, et al.: Data.table: Extension of Data.frame. 2015.\n\nWickham H: ggplot2: Elegant Graphics for Data Analysis. Springer International Publishing. 2016. Publisher Full Text\n\nFroussios K, Schurch NJ, Mackinnon K, et al.: How well do RNA-Seq differential gene expression tools perform in a eukaryote with a complex transcriptome? bioRxiv. 2017. Publisher Full Text\n\nSoneson C, Matthes KL, Nowicka M, et al.: Isoform prefiltering improves performance of count-based methods for analysis of differential transcript usage. Genome Biol. 2016; 17: 12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeng N, Sanchez CG, Lasky JA, et al.: Detecting splicing variants in idiopathic pulmonary fibrosis from non-differentially expressed genes. PLoS One. 2013; 8(7): e68352. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZerbino DR, Achuthan P, Akanni W, et al.: Ensembl 2018. Nucleic Acids Res. 2018; 46(D1): D754–D761. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFroussios K, Schurch N: bartongroup/RATS: RATs 0.6.5 - R source package (Version 0.6.5). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.2556564\n\nLove MI, Soneson C, Patro R: Swimming downstream: statistical analysis of differential transcript usage following Salmon quantification [version 3; referees: 3 approved]. F1000Res. 2018; 7: 952. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeng N, Puetter A, Zhang K, et al.: Isoform-level microRNA-155 target prediction using RNA-seq. Nucleic Acids Res. 2011; 39(9): e61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQuinn T: peakRAM: Monitor the total and peak RAM used by an expression or function. 2017. Reference Source\n\nWu PY, Phan JH, Wang MD: Assessing the impact of human genome annotation choice on RNA-seq expression estimates. BMC Bioinformatics. 2013; 14 Suppl 11: S8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGaralde DR, Snell EA, Jachimowicz D, et al.: Highly parallel direct RNA sequencing on an array of nanopores. Nat Methods. 2018; 15(3): 201–206. PubMed Abstract | Publisher Full Text\n\nLagarde J, Uszczynska-Ratajczak B, Carbonell S, et al.: High-throughput annotation of full-length long noncoding RNAs with capture long-read sequencing. Nat Genet. 2017; 49(12): 1731–1740. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLove KR, Shah KA, Whittaker CA, et al.: Comparative genomics and transcriptomics of Pichia pastoris. BMC Genomics. 2016; 17: 550. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang B, Tseng E, Regulski M, et al.: Unveiling the complexity of the maize transcriptome by single-molecule long-read sequencing. Nat Commun. 2016; 7: 11708. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "44979",
"date": "28 Feb 2019",
"name": "Sophie Shaw",
"expertise": [
"Reviewer Expertise Bioinformatics analysis of next generation sequencing data from a range of applications including varied projects using RNA sequencing data",
"and analysis of data from a broader scope of science including genome assembly and microbial community 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\nFroussios et al. have presented here a new tool, RATS, for the identification of differential transcript usage from transcript abundance estimates. RATs was benchmarked and compared to the existing tools DRIM-Seq and SUPPA2 across four different datasets. False positive rate, false negative rate, sensitivity and Matthews correlation coefficient were all measured. When considered as a whole, RATs was found to outperform the other tools. Differing results due to the version of the reference genome used are also discussed. This is a nicely presented manuscript, with well thought out comparisons. The tool will make a good addition to existing RNA sequencing analysis pipelines, especially as the field moves towards alignment free methods.\n\nThe rationale for the development of this tool is clearly stated, as there are only a few tools which carry out DTU detection from alignment-free RNA-seq quantifications. The majority of existing tools for DTE and DTU are designed for use with alignment- and assembly-based methods. Of the existing tools described, each has specific uses, and RATs has been presented as a broad \"differential transcript usage\" identification tool.\n\nThe methods of the analysis have been described well, and overall are technically sound. I would like to see an expansion on the description of the statistical method underpinning RATs. Although G-test of independence is cited, a brief description of what this entails and how it differs from existing tools would aid in the understanding of how the tool functions.\n\nHowever, I have some suggestions concerning the comparison of tools and the datasets selected. With regards to the selection of tools for comparison, SwitchSeq and iso-KTSP are mentioned within the introduction as being able to use transcript abundance estimates, however are not compared to. I assume that this is because they are too specialist in their identification of differential transcript usage and/or isoform switching, but I think the decision to not compare to these tools should be more explicit. The authors have not mentioned the recent pre-print from Cmero et al. (2019)1 which discusses the development of methods for DTU detection from alignment-free datasets using equivalence classes. The paper uses the same simulated datasets for benchmarking of the tool, and should be considered as another tool to compare to RATs. If this is not deemed as an equivalent method, it should at least be discussed in this manuscript.\n\nWith regards to the datasets tested, the published human data set which is shown here is not directly confirming the accuracy of RATs, as the authors show that the qPCR validation within the original study may be inaccurate, and underlying issues are present due to the reference genome version. Although the dataset is being used to compare RATs to SUPPA and DRIM-Seq, it is not validating the tool. I think that this manuscript would benefit from comparison of the three tools using another \"real-life\" dataset, which has been validated in some way, to support that RATs is detecting known DTU.\n\nMethods for tool development and testing are clearly described, apart from with false positive testing with A. thaliana dataset. The authors should include details on how the transcript abundances were produced for this (using Kallisto or Salmon? Any other pre-processing?). All datasets used are publicly available with accession numbers given. Additional data is provided within published links; however, these would benefit from a simple readme file, which explains the contents of each extended data file so the reader doesn't need to search through them.\n\nWithin the results, it would be nice to see more discussion on the impact of the bootstrapping information used by RATs. I think that this is a really beneficial part of this tool and this has not been demonstrated enough. It should also be made clearer if this bootstrapping information is obtained solely from Salmon/Kallisto or if RATs implements it's own bootstrapping.\n\nAlthough the testing of the simulated datasets does show that RATs outperforms DRIM-Seq and SUPPA2, I don't feel that as it stands you can conclude that the analysis of the published dataset shows this. When comparing to the published findings, SUPPA2 shows better results with confirmation of the qPCR results. As I've mentioned above, I would find another \"real-life\" dataset for comparison, or simply re-word the conclusion so that this isn't overstated.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Partly\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": []
},
{
"id": "44977",
"date": "04 Mar 2019",
"name": "James P. B. Lloyd",
"expertise": [
"Reviewer Expertise I have research experience in wet lab biology and dry lab (computational analysis). I have used many computational tools like RATS for analyses with real 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\n‘RATs’ is addressing an important problem: how to quantify changes in transcript isoform usage. Other tools, like ‘Sleuth’, address a related problem, which is differential transcript expression. Both RATs and Sleuth take advantage of the bootstrapping data that tools like ‘kallisto’ and ‘Salmon’ generate when quantifying transcript isoform abundance. By taking advantage of bootstrapping, such tools can estimate the technical variation within the data, to better look for differential changes. Transcript isoform usage is often linked to changes in alternative splicing or isoform specific decay rates (e.g. from NMD). Therefore, having a tool to accurately find changes in isoform usage is vital to our ability to address a range of biological problems.\n\nI have tried a version of RATs. I found that it was easy to install and easy to use. Being able to install bioinformatics software is no guarantee (Mangul et al., 20181). A bonus of RATs is that several figures can be generated from the data within the tool. This was simple to do, but allowed for you to visualize your data in a straightforward but powerful way. This is one of the rare tools that just works and was relatively intuitive and well-documented.\n\nThe paper uses sensible approaches to compare RATs to other tools, including ‘SUPPA’ and ‘DRIMSeq’ and the authors found that RATs performed at a similar or better level than the other tools.\n\nOne minor point that could be better explained is how RATs uses the bootstrapping data. Does it use it to simply throw out highly variable genes (decreasing FP rate) or does it help get closer to the true rate of biological variation, thus increasing the true positive rate?\n\nIn the methods, it would be good to see more explanation on how RATs does its pre-filtering. For example, if a transcript has zero expression in one treatment but a modest to high expression in the other treatment, would RATs keep this transcript or discard it? This would be of interest to people working on RNA decay pathways, such as NMD.\n\nOne thing that I would love to see is a comparison of RATs to the DEXSeq/DRIMSeq approach used to address differentiation transcript isoform usage (Love et al., 20182). This tool appeared to also perform well in the publication of this approach, where they used sim data. Therefore I think a comparison to RATs here, using sim and real data (human) would be appropriate.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Partly\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": "44975",
"date": "18 Mar 2019",
"name": "Michael I. Love",
"expertise": [
"Reviewer Expertise Statistical methods development for RNA-seq and other genomic assays"
],
"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 new method for detection of DTU from RNA-seq data, which uniquely leverages quantification uncertainty in the form of inferential replicates. I am not aware of other methods specifically designed to detect DTU as opposed to change in total expression level of the gene, which take into account quantification uncertainty. It is therefore a useful contribution to the methods literature. The authors have taken some length to assess their method against other popular methods on real and simulated datasets, and investigating individual genes with qRT-PCR validation in detail.\nI have some concerns about the conclusions from the evidence provided in the article, and additionally have requests for further details about the methods, which should be presented in the article itself.\n\nMajor comments:\n1) The methods are not sufficiently described, I have the following questions:\nWhat is the input to RATs? Is it TPM or counts or scaledTPM? Should the library size differences be removed prior to providing to RATs or does RATs take care of library size differences internally?\n\nCan the methods described all analyze the same type of experiment, are they all restricted to two-group analyses? Can any of them control for batch effects?\n\nWhat are the default pre-filtering and post-filtering settings? What is the default minimum abundance threshold or proportion threshold for an isoform to be considered expressed? What is the default effect size cutoff, and how is it implemented per isoform, per gene? What is the default fraction for determining that evidence of DTU is not substantiated across inferential replicates? Likewise, what default fraction for biological replicate variation?\n\n2) I didn't understand why abundance thresholds were not used, as described here, \"No transcript abundance pre-filter was imposed on any of the three DTU tools,\" and also \"As SUPPA2 offered no abundance pre-filtering, RATs and DRIM-Seq were run with abundance threshold values of 0.\"\n\nAs shown in Soneson et al. (20161) and Love et al. (20182), performance of a number of DTU methods is greatly improved by filtering out lowly expressed transcripts. It can be inferred from the title of the former paper: \"Isoform prefiltering improves performance of count-based methods for analysis of differential transcript usage\".\n\nSUPPA2 does have an abundance pre-filtering option, which was used in Love et al. (20182):\n\"We enabled a filter to remove transcripts with less than 1 TPM. TPM filtering is a command-line option available during the diffSplice step of SUPPA2 and this greatly improved the running time without loss of sensitivity\".\nFrom the SUPPA2 manual: \"-th | --tpm-threshold: Minimum expression (calculated as average TPM value within-replicates and between-conditions) to be included in the analysis. (Default: 0).\"\nGiven that all the methods have abundance and/or proportion filters available, that filters are recommended by at least two of the three methods in their documentation (DRIMSeq and RATs), and at least two independent review papers (not introducing methods) have shown that abundance and/or proportion filtering improves performance of methods, I can't see why the choice was made to not use filters.\n\n3) It is mentioned in the false positive analysis that the median FP fraction was less than 0.05 and a horizontal line is drawn on Fig 2A and B. This is misleading, as the adjusted p-values are being thresholded at 0.05 (I assume), and in a null comparison the rate of false positives from an adjusted p-values should be 0, not 0.05. Drawing or mentioning a 0.05 cutoff would be relevant for the p-values (uncorrected), but has no bearing on the adjusted p-values. This may confuse readers.\n\n4) The authors repeatedly refer to the reported effect size in DRIMSeq being an issue for comparison across methods, e.g. \"Direct comparison with DRIM-Seq is complicated by different methods for measuring DTU effect-sizes between the tools\", but this is only an issue to the extent that the authors wish to perform post-hoc filtering on effect size. It is not an issue for null hypothesis testing without post-hoc filtering, because all methods are testing against the null that the underlying proportion of expression across isoforms has the same distribution for control samples and treated samples. However, I agree that for post-hoc filtering, one may want to filter the methods in a similar manner. It should be easy to filter the DRIMSeq results directly on absolute difference in isoform proportion, for example in Love et al. (20182) we performed post-hoc filtering for DRIMSeq on the SD of proportions across all samples using a 6-line R function.\nAs the likelihood ratio statistic should be 1-1 and monotonic with the p-value for DRIMSeq (if the degrees of freedom is constant across genes or transcripts), then I would not compare effect size filtering with likelihood ratio filtering, as the latter is simply filtering the p-value at a lower threshold.\n\nMinor comments:\nIn the Introduction, the authors state \"there is little justification for choosing DGE over DTE in the study of complex transcriptomes\". The authors imply that gene-level and transcript-level analysis are mutually exclusive analyses, when they are not, and so I would suggest to reword or reconsider this statement. I and others have encouraged assessment of total changes in gene expression (DGE) as well as changes in isoform proportion (DTU), as both may be present in an experiment and both may be of biological importance to the system being studied. DGE has the property that the majority of inferential uncertainty which exists in an RNA-seq sample is removed (because it occurs across isoforms within genes), leaving inferential uncertainty from reads mapping across gene loci, but this property of reduced uncertainty does not preclude a transcript-level analysis. While DTE has advantages, the above sentence claiming that DGE has none overstates a more complex situation in my opinion.\n\nThroughout the paper, the authors refer to \"DRIMSeq\" as \"DRIM-Seq\" which is minor but different than the software and publication.\n\nFor what it's worth, the transcript-level test is similar conceptually to the current implementation of testForDEU() in DEXSeq which compares the expression of each feature to the sum of expression from all other features of the gene (this is also different from the test described in the original DEXSeq publication). Running DEXSeq on transcript estimated counts with testForDEU() was tested on simulated data in Soneson et al. (20161) and Love et al. (20182), and so such an approach has some evidence of working well for detection of isoform changes within a gene.\n\nI didn't understand what was meant by the following: \"the tables record the full provenance of the results\".\n\nIt is stated that, \"The performance results of RATs on these simulated datasets are in good agreement with those presented in Love et al (2018)\". However, this seems to be not clearly the case, which may be due to differences in the simulated data in the two articles, or some other reason. In the present article, DRIMSeq is reported as having lower sensitivity with lower achieved FDR than other methods, SUPPA2 has higher sensitivity and higher FDR, and RATs with various filter thresholds falls in between. In Love et al. (20182), DRIMSeq had the opposite performance: higher sensitivity but higher FDR relative to SUPPA2 and RATs run with default filters. However interpretation is made difficult by all the filtering options in Figure 3. It would be easier to compare perhaps if an additional supplementary plot to Figure 3 was made with only the default filter thresholds instead of the filter threshold ranges for all methods. The main commonality across the two benchmarks seems to be that RATs can achieve higher sensitivity than SUPPA2 while maintaining the same precision, for the 5% nominal FDR threshold.\n\nThis sentence needs to be made more specific, or else it could be misleading: \"As a consequence, the qRT-PCR intensities measured in the original study are actually impossible to interpret in the context of the updated annotation and the originally reported conclusion is likely wrong.\" Specifically which conclusion is likely wrong? From the analysis, it seemed like there is not a problem with the original qRT-PCR intensities and interpretation for at least one of the three genes.\n\nWhy is it perplexing that \"18% of the SUPPA2 results are rejected due to the effect size filter\". I didn't follow the authors in that statement.\n\nIt is stated, \"Existing tools rely on the mean isoform abundances...\". This implies that the mean of inferential replicates is used for statistical testing. It's perhaps subtly different, other methods are typically using the maximum likelihood estimate, which may be different than the mean of the bootstrap distribution, and different than the mean of the Gibbs sampling distribution. I would just say that other tools do not make use of inferential replication.\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": []
}
] | 1
|
https://f1000research.com/articles/8-213
|
https://f1000research.com/articles/7-448/v1
|
11 Apr 18
|
{
"type": "Opinion Article",
"title": "A novel ultrasound technique to detect early chronic kidney disease",
"authors": [
"Dulitha K. Hewadikaram",
"Mudhitha Bandara",
"Amal N. Pattivedana",
"Hiran H. E. Jayaweera",
"Kithsiri M. Jayananda",
"W. A. Monica Madhavi",
"Aruna Pallewatte",
"Channa Jayasumana",
"Sisira Siribaddana",
"Janaka P. Wansapura",
"Dulitha K. Hewadikaram",
"Mudhitha Bandara",
"Amal N. Pattivedana",
"Hiran H. E. Jayaweera",
"Kithsiri M. Jayananda",
"W. A. Monica Madhavi",
"Aruna Pallewatte",
"Channa Jayasumana",
"Sisira Siribaddana"
],
"abstract": "Chronic kidney disease (CKD) of unknown etiology is recognized as a major public health challenge and a leading cause of morbidity and mortality in the dry zone in Sri Lanka. CKD is asymptomatic and are diagnosed only in late stages. Evidence points to strong correlation between progression of CKD and kidney fibrosis. Several biochemical markers of renal fibrosis have been associated with progression of CKD. However, no marker is able to predict CKD consistently and accurately before being detected with traditional clinical tests (serum creatinine, and cystatin C, urine albumin or protein, and ultrasound scanning). In this paper, we hypothesize that fibrosis in the kidney, and therefore the severity of the disease, is reflected in the frequency spectrum of the scattered ultrasound from the kidney. We present a design of a simple ultrasound system, and a set of clinical and laboratory studies to identify spectral characteristics of the scattered ultrasound wave from the kidney that correlates with CKD. We believe that spectral parameters identified in these studies can be used to detect and stratify CKD at an earlier stage than what is possible with current markers of CKD.",
"keywords": [
"Chronic Kidney Disease of unknown etiology",
"Ultrasound spectral characteristics",
"Kidney fibrosis"
],
"content": "Introduction\n\nChronic kidney disease (CKD) is a major public health challenge and a leading cause of morbidity and mortality1. About 8 – 16% of the world population is affected by CKD2–4 with increased risk for end-stage renal disease, cardiovascular disease, and death5. To date, no specific treatment has shown to arrest the progression of CKD, except dialysis or kidney transplantation1. Considering the high cost of renal replacement therapy, the growing prevalence of CKD has implications for health and social care systems6 especially for developing nations. New variety of CKD has been identified among paddy farmers (known as CKD of unknown etiology (CKDu) or Chronic Interstitial Nephropathy among Agricultural Communities (CINAC)) in the North Central Province of Sri Lanka7. One fifth of the population in Anuradhapura, Polonnaruwa and Badulla districts suffer from CKDu and it has already become a major public health issue in Sri Lanka8\n\nCKD is silent killer because it starts insidiously and progresses slowly until end stage renal disease. The main challenges to improve outcomes in patients with CKD are the inability to identify patients with CKD in early subclinical stages9–11.\n\nEvidence suggests kidney fibrosis occurs in every type of CKD and leads to progressive and irreversible loss of renal function12,13. Progressive deposition of extracellular matrix in glomeruli (glomerulosclerosis) and/or interstitial space (tubulointerstitial fibrosis) is known as kidney fibrosis. Recent studies have hypothesized several biomarkers of renal fibrosis. Among them are, transforming growth factor-β1 (TGF-β1) a pro-fibrotic cytokine measured in urine and serum, bone morphogenetic protein-7 (BMP-7) recognized as a natural antagonist to TGF-β1 measured in serum and epidermal growth factor (EGF), a tubule-specific protein critical for cell differentiation and regeneration measured in urine. These markers of renal fibrosis have been associated with progression of CKD as measured by eGFR. However, no marker is able to predict CKD before being detected with traditional clinical tests (serum creatinine, and cystatin C, urine albumin, and ultrasound scanning) consistently and accurately. New biomarkers point to generalized processes that cause fibrosis, but they do not directly reflect kidney pathology. Biomarkers measured in serum or urine from CKD patients may not reflect the degree of kidney fibrosis and these should be corroborated with actual pathological measures of the kidney fibrosis in order to predict a given patient’s outcome.\n\nIn the context of quantification of fibrosis, several different techniques such as renal biopsy, magnetic resonance imaging (MRI), ultrasound (US) scanning can be utilized14,15. However, there are some limitations when it comes to renal biopsy as it does not cover the entire organ. The extent of fibrosis can be quantified by renal biopsy but it is an invasive procedure with associated risks16.\n\nMRI is ideally suited to quantify fibrosis via late Gadolinium enhancement and relaxation time measurement14,15,17. Previously, we have shown that young kidney patients on maintenance dialysis develop myocardial fibrosis quantifiable via MRI T2 relaxation time15. But MRI is an expensive modality, which is not readily accessible to the general population at risk of CKDu in Sri Lanka.\n\nFibrosis decreases the elasticity of tissue, hence measures of tissue elasticity is a surrogate marker of fibrosis. There are several techniques to measure elasticity in tissue using US. Among them transient elastography is the most common and widely used method. It measures tissue deformation while applying external pressure to the organ. Due to its retroperitoneal position this is not feasible in the kidney though it has been successfully used to quantitate liver fibrosis.\n\nAcoustic radiation force impulse (ARFI) imaging and shear wave velocity (SWV) are two other US elastography techniques. However, the reliability of these techniques in measuring kidney fibrosis has not been consistent18–20. Recent feasibility studies show that both ARFI and SWV failed to correlate with kidney fibrosis19,20. Unlike liver, the kidney is not homogenous in tissue character; it is more perfused, with two distinct zones and pathologically more complex. Therefore in our opinion, and as evident by these studies, elasticity of the kidney is probably a less reliable surrogate for kidney fibrosis.\n\n\nHypothesis\n\nOur approach to developing a non-invasive imaging method to detect early signs of CKD is based on the following hypothesis:\n\nDisease severity in CKD is associated with changes in the Fourier transform of the scattered ultrasound waves (Radio Frequency Spectrum) from the cortex of the CKD kidney.\n\nThe rationale for our hypotheses is based on the ultrasound physics.\n\nThe speckle patterns in B-Mode ultrasound images is the result of interference of scattered ultrasound waves from scatterers whose size is much smaller than the ultrasound wave length. B-mode ultrasound images are constructed from the amplitude modulation of the time domain scattered signal, known as the Radio Frequency echo (RF echo). In B-mode ultrasound the frequency dependent information of the RF echo is not utilized. However, theoretical and phantom studies21 have shown that the frequency spectrum of the RF echo (RF spectrum), can be related to microstructural properties such as shape, size, density and acoustical properties of tissue. Thus, changes in scatterer properties in tissue may affect the RF spectrum. On the other hand, due to its size and structure, accumulation of fibrosis could change the scatterer properties of tissue. Thus, it is plausible that increasing fibrosis in the CKD kidney, and therefore the severity of the disease, is associated with changes in the RF spectrum. This hypothesis is reinforced by studies in the liver showing correlation between B-mode ultrasound features and the presence of fibrosis22–24.\n\nA simple ultrasound system capable of RF echo signal acquisition and spectral analysis will be constructed to perform ex vivo and clinical studies. In the ex vivo experiment, correlations between the RF spectrum, speckle patterns and tissue characteristics will be investigated. In the clinical study, correlations between the RF spectrum, speckle pattern and the CKD stage will be investigated. In both cases speckle patterns will be used as the link between RF spectrum and fibrosis because speckle patterns are known to correlate with fibrotic stage in the liver22–24.\n\n\nThe portable US system\n\nMajority of the affected population consist of rural farmers who cannot afford to undergo regular medical screenings. Here we present an affordable and portable ultrasound system with the proposed technology that can be taken to the people at risk, where they live.\n\nThe proposed portable ultrasound probe system will have five major components. They are: waveform generator, high voltage amplifiers, a controlling unit consist of a Field Programmable Gate Array (FPGA) and an ARM microcontroller, ultrasound transducer, and an analog-to-digital converter. A block diagram of the system is shown in Figure 1.\n\nThe proposed system will have five major components: Pulse train generator, high voltage amplifiers, A controlling unit consist of a Field Programmable Gate Array (FPGA) and an ARM microcontroller, ultrasound transducer, and an analog-to-digital converter.\n\nThe controller will initiate a series of 1 – 10 MHz high voltage pulses through the ultrasonic transducer. The echo received by the same transducer, after suitable amplification and filtering will be digitized at the rate of 125 MS/s. This data will be processed by the microcontroller and then will be transmitted to a laptop computer via Wi-Fi for further analysis and display.\n\nA single element immersion transducer (Olympus V310-SU 5 MHz and V312-SU 10 MHz) will be used. Data analysis software will be developed and installed on the PC.\n\n\nProposed ex vivo experiment\n\nIn the laboratory experiment, RF spectrum and speckle pattern analysis will be performed on a range of ex vivo tissue (e.g. bovine liver, kidney tissue, etc.). The tissue sample will be completely submerged in degassed water (0.9% saline solution) and a 5 MHz or 10 MHz single element 0.25 in elemental diameter; immersion transducer (Olympus-V310-SU, Band width 16–24 MHz at -6 dB) will be used for scanning. Tissue sample will be held stationary and the transducer will move 200-micrometer step using computer controlled micro position system. RF echo data will be acquired using the proposed US system. A Hamming window will be applied to RF signal followed by fast Fourier transform (FFT) to calculate the RF power spectrum21. The RF spectrum will be normalized by a spectrum obtained from a standard phantom with identical acquisition parameters. The calibrated power spectrum will be fitted to a linear model over its bandwidth. This will give the standard spectral parameters: spectral intercept (dB; extrapolation to zero frequency) and spectral slope. In addition to the standard spectral parameters we will explore other statistical measures of the spectrum in order to identify attributes of the RF spectrum that are most sensitive to tissue characteristics.\n\nThe RF data from different ex vivo samples will be used to construct a B-Mode ultrasound image of the sample. This image will be analyzed for speckle features employing standard techniques such as statistical methods, model-based approaches, signal processing and geometrical analysis25. The speckle feature parameters will be tested on their sensitivity to differentiate ex vivo tissue types. The best performing speckle feature parameters will be analyzed to find correlations with RF spectrum parameters.\n\n\nProposed clinical study\n\nIn the clinical study, ultrasound imaging will be performed on human subjects with approval from the institutional ethical committee. The standard diagnostic criteria will be applied to diagnose CKD participants and they will be classified into stages (five) according to eGFR by using CKD-EPI equation26. The sample size will be based on the uncertainty of textural parameters determined in the experimental work.\n\nB-mode ultrasound imaging will be performed with standard imaging equipment to depict long-axis and transverse views of the kidneys. Kidney longitudinal length (Bi-parietal) will be measured for both kidneys. Additionally, kidney size, cortical echogenicity, parenchymal thickness and cortico-medullary demarcation will be recorded. The proposed US probe will be used to acquire RF echo data. The RF echo data will be transmitted to a remote server where the RF spectrum will be analyzed to quantitate RF spectrum parameters found in the laboratory experiment.\n\nAnonymized B-mode data will be used to stratify CKD patients using speckle parameters developed in experimental work. The Spearman’s correlation test will be used to assess any correlations between the clinical and biochemical data with B-mode speckle parameters. The positive predictive values of the B-mode speckle scoring system and RF spectrum parameters will be compared with the results of the CKD stages. It is expected that the combination of ex vivo and clinical study results will enable us to identify an optimal set of RF spectrum parameters that will be used to diagnose early signs of CKD.\n\nThe successful completion of this project will result in, a novel ultrasound parameter of CKD that can detect the disease at early stages and technology to construct a device that can make noninvasive diagnostic measurements of the kidney. A simple diagnostic tool that is portable will have significant impact on the future studies of CKD. This ultrasound device technology that will be developed in this study is potentially patentable and has a commercial value.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nFunding by National Research Council, Vidya Mawatha, Sri Lanka (NRC 16-044).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nLozano R, Naghavi M, Foreman K, et al.: Global and regional mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012; 380(9859): 2095–128. PubMed Abstract | Publisher Full Text\n\nCoresh J, Byrd-Holt D, Astor BC, et al.: Chronic kidney disease awareness, prevalence, and trends among U.S. adults, 1999 to 2000. J Am Soc Nephrol. 2005; 16(1): 180–8. PubMed Abstract | Publisher Full Text\n\nLevey AS, Atkins R, Coresh J, et al.: Chronic kidney disease as a global public health problem: approaches and initiatives - a position statement from Kidney Disease Improving Global Outcomes. Kidney Int. 2007; 72(3): 247–59. PubMed Abstract | Publisher Full Text\n\nJha V, Garcia-Garcia G, Iseki K, et al.: Chronic kidney disease: global dimension and perspectives. Lancet. 2013; 382(9888): 260–72. PubMed Abstract | Publisher Full Text\n\nWhitman IR, Feldman HI, Deo R: CKD and sudden cardiac death: epidemiology, mechanisms, and therapeutic approaches. J Am Soc Nephrol. 2012; 23(12): 1929–39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGlobal Burden of Disease 2013 Risk Factors Collaborators, Forouzanfar MH, Alexander L, et al.: Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks in 188 countries, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet. 2015; 386(10010): 2287–323. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJayasumana C, Gunatilake S, Siribaddana S: Simultaneous exposure to multiple heavy metals and glyphosate may contribute to Sri Lankan agricultural nephropathy. BMC Nephrol. 2015; 16: 103. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJayatilake N, Mendis S, Maheepala P, et al.: Chronic kidney disease of uncertain aetiology: prevalence and causative factors in a developing country. BMC Nephrol. 2013; 14: 180. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKavvadas P, Dussaule JC, Chatziantoniou C: Searching novel diagnostic markers and targets for therapy of CKD. Kidney Int Suppl (2011). 2014; 4(1): 53–57. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWong MG, Pollock CA: Biomarkers in kidney fibrosis: are they useful? Kidney Int Suppl (2011). 2014; 4(1): 79–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJu W, Nair V, Smith S, et al.: Tissue transcriptome-driven identification of epidermal growth factor as a chronic kidney disease biomarker. Sci Transl Med. 2015; 7(316): 316ra193. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoor P, Ostendorf T, Floege J: Renal fibrosis: novel insights into mechanisms and therapeutic targets. Nat Rev Nephrol. 2010; 6(11): 643–56. PubMed Abstract | Publisher Full Text\n\nMuñoz-Félix JM, González-Núñez M, Martínez-Salgado C, et al.: TGF-β/BMP proteins as therapeutic targets in renal fibrosis. Where have we arrived after 25 years of trials and tribulations? Pharmacol Ther. 2015; 156: 44–58. PubMed Abstract | Publisher Full Text\n\nDiwan A, Wansapura J, Syed FM, et al.: Nix-mediated apoptosis links myocardial fibrosis, cardiac remodeling, and hypertrophy decompensation. Circulation. 2008; 117(3): 396–404. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMalatesta-Muncher R, Wansapura J, Taylor M, et al.: Early cardiac dysfunction in pediatric patients on maintenance dialysis and post kidney transplant. Pediatr Nephrol. 2012; 27(7): 1157–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSerón D: Risk factors associated with the deterioration of renal function: the role of protocol biopsies. Prilozi. 2007; 28(1): 291–302. PubMed Abstract\n\nWansapura J, Hor KN, Mazur W, et al.: Left ventricular T2 distribution in Duchenne muscular dystrophy. J Cardiovasc Magn Reson. 2010; 12: 14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZaffanello M, Piacentini G, Bruno C, et al.: Renal elasticity quantification by acoustic radiation force impulse applied to the evaluation of kidney diseases: a review. J Investig Med. 2015; 63(4): 605–12. PubMed Abstract | Publisher Full Text\n\nWang L, Xia P, Lv K, et al.: Assessment of renal tissue elasticity by acoustic radiation force impulse quantification with histopathological correlation: preliminary experience in chronic kidney disease. Eur Radiol. 2014; 24(7): 1694–9. PubMed Abstract | Publisher Full Text\n\nTakata T, Koda M, Sugihara T, et al.: Renal shear wave velocity by acoustic radiation force impulse did not reflect advanced renal impairment. Nephrology (Carlton). 2016; 21(12): 1056–62. PubMed Abstract | Publisher Full Text\n\nLizzi FL, Feleppa EJ, Kaisar Alam S, et al.: Ultrasonic spectrum analysis for tissue evaluation. Pattern Recognit Lett. 2003; 24(4–5): 637–58. Publisher Full Text\n\nGao S, Peng Y, Guo H, et al.: Texture analysis and classification of ultrasound liver images. Biomed Mater Eng. 2014; 24(1): 1209–16. PubMed Abstract | Publisher Full Text\n\nCao G, Shi P, Hu B: Liver fibrosis identification based on ultrasound images. Conf Proc IEEE Eng Med Biol Soc. 2005; 6: 6317–20. PubMed Abstract | Publisher Full Text\n\nLayer G, Zuna I, Lorenz A, et al.: Computerized ultrasound B-scan texture analysis of experimental diffuse parenchymal liver disease: correlation with histopathology and tissue composition. J Clin Ultrasound. 1991; 19(4): 193–201. PubMed Abstract | Publisher Full Text\n\nDamerjian V, Tankyevych O, Souag N, et al.: A Speckle characterization methods in ultrasound images – A review. IRBM. 2014; 35(4): 202–13. Publisher Full Text\n\nLevey AS, Stevens LA, Schmid CH, et al.: A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009; 150(9): 604–12. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "41738",
"date": "10 Jan 2019",
"name": "Gaetano Lucisano",
"expertise": [
"Reviewer Expertise Ultrasound in Nephrology",
"Transplant Medicine",
"Transplant 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\nThis is the proposal of a prospective study aimed to assess the ability of a novel portable ultrasound (US) system to estimate the amount of renal fibrosis. The study would first consist in an ex-vivo experiment in which the US system is tested on bovine solid organs (kidney and liver), ultimately leading to the clinical study on human subjects. In the first instance, the novel US system would be employed as a screening tool for the early diagnosis and staging of chronic kidney disease in the rural population of Sri Lanka.\nThe idea and purposes are interesting. Few comments:\nIt is not clear whether the data obtained ex-vivo would be compared against histopathology data, as this would be imperative before proceeding with the clinical study. In this case the investigators should make sure that bovine kidneys with different degrees of fibrosis are analysed in order to properly assess the sensitivity and specificity of the US system.\n\nIt would be interesting to have an estimate of the saved costs of this technique employed for the rural population of Sri Lanka, as the application could be extended to the many other populations across the world with limited access to the healthcare system.\n\nIn my view, it is important to point out that, if successful, this US system is aimed to flag patients with suspected CKD - who will eventually need a biochemical estimation of the degree of CKD, as I doubt this system would have sufficient sensitivity to warrant a fairly accurate monitoring of the progression of the CKD (changes in the degree of fibrosis detected by the US system would probably correspond to dramatic decreases of the eGFR).\n\nIn the clinical study, the obtained B-mode parameters should be corrected for anthropometric features, such as the body height. This would strengthen the degree of correlation between CKD stage and US parameters (Lucisano et al., 20151).\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": "4434",
"date": "20 Feb 2019",
"name": "Janaka Wansapura",
"role": "Author Response",
"response": "Response to reviewer:1) It is not clear whether the data obtained ex-vivo would be compared against histopathology data, as this would be imperative before proceeding with the clinical study. In this case the investigators should make sure that bovine kidneys with different degrees of fibrosis are analysed in order to properly assess the sensitivity and specificity of the US system.We agree with the reviewer’s concerns and have now included a section for sensitivity analysis as follows:To assess the sensitivity and the specificity of the US system, histopathology data of human kidneys will be used. Autopsied human kidneys with different level of CKD will be collected form the pathologists. Kidney fibrosis will be determined by calculating the percentage of Masson’s Trichrome stained area of interstitial fibrosis per total area of kidney tissue. The stained areas will be analyzed with available image software, and the levels of lesion formation will be expressed as percent lesion area per total area. Kidney samples will be scanned using the immersion US probe. Back scattered US signal will be collected and analyzed to determine the sensitivity of RF and speckle feature parameters to different levels of kidney fibrosis in the autopsied human kidneys.2) It would be interesting to have an estimate of the saved costs of this technique employed for the rural population of Sri Lanka, as the application could be extended to the many other populations across the world with limited access to the healthcare system.The following text was added:It is estimated that a prototype of the proposed US system can be built for approximately USD 1400.3) In my view, it is important to point out that, if successful, this US system is aimed to flag patients with suspected CKD - who will eventually need a biochemical estimation of the degree of CKD, as I doubt this system would have sufficient sensitivity to warrant a fairly accurate monitoring of the progression of the CKD (changes in the degree of fibrosis detected by the US system would probably correspond to dramatic decreases of the eGFR)We understand the reviewer’s concern regarding the sensitivity of the proposed US system. The proposed clinical study, by way of testing the hypothesis of this paper, is only aimed at establishing a novel set of US based markers for kidney fibrosis. To study the sensitivity and specificity of these makers in relation to CKD, a rigorous clinical study must be performed and the data must be corroborated with established markers such as eGFR. We think such a study is beyond the scope of this hypothesis paper.4) In the clinical study, the obtained B-mode parameters should be corrected for anthropometric features, such as the body height. This would strengthen the degree of correlation between CKD stage and US parameters (Lucisano et al., 20151).As proposed in this study, both ex vivo and clinical study results will be used to identify optimal set of RF spectrum parameters that will lead to identifying CKD at early stages. As Lucisano et al. observed, all clinical results such as one-axis and transverse views of the kidneys, Kidney longitudinal length (Bi-parietal), kidney size, cortical echogenicity, parenchymal thickness and cortico-medullary demarcation, etc. will be corrected for anthropometric features such as body height etc.The following text was added to the manuscript:All clinical results will be normalized for anthropometric features such as body height etc. (27)."
}
]
},
{
"id": "41537",
"date": "23 Jan 2019",
"name": "Federico Nalesso",
"expertise": [
"Reviewer Expertise CKD",
"AKI",
"US in AKI/CKD. Critical care nephrology."
],
"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 introduces a new very interesting ultrasound examination to be used for the study of CKD.\nThe parameters used to perform the ultrasound measurements to be correlated with renal fibrosis appear to be very difficult to measure and operator-dependent. The standardization appears difficult and very difficult to obtain. The vascular component during fibrosis is not taken into consideration and the correlation of the new ultrasound measurement with renal resistance indices is not analyzed.\nCKD is reduced to the fibrotic component only, without considering the vascular alterations that underlie interstitial changes.\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? No\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-448
|
https://f1000research.com/articles/7-1392/v1
|
03 Sep 18
|
{
"type": "Research Article",
"title": "Purchase of medications without prescription in Peru: a cross-sectional population-based study",
"authors": [
"Akram Hernández-Vásquez",
"Christoper A. Alarcon-Ruiz",
"Deysi Díaz-Seijas",
"Luisa Magallanes-Quevedo",
"Diego Rosselli",
"Christoper A. Alarcon-Ruiz",
"Deysi Díaz-Seijas",
"Luisa Magallanes-Quevedo",
"Diego Rosselli"
],
"abstract": "Background: Low availability of medicines in health services, self-medication, inadequate use of medicines, and inadequate dispensing practices in pharmacies are frequent problems in Peru. We aimed to evaluate how frequent the purchase of medications without medical prescription is in Peru, and which factors are associated with this practice. Methods: We conducted a secondary analysis of the 2016 ENSUSALUD national survey data. Purchase of medicines that require a prescription was measured as a dichotomous coded as bought one or more medicines that requires medical prescription or bought medicines that do not require a prescription. Crude and adjusted prevalence ratios (PR) and their 95% confidence intervals (95% CI) were calculated using Poisson regressions model with robust variance to assess the association of purchasing of medicines that require prescriptions with sociodemographic factors. Results: There were 3858 participants in the dataset. The prevalence of purchasing medications without prescriptions was 47.2%. History of having previously consumed the same medication (31.6%), and the delay in receiving an appointment at health facilities (26.9%) were the main reasons to buy medications without a prescription. Regarding the recommendation of the medication purchased, the advice of the pharmacy, and remembering a previous old prescription, were the most frequent reasons (38.3%, and 25.9%, respectively). On the multivariable analysis, users that buy medications without prescription were more likely to be of aged 24-45; reside in the Amazon and Highlands regions; and self-consumption of the purchase. Individuals with Seguro Integral de Salud (Comprehensive Health Insurance) were less likely to buy medications without prescription. Conclusions: There is a high prevalence of prescription requiring medication being bought without one from pharmacies in Peru. It is necessary to include the evaluation of consumer patterns to develop strategies with the aim to regulate the consumption of prescription drugs in the Peruvian population.",
"keywords": [
"Nonprescription Drugs",
"Self Medication",
"Drug Utilization",
"Drug Seeking Behavior"
],
"content": "Introduction\n\nSelf-medication is the use of medicines to treat symptoms or disorders identified by the users without prescription, or the intermittent or continued use of a medicine prescribed by a physician to treat recurring or chronic symptoms or illnesses of the user or their family members, especially children or elderly people1. It is a common practice both in developed and developing countries that may cause potential risks due to an incorrect self-diagnosis, failures in treatment, risk of overdose, medical interactions, severe adverse effects, among others2,3.\n\nSeveral studies show that patterns of self-medication widely vary among populations, and these patterns are influenced by multiple factors, such as age, sex, income, expenses, orientation to self-care, level of education, medical knowledge, satisfaction with health services, and perception of illness1,4. A systematic review of publications from different countries show a prevalence of self-medication ranging between 12.7% and 71%5.\n\nIn Peru, important processes of health reform have been developed, meaning advances to achieve an increased coverage of universal insurance in health, and defense of users’ rights6. However, it was reported that 30.6% of patients had an insufficient access to medicines which patients requested in pharmacies of Peruvian health facilities. Furthermore, 20.7% of users of pharmacies of public health facilities where they had been treated were told to purchase medicines in an external pharmacy7. Other studies have shown that 13% of adults who purchased antibiotics for children under five years bought them in private pharmacies without medical prescription. These results could demonstrate a low availability of medicines in health facilities, which added to self-medication, an inadequate use of medicines, and inadequate dispensing practices in pharmacies, represent a critical problem in the Peruvian health system with prevalence of self-medication ranging from 10.5% to 87.8%8–11.\n\nIn Peru, the National Survey on User Satisfaction of Health Services (Encuesta Nacional de Satisfacción de Usuarios en Salud, ENSUSALUD) has been conducted since 2014, this survey includes information related to the purchase of medicines and its results are nationally representative12; however, to our best knowledge, studies using this survey have not been conducted. Such studies might assess a very frequent practice in the country such as the purchase of medicines without medical prescription.\n\nTo that effect, the objective of this study was to assess the prevalence of purchasing medicines which required a medical prescription in Peru without prescription, to know their characteristics, and to identify factors related to this practice.\n\n\nMethods\n\nWe conducted a secondary analysis of database from the Module of Pharmacies’ Users of the ENSUSALUD 2016 prepared by the National Superintendency of Health (Superintendencia Nacional de Salud, SUSALUD) and the National Institute of Statistics and Data Processing (Instituto Nacional de Estadística e Informática, INEI)12. All complete cases of people older than 15 years were included for the analysis.\n\nENSUSALUD includes questionnaires (available from the SUSALUD website) addressed to health professionals, executives of health services, and users of pharmacies and health services, and their objective is to monitor and assess the functioning and performance of the Peruvian health system, by studying the main participants in health care processes provided by health facilities13. The survey is conducted yearly using face-to-face questionnaires and in its 2016 edition, this survey took place from May 13 to July 9. ENSUSALUD is nationally representative, by applying a probabilistic, stratified and two-stages sampling with a simple random selection in the first stage and a systematic selection in the second stage, and using a level of confidence of 95% and a sampling error of +/-5%. ENSUSALUD obtains information of people older than 15 years who went to buy some medicine for themselves, their partners or their son/daughter in a pharmacy located up to two blocks around the health facilities selected to participate in the study.\n\nThe databases of ENSUSALUD 2016 public and can be obtained SUSALUD website.\n\nFor this analysis all the questions were obtained from the ENSUSALUD questionnaire. The dependent variable “purchase of prescription medicines without medical prescription” was determined by using the question “Did you buy this/these medicine(s) using a prescription?” for each one of the medicines purchased in order to determine if the interviewee had purchased at least one prescription medicine without medical prescription. The classification of each one of the medicines dependent on if they required a medical prescription or not for their sale was conducted by two investigators independently, based on Health Registration of Pharmaceutical Products of the General Bureau for Medicines, Drugs, and Inputs of Peru, in which official information of the national regulatory agency is included for all authorized medicines sold in Peru. The independent variables included are: sex, age group, level of education, type of health insurance, geographical area of residency, language, consumption of purchased products, and request of prescription by a pharmacist.\n\nWith regards to independent variables, it is important to point out that in Peru health insurance can correspond to one of the four subsystems: public system, which subsidizes health services to low-income population; social security system (EsSalud), which provides formal workers and their dependants with services; health system of armed forces and police (FF.AA.); and private system, available to people who can pay directly for a service or by a private insurance company14,15. The country is divided into 24 political-administrative departments which are distributed in three natural regions: Coast region, adjacent to the Pacific Ocean (included Metropolitan Lima, capital city of the country); Highlands region, where the Andes are located; and Jungle region, which is part of the Amazon rainforest.\n\nAdditionally, the following questions were included in the analysis: “Why did you buy medicines without prescription?” and “Who recommended you these medicines?” with the purpose of knowing the reasons for purchasing the medicine(s) and people responsible for recommending medicines purchased by survey respondents.\n\nThe analysis of data was conducted using Stata® software v14.2 (Stata Corporation, College Station, Texas, USA). In the first stage, characteristics of study populations were described by absolute frequencies and percentages, their weighted proportions were also estimated with a 95% confidence interval based on the purchase of medicines without prescription. The second stage was composed by a bivariate and multivariate statistical analysis, for that reason generalized linear models (GLM) of Poisson family and the log linking function were used to determine the association between the purchase of medicine without prescription and the independent variables. For the multivariate analysis, independent variables with a minimal association (p<0.2) with the dependent variable were included, and reasons of prevalence with 95% confidence interval (95% CI) were reported. The characteristics of sampling of the survey were specified, they include weightings according to strata, expansion factor, and design; for this purpose, the svy command was used. For all statistical testing a p value of <0.05 was considered statistically significant.\n\nThis study did not require any approval of an ethics committee, because it is a secondary analysis of a database of free access and public domain which does not identify survey respondents or pharmacies where purchases were made.\n\n\nResults\n\nThe Module of Pharmacies’ Users of ENSUSALUD included a total of 3858 survey respondents (expanded population: 3,529,791), of that 1892 participants purchased at least one prescription medicine without medical prescription, which represents a prevalence of 47.2% (95% CI: 45.3-49.1). The average age of the survey respondents who purchased without medical prescription was 38.4 ± 14.1. The survey respondents with private health insurance (57%), with university education (51.4%), who are residents of the Highlands region (57.1%) purchased more often without a medical prescription (Table 1). Additionally, the pharmacist in 69.6% of the purchases did not enquire for a prescription.\n\n*Weighted mass and specifications of sampling of ENSUSALUD 2016 were included\n\n**In Peru, there are 24 political-administrative departments which are distributed in three natural regions: Coast region, adjacent to the Pacific Ocean (included Metropolitan Lima, capital city of the country); Highlands region, where Andes are located; and Jungle region, which is part of the Amazon rainforest.\n\nSD: Standard deviation. FF.AA: Health System of Armed Forces and Police.\n\nThe precedent of previously using the medicine (31.6%), the delay in obtaining an appointment for health care (26.9%), and the delay for health care in health centers (21.5%) were the main causes of purchasing prescription medicines without medical prescription. Regarding the individual who recommended the purchased medicine (Table 2), a piece of advice of a salesperson of pharmacy/pharmacist, recalling a previous prescription and a piece of advice of a relative, were the more frequent reasons (38.3%, 25.9%, and 11.7%, respectively).\n\n*Question of multiple answer\n\nBivariate analysis found an association with a value p<0.2 between the purchase of medicines without prescription and all variables of the study (Table 3). The multivariate analysis showed that being in an age between 25-44 (PRa: 1.11; 95% CI: 1.01-1.22); living in the Highlands region (PRa: 1.39; 95% CI: 1.24-1.57) and Jungle region (PRa: 1.37; 95% CI: 1.23-1.52); and the consumption of purchased products themselves; were associated with a greater prevalence of purchasing medicines without a prescription. Furthermore, having Comprehensive Health Insurance (SIS, for Spanish) (PRa: 0.90; 95% CI: 0.83-0.98) and the pharmacist requesting a prescription (PRa 0.12; 95% CI: 0.10-0.15) were associated with a lower prevalence of purchasing without a prescription for the adjusted model (Table 3).\n\n*Regression model of Poisson with a log link function, robust variance and weighting of sampling.\n\n¶Adjusted by all variables showed in the column and which obtained a value of p<0.2 in crude analysis.\n\nPR: prevalence ratio. FF.AA: Health System of Armed Forces and Police.\n\n\nDiscussion\n\nIn this study, a high prevalence of purchasing prescription medicines without medical prescription was found according to the National Authority data. This practice was associated to the age group ranging from 25 to 44 years, inhabitants of the Highlands and Jungle regions, and when the purchase was for self-consumption. Having Comprehensive Health Insurance (SIS) and the pharmacist requesting a prescription would be factors that decrease the prevalence of purchase without prescription. Finally, the advice of the pharmacist, a salesperson of the pharmacy and/or any relative, and the previous use of the medication are cited information sources for the user going to a pharmacy to purchase a medicine without prescription.\n\nSelf-medication and the acquisition of medicines without prescription are common worldwide, with a prevalence range varying between 19% and 83%3,16,17. The Peruvian natural regions of Highlands and Jungle showed the highest incidences of these practices. Inadequate and non-effective access to medicines has been a common problem in Peru over recent years18, existing in the previously mentioned regions7, which would explain the results found in these regions of the country.\n\nAmong the reasons for the purchase of medicines without prescription, the delay to make an appointment or delay in receiving health care at the health centre, and previous knowledge of the medicine to be purchased stand out as the most important in our results; this situation was previously reported in national19 and international20,21 studies. That would reflect the discontent of the users regarding the health service in the Peruvian health centers, where an average of 18 days is expected to make an appointment and 104 minutes to be treated in an outpatient visit after their arrival to the health center for 201522.\n\nThe practice of purchasing medicines in the surveyed users is mainly focused on self-consumption. It is reported that most of these self-medication practices are developed in the context of self-consumption or for people under someone’s care, like children and elderly people4. A study carried out in eight Latin American countries about responsible self-medication showed that users have a positive attitude towards health care, they recognize that over-the-counter medicines are less dangerous than prescription medicine, they are also interested in reading the labels of the products to be informed about the possible adverse effects23. Thus, with a proper previous education for participants regarding the rational use of medicines, self-medication could produce potential benefits, like the improvement to the access to over-the-counter medicines, greater patient empowerment, and avoidance of unnecessary medical appointments amongst others24,25.\n\nThe control of the pharmacies over the sale of prescription products is important, including antimicrobial medication due to the reported increase in the resistance to these products related to self-medication. This situation occurs both in developed and developing countries11,25,26. Regarding a recommendation for the purchase of medicines, the pharmacist or salesperson of the pharmacy was the person that mainly provided this information. This situation has been previously reported in other regions around the world3. Previous conducted studies in the country report the recommendation of medicines other than those in the prescription or a change of prescription, including antibiotics, by the pharmacist or personnel of pharmacies, despite they are not entitled for these practices27,28. The advice of a relative and the previous consumption of a medicine are other important sources of recommendation for the purchase without prescription. Studies performed in a district of Metropolitan Lima, capital city of Peru, and in Cajamarca, had the same findings8,11. One of the main problems of this practice is the consumption of medicines that require prescription, such as antibiotics. Antimicrobial resistance has been shown to be associated with inappropriate antibiotic prescription, either for the wrong indication or incorrect duration of treatment16,29,30. Allergies to drugs or use of antibiotics to treat viral diseases are other frequent problems27,28.\n\nFrom these finding it can be seen that it is crucial to apply measures that guarantee effective access to medicines in health facilities, and to establish procedures that allow verification of proper practices in the purchase of prescription medicines in private pharmacies. The proposals should be an initiative of the national health authority (Ministry of Health) in order to end this phenomenon, taking into consideration the characteristics of people who purchase medicines without prescription.\n\nAmong the measures to be included in such a policy are: rationalization of the medicines provision and improvements in the main four stages of the medicines management cycle (selection, purchase, distribution, and use) through actions like: the development of information and control systems; the implementation of proper storage mechanisms and development of clinical practice guidebooks procedures (to promote rational prescription); optimization of the efficiency of investments in supply systems for public provision of medicines; implementation of centralized medication purchases; national and joint international bidding to reduce drug prices; rational prescription, distribution and consumption; among others31.\n\nThe national representation of the results and the characterization of each one of the medicines to establish if they require a medical prescription for their purchase are the main strengthens of the study. Also, this is one of the first studies in the country to analyze the relation between socioeconomic factors and the purchase of prescription medicines without medical prescription at national level. However, given that it is a data analysis from secondary sources, there is a possibility that the collected data are not accurate. Another limitation is that it does not allow to establish a causal connection due to the cross-sectional design of the study. Also, ENSUSALUD does not include some morbidity variables in the literature that can affect results17. Despite that, we consider that our analysis is a good approach to the study of this problem, which represents a greater negative impact in countries with segmented health systems or in developing countries, as in the Peruvian case.\n\nIn conclusion, the purchase of prescription medicines without medical prescription is a common practice within the Peruvian health system and is more frequent in regions with a lower supply of medical facilities or access to health services. The development of future programs that allow and/or limit the access to medicines should evaluate the current practices of sale of products without the required prescription in the Peruvian population and the practices of self-medication to strengthen the sale control of these products and improve the access in regions in greater need.\n\n\nData availability\n\nThe 2016 Encuesta Nacional de Satisfacción de Usuarios en Salud (ENSUSALUD) data are available at from the National Superintendency of Health (Superintendencia Nacional de Salud, SUSALUD) website: http://portal.susalud.gob.pe/blog/base-de-datos-2016/",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nSupplementary material\n\nQuestionnaires used for this study are available in Spanish from the National Superintendency of Health (Superintendencia Nacional de Salud, SUSALUD) website: http://portal.susalud.gob.pe/wp-content/uploads/archivo/encuesta-sat-nac/2016/Cuestionario-Usuarios-en-boticas-y-farmacias.pdf\n\n\nReferences\n\nWorld Health Organization: Guidelines for the regulatory assessment of medicinal products for use in self-medication. [Accessed 20 June 2016]. Reference Source\n\nMontastruc JL, Bondon-Guitton E, Abadie D, et al.: Pharmacovigilance, risks and adverse effects of self-medication. Therapie. 2016; 71(2): 257–62. PubMed Abstract | Publisher Full Text\n\nKhalifeh MM, Moore ND, Salameh PR: Self-medication misuse in the Middle East: a systematic literature review. Pharmacol Res Perspect. 2017; 5(4): e00323. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorld Self-medication Industry: What is self-medication? [Accessed 10 June 2017]. Reference Source\n\nShaghaghi A, Asadi M, Allahverdipour H: Predictors of Self-Medication Behavior: A Systematic Review. Iran J Public Health. 2014; 43(2): 136–46. PubMed Abstract | Free Full Text\n\nVelasquez A, Suarez D, Nepo-Linares E: [Health sector reform in Peru: Law, governance, universal coverage, and responses to health risks]. Rev Peru Med Exp Salud Publica. 2016; 33(3): 546–55. PubMed Abstract\n\nMezones-Holguín E, Solis-Cóndor R, Benites-Zapata VA, et al.: [Institutional differences in the ineffective access to prescription medication in health care centers in Peru: analysis of the National Survey on User Satisfaction of Health Services (ENSUSALUD 2014)]. Rev Peru Med Exp Salud Publica. 2016; 33(2): 205–14. PubMed Abstract\n\nLlanos-Zavalaga LF, Contreras-Ríos CE, Velásquez-Hurtado JE, et al.: [Self-medication in five provinces from Cajamarca]. Rev Med Hered. 2001; 12(4): 127–33.\n\nMiní E, Varas R, Vicuña Y, et al.: [Self-medication behavior among pregnant women user of the Instituto Nacional Materno Perinatal, Peru 2011]. Rev Peru Med Exp Salud Publica. 2012; 29(2): 212–7. PubMed Abstract\n\nPillaca-Medica ML, Carrión-Dominquez K: Automedicación en personas adultas que acuden a boticas del Distrito Jesús Nazareno, Ayacucho 2015. An Fac med. 2016; 77(4): 387–92. Publisher Full Text\n\nHermoza-Moquillaza R, Loza-Munarriz C, Rodríguez-Hurtado D, et al.: Automedicación en un distrito de Lima Metropolitana, Perú. Rev Med Hered. 2016; 27(1): 15–21. Publisher Full Text\n\nSuperintendencia Nacional de Salud: [National Survey of Healthcare Users Satisfacion (ENSUSALUD 2016)]. [Accesed 20 Jun 2017]. Reference Source\n\nInstituto Nacional de Estadística e Informática: [National Survey of Healthcare Users Satisfacion 2016. Final Report]. [Accessed 20 Nov 2017]. Reference Source\n\nSánchez-Moreno F: [The national health system in Peru]. Rev Peru Med Exp Salud Publica. 2014; 31(4): 747–753. PubMed Abstract\n\nAlcalde-Rabanal JE, Lazo-González O, Nigenda G: [The health system of Peru]. Salud Publica Mex. 2011; 53 Suppl 2: S243–4. PubMed Abstract\n\nAlhomoud F, Aljamea Z, Almahasnah R, et al.: Self-medication and self-prescription with antibiotics in the Middle East-do they really happen? A systematic review of the prevalence, possible reasons, and outcomes. Int J Infect Dis. 2017; 57: 3–12. PubMed Abstract | Publisher Full Text\n\nLocquet M, Honvo G, Rabenda V, et al.: Adverse Health Events Related to Self-Medication Practices Among Elderly: A Systematic Review. Drugs Aging. 2017; 34(5): 359–65. PubMed Abstract | Publisher Full Text\n\nContribuyentes por respeto: [Drug evaluation and purchase in the Peruvian state. Two sides of the same shady coin]. [Accessed 08 December 2017].\n\nEcker L, Ruiz J, Vargas M, et al.: [Prevalence of purchase of antibiotics without prescription and antibiotic recommendation practices for children under five years of age in private pharmacies in peri-urban areas of Lima, Peru]. Rev Peru Med Exp Salud Publica. 2016; 33(2): 215–23. PubMed Abstract\n\nJafari F, Khatony A, Rahmani E: Prevalence of self-medication among the elderly in Kermanshah-Iran. Glob J Health Sci. 2015; 7(2): 360–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShankar PR, Partha P, Shenoy N: Self-medication and non-doctor prescription practices in Pokhara valley, Western Nepal: a questionnaire-based study. BMC Fam Pract. 2002; 3: 17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nInstituto Nacional de Estadística e Informática: [National Survey of Healthcare Users Satisfacion 2015. Final Report]. [Accessed 10 October 2017]. Reference Source\n\nBolaños H: Responsible self-medication in Latin America. Drug Inf J. 2005; 39(1): 99–107. Publisher Full Text\n\nWorld Health Organization: Report of the WHO Expert Committee on National Drug Policies 1995. [Accessed 16 December 2017]. Reference Source\n\nHughes CM, McElnay JC, Fleming GF: Benefits and risks of self medication. Drug Saf. 2001; 24(14): 1027–37. PubMed Abstract | Publisher Full Text\n\nRather IA, Kim BC, Bajpai VK, et al.: Self-medication and antibiotic resistance: Crisis, current challenges, and prevention. Saudi J Biol Sci. 2017; 24(4): 808–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVacca CP, Niño CY, Reveiz L: [Restriction of antibiotic sales in pharmacies in Bogotá, Colombia: a descriptive study]. Rev Panam Salud Publica. 2011; 30(6): 586–91. PubMed Abstract\n\nJara-Romero L, Camizán-Cunias A, Cornejo-Atoche D, et al.: [Alterations in drug dispensation by private sector pharmacies in the district of Chiclayo]. Rev cuerpo méd HNAAA. 2012; 5(1): 26–9. Reference Source\n\nGrigoryan L, Haaijer-Ruskamp FM, Burgerhof JG, et al.: Self-medication with antimicrobial drugs in Europe. Emerg Infect Dis. 2006; 12(3): 452–9. PubMed Abstract | Free Full Text\n\nDu Y, Knopf H: Self-medication among children and adolescents in Germany: results of the National Health Survey for Children and Adolescents (KiGGS). Br J Clin Pharmacol. 2009; 68(4): 599–608. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTobar F: [Lessons learned from drug supplies for primary health care]. Salud Publica Mex. 2008; 50 Suppl 4: S463–9. PubMed Abstract"
}
|
[
{
"id": "41709",
"date": "27 Dec 2018",
"name": "Sandipan Bhattacharjee",
"expertise": [
"Reviewer Expertise I believe that I am qualified to review this manuscript given my extensive experience with national level data sources."
],
"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 titled \"Purchase of medications without prescription in Peru: a cross-sectional population-based study\" is an interesting and relevant study from the perspective of Peru. The authors have done a good job. However, some of the following issues should be addressed prior to indexing.\nIn the abstract, the author mention age range of 24-45, but in the study they have used 25-44 years. This should be addressed.\n\nI would suggest suing the Jungle region term consistently in the abstract as well (instead of Amazon).\n\nIn the second paragraph of the Introduction, the authors mention \"several studies\" but only provide two studies as reference (1, 4). If it is only two studies, then it should be called a few or handful studies. Also, where were these studies conducted and how are they relevant in the context of Peru?\n\nIn the third paragraph of the Introduction section, some important statistics are provided. But it might be more appropriate to mention when these changes happened etc.\n\nThe weighted percentage column of Table 1 is confusing. Are they column percentages? If so, then why are they not adding up to 100%. The authors need to clarify this issue.\n\nIn Table 2, there is a mention of \"Other\" reason to purchase medicines without medical prescription. I am wondering if the respondents were allowed to provide comments on them? If yes, then some interesting comments should be provided either in this Table or in the text somewhere in the Results section.\n\nAnd finally, although this manuscript is rather well written, but in some places English editing is required for clarification (e.g. - second sentence of the Methods section of Abstract is not very clear to me).\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": "38798",
"date": "23 Jan 2019",
"name": "Marcus Tolentino Silva",
"expertise": [
"Reviewer Expertise epidemiology",
"health technology assessment"
],
"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 nationwide based study of Peru that measured self-medication by adults (>15 years old) in 2016. The authors showed that poor access to health services increases self-medication. I would like to add three suggestions. First, investigate colinearity between independent variables to avoid type I errors. Second, discuss the measurement bias: the self-medication was self-reported and may be influenced by memory bias (last 12 months) or non-response bias (for drugs for special control, like opioids, etc.). Third, add reference from Latin America1.\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": "43615",
"date": "04 Feb 2019",
"name": "Fernando M. Runzer-Colmenares",
"expertise": [
"Reviewer Expertise Geriatrics"
],
"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\nInteresting article, finely written. Introduction reports adequately about antecedents and justification of research. Methods reports designs, variables, and statistical analysis. Results and analysis are performed adequately in my opinion and discussion is enriched with comparison with recent publications about the research question, that's why I consider this paper as suitable for publication. The authors may detail the criteria used to include variables in the regression model.\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": []
}
] | 1
|
https://f1000research.com/articles/7-1392
|
https://f1000research.com/articles/7-1083/v1
|
16 Jul 18
|
{
"type": "Research Article",
"title": "A nested randomised controlled trial of a newsletter and Post-it® note did not increase postal questionnaire response rates in a falls prevention trial",
"authors": [
"Sara Rodgers",
"Illary Sbizzera",
"Sarah Cockayne",
"Caroline Fairhurst",
"Sarah E. Lamb",
"Wesley Vernon",
"Judith Watson",
"Catherine Hewitt",
"David J. Torgerson",
"Illary Sbizzera",
"Sarah Cockayne",
"Caroline Fairhurst",
"Sarah E. Lamb",
"Wesley Vernon",
"Judith Watson",
"Catherine Hewitt",
"David J. Torgerson"
],
"abstract": "Background: Attrition (i.e. when participants do not return the questionnaires) is a problem for many randomised controlled trials. The resultant loss of data leads to a reduction in statistical power and can lead to bias. The aim of this study was to assess whether a pre-notification newsletter and/or a handwritten or printed Post-it® note sticker, as a reminder, increased postal questionnaire response rates for participants of randomised controlled trials. Method: This study was a factorial trial embedded within a trial of a falls-prevention intervention among men and women aged ≥65 years under podiatric care. Participants were randomised into one of six groups: newsletter plus handwritten Post-it®; newsletter plus printed Post-it®; newsletter only; handwritten Post-it® only; printed Post-it® only; or no newsletter or Post-it®. The results were combined with those from previous embedded randomised controlled trials in a meta-analysis. Results: The 12-month response rate was 803/826 (97.2%) (newsletter 95.1%, no newsletter 99.3%, printed Post-it® 97.5%, handwritten Post-it® 97.1%, no Post-it® 97.1%). Pre-notification with a newsletter had a detrimental effect on response rates (adjusted odds ratio (OR), 0.14; 95% CI, 0.04 to 0.48; p<0.01) and time to return the questionnaire (adjusted hazard ratio, 0.86; 95% CI, 0.75 to 0.99; p=0.04). No other statistically significant differences were observed between the intervention groups on response rates, time to response, and the need for a reminder. Conclusions: Post-it® notes have been shown to be ineffective in three embedded trials, whereas the evidence for newsletter reminders is still uncertain.",
"keywords": [
"Randomised controlled trial",
"randomisation",
"embedded trial",
"newsletter",
"Post-it® note",
"response rate"
],
"content": "Introduction\n\nPostal questionnaires represent a cost-effective and convenient way of collecting participant-reported outcome data in health research, such as in randomised controlled trials (RCTs). However, attrition (i.e. when participants do not return the questionnaires) is a problem for many RCTs. The resultant loss of data leads to a reduction in statistical power and can lead to bias1. Although a number of strategies have been found to reduce attrition1,2 few of these have been evaluated in the context of healthcare RCTs. A recent systematic review highlighted the need for further research into methods of retaining participants in RCTs3.\n\nA Cochrane systematic review4 evaluating 110 different strategies to improve response rates to postal questionnaires in randomised controlled trials identified pre-notification as an effective strategy. The odds of response were increased by nearly half when participants were pre-notified of the impending arrival of the questionnaire (odds ratio (OR), 1.45; 95% CI, 1.29 to 1.63). Although there have been several studies evaluating different methods of pre-notification (such as letters, postcards or telephone calls) very few of these have been conducted in a healthcare setting. Only one RCT has evaluated the effectiveness of a pre-notification newsletter to increase response rates5. This study found a statistically significant increase in response rates (OR, 1.45; 95% CI, 1.01 to 2.10) among participants allocated to receive the pre-notification newsletter.\n\nThe Cochrane review4 also reported that the appearance of the questionnaire (e.g., making questionnaire materials more personal by using handwritten signatures) can affect response rates. For example, the odds of response increased by a quarter when addresses were handwritten compared to using computer-printed labels (OR, 1.25; 95% CI, 1.08 to 1.45). We are also aware of six studies that evaluated the effectiveness of attaching a Post-it® note to questionnaires to increase response rate6,7,8; four of these were undertaken within an academic setting and reported a statistically significant increase (p<0.05) in responses rates when personalised Post-it® notes were used3,6.\n\nAt the York Trials Unit we have a programme of undertaking studies within a trial (SWATs)9 that aim to evaluate simple interventions to increase response rates to postal questionnaires. Methods of pre-notification and Post-it® notes are relatively inexpensive, so even a small benefit is likely to be cost-effective. A single embedded trial will often not have the statistical power to detect a modest difference if there truly was one present; therefore, we have a strategy of repeating our SWATs in order to conduct meta-analyses to strengthen the evidence base. With respect to pre-notification, our previous trial showed a small absolute difference in favour of the intervention, which was borderline statistically significant (p=0.05)5, whereas our two previous studies of Post-it® notes7,8 produced identical, non-statistically significant ORs (0.97) favouring the control group (no Post-it® note).\n\nWe conducted a SWAT to evaluate the effectiveness of a pre-notification newsletter and/or applying a handwritten or printed Post-it® note as a means of increasing response rates to the 12-month follow-up questionnaire sent to participants in the REFORM trial. This paper presents the results of this sub-study. We also present the results of a meta-analysis of the three ‘Post-it® notes’ and two ‘pre-notification using a newsletter’ studies to increase questionnaire response rates in RCTs of health treatments.\n\n\nMethods\n\nThis trial was embedded within the National Institute for Health Research Health Technology Assessment (NIHR HTA) programme funded REFORM (REducing Falls with ORthoses and a Multifaceted podiatry intervention) study (registration number ISRCTN68240461; registration date, 1st July 2011; http://www.isrctn.com/ISRCTN68240461)10, which aimed to evaluate the clinical and cost effectiveness of a podiatry intervention for the prevention of falls in older people. Ethical approval for the REFORM study and this embedded sub-study was given by National Research Ethics Service East of England – Cambridge East Research Ethics Committee (REC reference 11/EE/0379) and the University of York, Department of Health Sciences Research Governance Committee.\n\nParticipants in the REFORM study who were due to be sent their 12-month follow-up questionnaire were included in this nested RCT. Participants who had asked to be withdrawn from the REFORM study or who did not wish to receive a questionnaire at this time point were excluded. Supplementary File 1 contains the full trial protocol of the REFORM study.\n\nWe undertook a three-by-two SWAT. Participants were allocated to one of six arms using block randomisation with a block size of 18, stratified by REFORM treatment group allocation. An independent data manager who was not involved in the recruitment of participants generated the allocation sequence by computer and allocated participants in a 1:1:1:1:1:1 ratio.\n\nParticipants were assigned to one of the following six groups: pre-notification newsletter plus handwritten Post-it® note applied to the questionnaire; newsletter plus printed Post-it®; newsletter only; handwritten Post-it® note only; printed Post-it® note only; or neither newsletter nor Post-it® note. The newsletter contained information regarding trial progress, including the geographical location and number of participants recruited, anonymised quotes from participants about what they thought of the study, and a reminder about the importance of the trial and of completing and returning postal questionnaires. The newsletter was tailored to the main trial treatment groups, with the newsletter sent to the intervention group addressing issues raised by participants about undertaking exercises and wearing orthotics. The newsletter was posted to participants 3 weeks prior to posting the 12-month questionnaire. Those participants randomised to not receive the pre-notification newsletter were sent the newsletter eight weeks after the questionnaire was sent out. The wording on the Post-it® note was “Please take a few minutes to complete this for us. Thank you! Sarah”. In order to minimise the possibility of heterogeneity, the wording (except for the name), text size and font on the Post-it® note was the same as that used for the studies by Tilbrook et al.7 and Lewis et al.8 and the Post-it® note was placed in the same location, on the top right hand corner of the questionnaire. Two researchers and three trial secretaries wrote the text of the handwritten Post-it® notes and every effort was made to ensure the format of the message was consistent. Participants also received an unconditional £5 note with their final follow up.\n\nThe date participants were sent and returned their postal questionnaires was recorded. Participants who did not return their follow-up questionnaire within 2 weeks were sent up to two postal reminders, 2 weeks apart by post, text or email, according to the participant’s preference, followed by a telephone reminder 1 week later.\n\nThe primary outcome was questionnaire response rate defined as the proportion of participants that returned their 12-month postal follow-up questionnaire to York Trials Unit.\n\nThe secondary outcomes were: time to response, defined as number of days between the questionnaire being mailed out to a participant and the questionnaire being recorded as returned to York Trials Unit; and the proportion of participants that needed a reminder.\n\nAll statistical analyses were conducted in Stata version 14 (StataCorp. 2015. Stata Statistical Software: Release 14. College Station, TX: StataCorp LP) using two-sided tests at the 5% significance level on an intention-to-treat basis. Age at randomisation into the main REFORM trial, gender and main trial allocation are summarised by randomised sub-study group. This factorial trial is reported as recommended by Montgomery et al.11 Response rates were calculated for each intervention. A logistic regression model containing the two interventions (Post-it® note and newsletter), age, gender and REFORM treatment allocation was performed. Adjusted ORs and corresponding 95% CIs were obtained from this model. The presence of an interaction between the two interventions was also tested by introducing the interaction term of the intervention into the logistic model.\n\nTime to return the 12-month follow-up questionnaire was calculated as the number of days from the date the questionnaire was sent out, to the date it was returned. Median time to return was calculated for all participants who returned their questionnaire. For the time-to-event analysis, questionnaires that were not returned or returned 6 weeks (42 days) or more after being sent were treated as censored. Time to questionnaire return was plotted for both interventions using Kaplan-Meier survival curves, and the log-rank test was used to compare the randomised groups within each intervention. A Cox proportional hazards regression model containing the two interventions, age, gender and REFORM treatment allocation was performed; adjusted hazard ratios (HR) and corresponding 95% CIs were obtained. The proportion of participants requiring a reminder was analysed using a similarly adjusted logistic model.\n\nAn aggregated fixed effect meta-analysis of this study with the study reported by Mitchell et al.5 evaluated the effect of sending a newsletter before receiving the questionnaire to improve response rates. A second aggregated data meta-analysis was conducted incorporating the results of this study and those by Tilbrook et al.7 and Lewis et al.8 in order to evaluate the effect of receiving a questionnaire with an attached Post-it® note on response rates.\n\nSupplementary File 2 contains a completed CONSORT checklist for this study.\n\n\nResults\n\nA total of 1010 participants were recruited into the REFORM study and randomised to receive a multifaceted podiatry intervention or usual care. In total, 917 (90.8%) reached the 12-month time point and were sent a follow-up questionnaire, of which 826 (90.1%) were randomised into the nested RCT (due to a delay in the start of the sub-study): 135 to receive the newsletter and the handwritten Post-it® note; 138 to receive the newsletter and the printed Post-it® note; 137 to receive the newsletter only; 137 to receive the handwritten Post-it® note only; 136 to receive the printed Post-it® note only; and 143 to receive neither the newsletter nor the Post-it® note (Figure 1). Participants had a mean age of 78 years (range, 65 to 96 years), and were predominantly female (n=509, 61.6%). Age and main trial allocation were balanced between the six groups, whereas a small chance imbalance for gender can be seen in the groups receiving the newsletter compared to those not receiving it: the presence of women tended to be higher in the groups not receiving the newsletter (65.6% vs 57.7%) (Table 1).\n\nThe total number of participants returning the 12-month follow-up questionnaire was 803 of 826 (97.2%), 390 of 410 (95.1%) of those who received the newsletter, and 413 of 416 (99.3%) of those who did not receive it. The difference in response rates between these two groups was statistically significant (adjusted OR, 0.14; 95% CI, 0.04 to 0.48; p<0.01) (Table 2). With respect to the Post-it® note intervention, 272 of 280 (97.1%) participants who received no Post-it® note, 267 of 274 (97.5%) participants who received the printed Post-it® note, and 264 of 272 (97.1%) who received the handwritten Post-it® note returned their questionnaire. The Post-it® note intervention did not show a statistically significant effect on the response rate (printed Post-it® vs no Post-it®: adjusted OR, 1.06; 95% CI, 0.37 to 3.01; p=0.92; handwritten Post-it® vs no Post-it®: adjusted OR, 0.91; 95% CI, 0.33 to 2.49; p=0.85). There was no statistically significant interaction between the interventions.\n\n1Logistic regression; 2Cox regression. All models contained both the newsletter and Post-it® note intervention terms and were adjusted for age, gender and main trial allocation. SE, standard error; OR, odds ration; HR, hazard ratio.\n\nTime to return ranged from 3 to 101 days. Among the participants who responded, the median time taken to return the 12-month questionnaire was 11 days, both overall and in each intervention group (i.e. no newsletter sent, newsletter sent, no Post-it® note, printed Post-it® note, and handwritten Post-it® note). In total, 793 (96.0%) participants returned the questionnaire within 6 weeks (no newsletter: n=407, 97.8%; newsletter: n=386, 94.2%; no Post-it® note: n=271, 96.8%; printed Post-it® note: n=263, 96.0%; and handwritten Post-it® note: n=259, 95.2%). There was evidence of a difference in time to return between those who received the newsletter and those who did not (adjusted HR, 0.86; 95% CI, 0.75 to 0.99; p=0.04) (Figure 2; Table 2). The Post-it® note intervention did not appear to have any effect on time to return (printed Post-it® vs no Post-it®: adjusted HR, 0.95; 95% CI, 0.80 to 1.13; p=0.55; handwritten Post-it® vs no Post-it®: adjusted HR, 0.90; 95% CI, 0.76 to 1.07; p=0.22) (Figure 3; Table 2). There was no statistically significant interaction between the interventions.\n\nOverall 125 (15.1%) participants required a reminder following 2 weeks of questionnaire non-response (newsletter: n=69, 16.8%; no newsletter: n=56, 13.5%; no Post-it® note: n=36, 12.9%; printed Post-it® note: n=41, 15.0%; handwritten Post-it® note: n=48, 17.7%). There was no evidence of a difference in the proportion of participants requiring a reminder between the groups (newsletter vs no newsletter: adjusted OR, 1.30; 95% CI, 0.88 to 1.91; p=0.19; printed Post-it® vs no Post-it®: adjusted OR, 1.20; 95% CI, 0.74 to 1.94; p=0.47; handwritten Post-it® vs no Post-it®: adjusted OR, 1.47; 95% CI, 0.92 to 2.36; p=0.11) (Table 2).\n\nWe combined the two previous Post-it® note studies with the study described in this paper. Because there was no material difference in response rates between the printed and handwritten Post-it® note (i.e., 97.5% vs 97.1%) in this study we combined these two groups in the meta-analysis (Post-it® note vs no Post-it® note: adjusted OR, 0.98; 95% CI, 0.40 to 2.37). The pooled OR was 0.97 (favouring no Post-it® note) but was not statistically significant (95% CI, 0.70 to 1.35; p=0.87) (Figure 4). No heterogeneity was observed (I2=0%). For the prior notification by newsletter, the meta-analysis (Figure 5) showed significant heterogeneity (I2=92%) with a non-statistically significant effect estimate favouring the intervention (pooled OR, 1.19; 95% CI, 0.84 to 1.70; p=0.33).\n\n\nDiscussion\n\nWe undertook a three-by-two randomised SWAT of pre-notification using a study newsletter and of attaching Post-it® notes (printed or handwritten) to postal questionnaires to improve response rates. The trial was embedded at the final (12-month) follow-up time point of the NIHR HTA-funded REFORM RCT. There was evidence that sending a study newsletter 3 weeks prior to the 12-month questionnaire had a detrimental effect on the response rate (adjusted OR, 0.14; 95% CI, 0.04 to 0.48; p<0.01) and time to return the questionnaire (adjusted HR, 0.86; 95% CI, 0.75 to 0.99; p=0.04); however, the raw difference in response rates was small (95.1% vs 99.3%). A small imbalance in gender among the six groups was observed at randomisation, but gender was adjusted for in all analyses. A previous SWAT of a pre-notification newsletter5, conducted in an older female population, showed a positive finding, which was in line with the Cochrane review4 of pre-notification approaches to enhance survey returns. A meta-analysis combining that trial with ours produced a small, non-statistically significant effect favouring pre-notification; however, there was significant heterogeneity in the results.\n\nResponse rates across the groups receiving a printed Post-it® note on their questionnaire, a handwritten Post-it® note and no Post-it® note were all very similar (97.5, 97.1 and 97.1%, respectively). There was no statistically significant difference between the groups in terms of response rate, time to return the questionnaire, and requiring a reminder. This lack of effect on response rates has now been demonstrated across three separate trials. The first trial was among patients with neck pain (mean age, 53 years)7, the second trial was among older patients (mean age, 74 years) at risk of depression8 with the current trial among a similar age group (mean age, 76 years), but no risk/diagnosis of depression. The consistent results suggest that it is not worthwhile undertaking further trials of this intervention among a middle-aged or older population. There may be merit, however, in testing this intervention in a younger population where response rates may be lower.\n\nNo statistically significant differences were observed in the proportion of participants requiring a reminder between the groups.\n\nResponse rates in the six groups all exceeded 94%, making significant improvement difficult. These simple interventions were relatively inexpensive but not cost-free due to the price of printing the newsletters and the printed Post-it® notes, and staff time to handwrite the Post-it® notes. A cost-effectiveness analysis was not performed since a benefit was not observed.\n\n\nConclusions\n\nIn summary, in this reasonably sized trial of 826 participants, we found no evidence of a benefit of handwritten or printed Post-it® notes on questionnaire response rates. We also found a negative effect of a pre-notification newsletter; however, a meta-analysis suggests the evidence is still uncertain.\n\n\nData availability\n\nDataset 1. Raw data concerning patient demographics, type of reminder received and the returning of the questionnaire. DOI: 10.5256/f1000research.14591.d20291012",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe REFORM study was funded by the National Institute for Health Research (NIHR) Health Technology Assessment (HTA) Programme (Programme grant number 09/77/01). This SWAT was funded by York Trials Unit.\n\n\nAcknowledgements\n\nThis article has been written by the authors on behalf of the REFORM Study Team.\n\n\nSupplementary material\n\nSupplementary File 1. Complete trial protocol for the REFORM study.\n\nClick here to access the data.\n\nSupplementary File 2. Completed CONSORT checklist.\n\nClick here to access the data.\n\n\nReferences\n\nMcColl E, Jacoby A, Thomas L, et al.: Design and use of questionnaires: a review of best practice applicable to surveys of health service staff and patients. Health Technol Assess. 2001; 5(31): 1–256. PubMed Abstract | Publisher Full Text\n\nHewitt CE, Kumaravel B, Dumville JC, et al.: Assessing the impact of attrition in randomized controlled trials. J Clin Epidemiol. 2010; 63(11): 1264–70. PubMed Abstract | Publisher Full Text\n\nBrueton VC, Tierney J, Stenning S, et al.: Strategies to improve retention in randomised trials. Cochrane Database Syst Rev. 2013; (12): MR000032, 1–126. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdwards PJ, Roberts I, Clarke MJ, et al.: Methods to increase response to postal and electronic questionnaires. Cochrane Database Syst Rev. 2009; (3): MR000008. PubMed Abstract | Publisher Full Text\n\nMitchell N, Hewitt CE, Lenaghan E, et al.: Prior notification of trial participants by newsletter increased response rates: a randomized controlled trial. J Clin Epidemiol. 2012; 65(12): 1348–52. PubMed Abstract | Publisher Full Text\n\nGarner R: Post-it® Note Persuasion: A sticky influence. J Consum Psychol. 2005; 15(3): 230–37. Publisher Full Text\n\nTilbrook HE, Becque T, Buckley H, et al.: Randomized trial within a trial of yellow ‘post-it notes’ did not improve questionnaire response rates among participants in a trial of treatments for neck pain. J Eval Clin Pract. 2015; 21(2): 202–4. PubMed Abstract | Publisher Full Text\n\nLewis H, Keding A, Bosanquet K, et al.: An randomized controlled trial of Post-it® notes did not increase postal response rates in older depressed participants. J Eval Clin Pract. 2017; 23(1): 102–107. PubMed Abstract | Publisher Full Text\n\nAdamson J, Hewitt CE, Torgerson DJ: Producing better evidence on how to improve randomised controlled trials. BMJ. 2015; 351: h4923. PubMed Abstract | Publisher Full Text\n\nCockayne S, Rodgers S, Green L, et al.: Clinical effectiveness and cost-effectiveness of a multifaceted podiatry intervention for falls prevention in older people: a multicentre cohort randomised controlled trial (the REducing Falls with ORthoses and a Multifaceted podiatry intervention trial). Health Technol Assess. 2017; 21(24): 1–198. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMontgomery AA, Peters TJ, Little P: Design, analysis and presentation of factorial randomised controlled trials. BMC Med Res Methodol. 2003; 3: 26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRodgers S, Sbizzera I, Cockayne S, et al.: Dataset 1 in: A nested randomised controlled trial of a newsletter and Post-it® note did not increase postal questionnaire response rates in a falls prevention trial. F1000Research. 2018. Data Source"
}
|
[
{
"id": "36069",
"date": "03 Aug 2018",
"name": "Shaun P. Treweek",
"expertise": [
"Reviewer Expertise Trial 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 is a well-written article that was great to read. I particularly liked the recommendations in Discussion re: whether it was worth doing more SWAT evaluations. It would be good to see that sort of clarity more often.\nI only have a few comments, which are listed below.\nInterventions Would it be possible to see the newsletter, or at least the bit relevant to this study? We have the Post-It text but not the newsletters.\nManagement of postal questionnaires I’m guessing that the content of the reminders was the same for all participants regardless of which arm they were allocated to but could you confirm this?\nTable 1 The gender imbalance does look odd to me. I know that you say that it is due to chance and that is of course plausible but it differs by up to about 15% across the interventions for women and by up to about 15% for men. These percentages amount to 10 - 20 or so individuals in a total sample size of 135-143. I’m not sure that size of difference would come about just by chance although, of course, it could. Are you sure it’s just chance, or a feature of the randomisation/blocking/something else?\n\nTable 2 Could you consider giving absolute differences for the primary outcome as well as the OR? ORs are always a bit tricky to interpret.\n\nFigures 4, 5 and linked text Two points. I think it would be good to do two GRADE assessments of the evidence included in the two forest plots. This isn’t as hard as it sounds. Depending on the design quality of the included studies (and the current 2018 one is good) my guess is that if the two other studies in Fig 4 are good studies, GRADE is high and for Fig 5 it’s moderate because of inconsistency, though you might pull it down for imprecision too (1.19 with a CI of 0.84 - 1.70 seems pretty wide to me). Regardless, I do think it would be good to say something about the certainty of the body of evidence and then link that to your recommendations in the Discussion.\nThe second point is that I wasn’t sure why you used a fixed effect model for Fig 5 and a random for Fig 4. My guess is that the random effects model is the one to go for (I’d be surprised if the only differences between studies is random error but that interventions, patients, context etc are at play too). Worth thinking about anyway, especially whether the intention was for the two forest plots to use different models.\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": "4414",
"date": "19 Feb 2019",
"name": "Sara Rodgers",
"role": "Author Response",
"response": "Thank you for your comments, which we address in turn below. We have now included the study update newsletter in supplementary material. This is correct, the reminders were standardised. We have clarified this in the text. We noticed that there was a mistake in the following sentence: ‘...the presence of women tended to be higher in the groups not receiving the newsletter (65.6% vs 57.7%)’ – the word ‘not’ has been removed. We can think of no other explanation than chance for the imbalance in gender between the groups. These are now included in the manuscript. Thank you for your suggestion. We have completed the GRADE assessments and included them in the text. We have also included the assessment tables as Supplementary materials. The text and analyses have been amended as appropriate so that both are a random effects meta-analysis, but we have introduced the fixed effects meta-analysis as a sensitivity analysis for the Post-it® notes given the lack of heterogeneity."
}
]
},
{
"id": "40423",
"date": "10 Dec 2018",
"name": "Phil J. Edwards",
"expertise": [
"Reviewer Expertise Systematic reviews of methods to increase response to postal and electronic questionnaires"
],
"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\nComments on the manuscript: Thank you for inviting us to review this manuscript. The authors should be commended for nesting this factorial study within the REFORM study, to provide further evidence on methods to increase response to postal questionnaires. It is vital that researchers conduct this type of nested trial for evidence-based methods research to progress in a cost-effective manner.\n\nOverall, we found this to be a well-designed, and well-conducted study. We have a few comments regarding the reporting of the results.\n\nTitle: We don’t think this is currently the best wording for the title - the “nested RCT” was not the intervention.\n\nABSTRACT Conclusions:\nThe authors say that the evidence for “newsletter reminders” is still uncertain, but we think that the authors meant to say “prenotification newsletters”.\n\nMETHODS\n\nInterventions:\n\nThe authors could make it clearer to what extent the newsletter was a prenotification intervention; for example, was there a letter with the newsletter explaining that the questionnaire was imminent? Or did text within the Newsletter explain that a questionnaire was imminent? It is currently unclear the extent to which the newsletter warned of the imminent questionnaire, and whether it tried to encourage participants to complete it and return it. Perhaps the Newsletter might be included in the Supplementary material?\n\nRESULTS Meta-analyses:\n\nIn the meta-analysis of the Post-it note interventions (figure 4), there is no evidence of heterogeneity among the studies (I-squared=0%) so a fixed effect model, rather than a random effect model, is appropriate. We expect that the 95% confidence interval will be narrower around the estimated odds ratio of 0.97, consistent with the conclusion that the study found no evidence of a benefit of the Post-it notes on increasing response.\n\nIn the meta-analysis of prenotification by newsletter interventions (figure 5), there is substantial heterogeneity among the studies (I-squared=92%), so in this case, a random effect model, rather than a fixed effect model is appropriate. We expect that the 95% confidence interval will be wider around the estimated odds ratio of 1.19, consistent with the conclusion that the magnitude of the effect on response of prenotification by newsletter remains uncertain, but that a moderate effect (e.g. OR=1.5) is still plausible.\n\nCONCLUSIONS We were not persuaded that a study with 826 participants is necessarily “a reasonably sized trial”. A study powered to detect an intervention effect equivalent to an odds ratio of OR=1.5 from a baseline response proportion of 97% would require 3,826 participants in each arm (80% power), or 5,121 participants in each arm (90% power). However, a statistically significant reduction in response with the prenotification newsletter was observed with the study sample of 826 participants, and so this may stand as the study result without any need for the authors to comment on the size of the sample.\n\nIn our most recent update to the Cochrane Review (cited in the manuscript), Forty-seven trials (79,651 participants) evaluated the effect on response of contacting participants before sending questionnaires. The odds of response were increased by a half when participants were pre-notified (OR 1.45; 95% CI 1.29 to 1.63). However, there was significant heterogeneity among the trial results (P < 0.00001). We have recently updated this meta-analysis, for an MSc Epidemiology dissertation (Woolf, B. 2018, unpublished data). In this update, 103 trials were included. Overall, pre-notification increased response 1.38 (95%CI: 1.27-1.49) (pooled result from a random effect model). However, when studies at high or unclear risk of bias were excluded the 95% confidence interval, for the pooled effect of the remaining eight studies, crossed the null. The meta-analysis also found several factors which explained some of the heterogeneity (e.g., the method of pre-notification, using a different method of delivering the pre-notification than the questionnaire, and the risk of bias of the included studies). However, heterogeneity was still present after accounting for these factors.\n\nGiven that the method of pre-notification appears to explain study differences, and that this study is only the second to explore the use of Newsletters as a type of pre-notification, it provides important evidence for further understanding this method for potentially reducing questionnaire non-response.\n\nIs the work clearly and accurately presented and does it cite the current literature? Yes, however, there are some minor modifications which would make the paper easier to read. For example, the presentation of the experimental conditions, although presented accurately, was not intuitively easy to grasp. We personally find it easier to understand factorial randomisation (especially when more complex than 2x2) if a matrix or decision tree is provided, for example, the one shown below.\n\nCondition\n\n________|_________\n\n|\n\n|\n\nNewsletter\n\nno Newsletter\n\n_________|________\n\n__|_________\n\n|\n\n|\n\n|\n\n|\n\n|\n\n| hand\n\nprinted\n\nno\n\nhand\n\nprinted\n\nno\n\nwritten postit\n\nPostit written Postit\n\nPostit\n\nPostit\n\nPostit\n\nIs the study design appropriate and is the work technically sound? Yes\nAre sufficient details of methods and analysis provided to allow replication by others? Yes. However, it would be useful if the length of time period for which a survey response would be included in the study was stated explicitly. Prima facie, varying the amount of time participants have for their response to be included in the study could change the results of any potential replication.\nIf applicable, is the statistical analysis and its interpretation appropriate? No, we believe that a random effect model, rather than a fixed effect model is appropriate in the meta-analysis of pre-notification by newsletter interventions. In addition, we would prefer to see more detail about the adjustments which were made for the main analysis. The authors present an odds ratio adjusted for gender. The adjustment was made because of a baseline imbalance after randomisation. However, it is unclear if the decision to make this adjustment was post hoc or part of a decision procedure in a pre-specified protocol. With this in mind, it would be useful if the authors presented 95% confidence intervals for the crude proportions of responses in each conditions to aid integration of these results, as well as the crude odds ratio.\nAre all the source data underlying the results available to ensure full reproducibility? Yes. However, we believe that the study authors should be more explicit about the source and selection methods of studies included in the meta-analyses.\n\nAre the conclusions drawn adequately supported by the results? See our comments above\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? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "4413",
"date": "19 Feb 2019",
"name": "Sara Rodgers",
"role": "Author Response",
"response": "Points addressed in turn: Thank you for your comments. We have addressed each comment in turn below. Additionally, the paper has been revised with minor amendments to improve clarity and general readability. Thank you, we have suggested the alternative title of “A study update newsletter or Post-it® note did not increase postal questionnaire response rates in a falls prevention trial: an embedded randomised factorial trial” Apologies, we did not intend to use the word “pre-notification” to suggest that specific text was included to remind participants of the impending arrival of their 12-month questionnaire. More accurately, we sent a ‘study update’ newsletter to participants, which we have now included in supplementary material. We have removed the use of the term “pre-notification” where necessary and clarified throughout the paper. We have now clarified the content of the newsletter and have added this as supplementary material. Thank you, the text and analyses have been amended as appropriate so that both are a random effects meta-analysis. We have also introduced the fixed effects meta-analysis as a sensitivity analysis given the lack of heterogeneity in the Post-it analysis. Updated forest plot for newsletter random effect meta-analysis (Figure 5). Thank you, we acknowledge the reviewer’s point and have removed the text: “in this reasonably sized trial of 826 participants,”. We have revised the CONSORT diagram according to the reviewer’s suggestion. All survey responses were included regardless of when they were returned. Text has been added to the manuscript to clarify this. The text and analyses have been amended as appropriate so that both are a random effects meta-analysis, and the forest plot for the newsletter meta-analysis (Figure 5) has been updated. The decision to adjust for gender was made prior to analysis and was not based on the chance imbalance observed in gender. The unadjusted odds ratios have been added to the manuscript, and the 95% CI for the proportions have been added to the CONSORT diagram. We have clarified that we meta-analysed the similar studies undertaken at York Trials Unit previously."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1083
|
https://f1000research.com/articles/8-200/v1
|
19 Feb 19
|
{
"type": "Systematic Review",
"title": "Health care seeking for maternal and newborn illnesses in low- and middle-income countries: a systematic review of observational and qualitative studies",
"authors": [
"Zohra S. Lassi",
"Philippa Middleton",
"Zulfiqar A. Bhutta",
"Caroline Crowther",
"Philippa Middleton",
"Zulfiqar A. Bhutta",
"Caroline Crowther"
],
"abstract": "Background: In low- and middle-income countries, a large number of maternal and newborn deaths occur due to delays in health care seeking. These delays occur at three levels i.e. delay in making decision to seek care, delay in access to care, and delay in receiving care. Factors that cause delays are therefore need to be understand to prevent and avoid these delays to improve health and survival of mothers and babies.\n\nMethods: A systematic review of observational and qualitative studies to identify factors and barriers associated with delays in health care seeking. Results: A total of 159 observational and qualitative studies met the inclusion criteria. The review of observational and qualitative studies identified social, cultural and health services factors that contribute to delays in health care seeking, and influence decisions to seek care. Timely recognition of danger signs, availability of finances to arrange for transport and affordability of health care cost, and accessibility to a health facility were some of these factors. Conclusions: Effective dealing of factors that contribute to delays in health care seeking would lead to significant improvements in mortality, morbidity and care seeking outcomes, particularly in countries that share a major brunt of maternal and newborn morbidity and mortality. Registration: PROSPERO CRD42012003236.",
"keywords": [
"Health care seeking",
"maternal health",
"neonatal health",
"developing countries",
"low- and middle-income countries"
],
"content": "Introduction\n\nThe majority of low- and middle-income countries (LMICs) have been unable to achieve the targets set for Millennium Development Goals (MDG) 4 and 51,2. Even with improvements in maternal and child mortality rates over past decades, 303,000 mothers and 5.9 million children under the age of 5 years died in 20153,4, with 99% of these deaths occurring in LMICs. LMICs lack financial and human resources and basic utilities including clean water, sanitation and education are not always readily available. Families in LMICs often unable to access and afford health care when required, and therefore, care seeking from non-skilled birth attendants is preferred when women give birth.\n\nRates of birthing at home are higher in LMICs, and usually skilled birth attendants (SBA) are not present5. In Sub-Saharan Africa, 50% of births occur at home with no skilled birth attendant; in South Asia, mothers and their families are the primary care givers of a third of all home births. In these regions, the inequalities are even higher among poorer people, particularly those living in very remote geographical areas6. While interventions to reduce poverty may require more time, training and deploying skilled birth attendants and upgrading emergency obstetric care are urgently needed7. Evidence suggests an association of skilled birth attendance with reduced neonatal mortality—77% of neonatal mortalities occur where coverage of skilled birth attendance is 50% or lower8.\n\nWhile a systematic review that assessed the determinants of skilled attendance or health facility use for delivery in LMICs has been performed9, there was no attempt to identify the barriers and facilitators of health care seeking for maternal and newborn illnesses in LMICs. Another systematic review on effectiveness trials that has also identified strategies that can improve maternal and newborn health care seeking10; however, a review of narrative and qualitative studies is required to identify barriers and enablers of health care seeking in LMICs. We aimed to systematically review observational and qualitative studies to identify factors associated with delays that lead to serious maternal and neonatal morbidity and mortality11. These delays occur at three levels: 1) identification and decision making to seek care; 2) arranging means to reach a health facility; 3) receiving adequate care at the health facility.\n\n\nMethods\n\nAll observational and qualitative studies from LMICs that assessed health care seeking behaviour or pattern for maternal and newborn health care and illnesses were included. We define health care seeking as ‘sequence of remedial actions that individuals undertake to rectify perceived ill-health’. The primary aim was to identify the barriers and enablers of maternal and newborn health care seeking and related pathways in LMICs. The protocol for this systematic review and meta-analysis has been registered with PROSPERO 2012: CRD42012003236.\n\nThe search engines PubMed, Medline, EMBASE, the Cochrane Library, and Google Scholar were initially searched up to Sep 16, 2016 and then searches were revised on September 27, 2017, but we found that data was saturated, and no new themes were emerged. Search terms were a combination of [(‘care seeking’ OR ‘care-seeking’ OR ‘health care’ OR ‘health care seeking’ OR ‘community based intervention*’ OR ‘community-based intervention*’) AND (mother* OR maternal OR women OR newborn* OR neonat*)] used as medical subject headings and keyword terms in the title/abstract. No language restrictions were applied. Grey literature and reference lists of included studies were also searched to identify studies. We considered studies from LMICs that assessed the factors associated with health care seeking for maternal and newborn illnesses in observational or qualitative studies. We did not consider studies on health care seeking for specific maternal and newborn illnesses such as jaundice etc. or for preterm babies. We considered recommendations for systematic reviewing of qualitative studies12. We used the PRISMA checklist PRISMA statement in reporting systematic reviews from the observational studies13. The 22-items STROBE checklist was used to assess the methodological quality of the cross-sectional studies14. Studies that fulfilled the methodological criteria of more than 18 points were classified as high quality, between 12–18 as moderate quality and below 12 were classified as low quality.\n\nZSL and PM independently reviewed the retrieved articles in two stages. First, relevance was assessed from the title and abstract and if relevance was still unclear, the full text was read. Any disagreement was referred to a third reviewer (CC or ZAB). Factors responsible for health care seeking patterns for maternal and newborn health from observational studies and qualitative studies were separately analysed. Study design, country of study, setting, participants, and results were recorded for each study. We performed a narrative synthesis of the findings from the included studies, as included studies were observational and qualitative in nature.\n\n\nResults\n\nOur search strategy identified 20,944 articles, of which 232 met the inclusion criteria (Figure 1). We found and analysed 159 original studies (162 published papers), of which 115 were observational studies and 44 were qualitative studies (characteristics of included studies are included as extended data15. Observational studies were moderate to high in quality upon quality assessment.\n\n\nQualitative findings for delays in care and pathways of care seeking\n\nHealth-care-seeking patterns are complex phenomena, often confounded by several interlinked factors such as education of mothers, socio-economic status and age. More than half of the included observational studies reported that poor maternal health care utilization and giving birth at home is associated with lack of antenatal care, age, parity, education and employment status of women13,16–74. On the other hand, seeking care for newborn illnesses depends on the severity of illness75–80 and gender of the baby, with preference being given to male children48. Studies have reported that adequate health care seeking from skilled health care providers leads to fewer deaths and morbidities81–83. Women who had good marital relationships with husbands were more likely to report receiving antenatal care and institutional birth55,56,84–90. Similarly, women who had good relationships with their mothers-in-law reported being able to attend or receive antenatal care, with the degree of bonding and communication of women with their mothers-in-law reported to be an important influencing factor91. We identified several qualitative themes in the section below that describe the reasons for delays in health care seeking and associated pathways.\n\nPrimary caregivers in all included studies were usually mothers; however, mothers-in-law, grandmothers, fathers, neighbours, traditional healers and opinion leaders in the community were among the many people involved as caregivers for mothers and babies. Across the studies, it was observed that mothers/families do seek care for neonatal illness80; however, complications during pregnancy are not considered as an illness and many signs are considered as normal, even when painful and constant92–102. In certain studies, bleeding was not considered to be a complication103, and in such situations decisions to seek health care were often delayed. Women who expressed pain verbally were considered as disobedient and therefore maintaining silence was considered appropriate104,105. Missing antenatal care visits were reported to be due to heavy and unavoidable workloads at home91, and a few studies reported that mothers-in-law privileged household chores over women’s health91,99. Some families perceived that some common neonatal symptoms should or cannot be treated at health facilities and therefore traditional care should be sought106.\n\nIn India, women during pregnancy are usually advised to be cautious while eating “hot” or “cold” food, and to eat less otherwise the baby can grow too large and therefore lead to a difficult birth92. A qualitative study from Pakistan (Baluchistan)107 described that the dai (traditional birth attendant (TBA)) usually places mustard oil on her fingers and massages the vaginal walls to ease the birth, and inserts vaginal and anal pessaries after birth to help shrink the uterus and to provide support for the uterus and backbone. They also prefer women to eat Goandh (edible gum) combined with turmeric powder, and dried dates in milk to induce heavy vaginal bleeding so that all unclean blood is drained from the body, thus predisposing to postpartum haemorrhage. In situations when the placenta does not expel normally, the dai enters her bare hands in the uterus or puts hair into the mother’s mouth to induce vomiting103. Eating vegetables rather than meat during pregnancy is preferred as it is considered to increase the production of breastmilk and freshens its taste108. During infant illnesses, mothers prefer to give ‘rabadi’ (prepared by cooking millet flour and yogurt), ‘khichchadi’ (a semi-liquid rice and pulses mixture) and ‘mateera’ (watermelon curry) to their febrile children in conjunction with breastmilk109.\n\nWhile illnesses, particularly in women who are not pregnant, are considered unimportant, evil spirits and fate (Allah’s will) are reported to be the cause of these illnesses103,110. Faith healing is important in many cultures. A study from Ghana111 named three major religions that practised faith healing and each religion has a specific healer. On the other hand, most of the communities in Asia and Africa believe that certain precautions during pregnancy or immediately after birth will ward off the evil eye (a gaze or stare superstitiously believed to cause harm) and will prevent the infant from getting sick92,112. This includes isolating women and their baby in a room for a certain period of time after childbirth and lighting a fire at the entrance where they are confined92. In order to prevent them from evil eyes, people reported keeping the pregnancy secret from people outside of close relations96.\n\nMothers may consult family and friends when the danger signs are not clear or unusually severe107,113. However, in severe illness, decision-making power can be switched to more experienced members of the extended family, which can cause significant delays in decision-making. Many of the studies reported that in scenarios when women had money, they hurried to pursue treatment options from a health care facility despite several familial pressures. A study from Tanzania reported95 that having an option of home birth was found to be a hurdle in emphasizing the importance of skilled birth care100,104,114–116. Trust for someone from the same community, sharing the same values and speaking the same language, was another factor that encouraged women to give birth at home and with a TBA38,117. However, it was apparent from the studies that if women continued to suffer, then they do seek care from western-trained care providers107.\n\nDecision-making emerged as a complex issue. Decision-making power is less likely to be with the woman and mostly rests with their partners and mothers-in-law. Women who had no income source were usually those who had no rights for decision making118. Several studies reported that the major barrier for institutional care was gaining permission from husbands92,93,104,119–124. Women are considered inferior to men and their disobedience often results in physical and emotional violence123. If husbands are absent, women face difficulties in receiving permission from her husband’s parents or other elders for seeking care and this results in even greater delays. Husbands and elders often have control over finances and women are mostly dependant on them38,93,99,108–110,117,121,122,125–131. Deciding to seek care can incur transportation costs, user fees, cost of medicines, and possibly ensuing costs of misdiagnosis and treatment failures93. Considering all these barriers, women often postpone seeking help, with the hope that the problem will subside on its own.\n\nWhen a family is willing to seek care and arranges the money required, other challenges such as physical transference of mothers and newborns to health facilities becomes a problem. The situation is even worse if complications arise at night, when risk of being attacked by criminals’ increases or when transport providers raise their taxi or car-hire charges96,118,121,122,125,127,130–143. Studies on people living in very remote areas reported factors such as distance to health facility and related transportation issues, lack of financial resources, encountering swollen rivers on the way, fear of encountering wild animals, shame about too many pregnancies or being of advanced age and pregnant as some of the critical reasons for not seeking care. Studies also reported other factors responsible for not seeking health care such as non-availability of staff at facility, rude behaviour of health care staff, and poor quality of care96. Fear of operative procedures was reported as a factor hindering care-seeking144. These were usually based on previous experience and contact with health care staff and the health care service received93,122,125,145,146.\n\nCost is another important barrier to seeking care from trained health professional. However a study from rural Mexico reported that cost of care from TBAs is sometimes higher than facility birth but women prefer them because they can give birth at home147. Many women also preferred giving birth at home because they preferred a squatting position for giving birth that was also endorsed by TBAs147. Relatives being not allowed at facilities during the childbirth was another factor expressed for giving birth at home147.\n\nWomen’s previous encounters with health care staff and facilities were reported as a key factor for decision making148. Further, many of the danger signs are not considered as pregnancy-related complications38,80,109,113,117,118,149, and thus families seek help from traditional healers, community health workers or drug sellers. Households often regard accessible and less expensive care such as herbal and home remedies or locally available drugs more highly150. Workers from these types of care were often praised as they give time to patients and consider their social and cultural aspects as well.\n\nWomen and families usually opt for medically qualified birth attendants where women are perceived to have possible birth complications. Where TBAs detected a complication at home, women were provided with referrals. Women also preferred SBA when they wanted to have a tubal ligation performed147. Studies reported that perceived fear of being torn in hospital, where Caesarean section was required151 discourages women to seek institutional care for childbirth104. Lack of privacy at care facilities and being examined in the open are other factors for not seeking care at clinics152.\n\nPregnant women or mothers with ill newborns usually experienced long waiting times when seeking hospital care93,104,153,154. Most of the included studies cited that health professionals have poor attitudes towards poor or pregnant women, which are stigmatizing118,131,144,145,155. Studies pointed out that health care staff examine women in hurry and at many occasions did not clarify their concerns. Staff may stigmatize women, criticising them for their age and number of pregnancies and judge them on their practices for family planning152. Staff behaviour is therefore a major barrier for accessing care151.\n\nSometimes, women are referred to another facility due to lack of trained staff and functional equipment and supplies that lead to further delays. Women may be asked to pay for fuel for the ambulance to take them to the other facility. They then may be required to pay for medicines and other supplies, and when stocks of these run out, there are further delays in receiving care118.\n\n\nDiscussion\n\nIt is often suggested that overwhelming maternal and neonatal mortalities and morbidities are closely linked with a number of interrelated delays that prevent a pregnant women or neonate from accessing the health care needed11. Each delay is closely related to services, logistics, facilities and conditions. Our review identified factors associated with delays (Figure 2) and the pathways for health care seeking in cases of illnesses (Figure 3). Although the pathways of seeking care were not similar across all the studies, choices usually followed the same pattern if not the same levels. Depending on predisposing factors (be it God’s will, past experiences, user affordability, accessibility, availability or acceptability), the first choices for seeking care were for spiritual healers and immediate elder members of the family and community such as mothers-in-law and TBAs, who not only hold a respected position in the community but are generally considered as experienced and knowledgeable people. If not gaining any benefits from the care received from the first level, women then consult pharmacists, homeopaths and quack healers or untrained village doctors. However, the last choice (can be second or third) is usually the trained doctors, nurses or lady health visitors in health facilities.\n\nIneffective or inequitable decision making was reported as the biggest hurdle for seeking care during illnesses. Several cultural, economic, and health system related factors confound this further. Prompt identification of danger signs, autonomy of decision making, availability of finances, accessibility to health facility, and perceived quality of care play a major role in institutional health care seeking. Distance and cost were highlighted as the two main reasons for causing delays in decision making. Inadequately equipped facilities further delays care156. Improvement in medical care seeking can be achieved if behaviour change communication interventions are contextualized and meet specific needs of the community. Similar findings have been reported by an earlier review on determinants of skilled birth care and institutional births9.\n\nThis review highlights the reasons for delays and the ramifications of these delays on morbidity and mortality outcomes. Delays at each level serve as barriers, and strategies to overcome these may help and empower the communities to select and make early decisions. Cultural norms, societal values along with limited financial resources were underscored as major hindering factors for care seeking. It is therefore important that health system reforms related to maternal and newborn health should consider societal and cultural barriers and practices to improve their health care seeking. A major obstacle is women’s self-sufficiency and lack of empowerment to make decisions about their health. A change brought about in the attitude of the family members with emphasis on the need for women’s autonomy in making these crucial health decisions will have an immediate positive impact. Women should have the right to choose where they give birth, although it is important to help the woman comprehend the risks associated with these options. This could be achieved by proper mobilization of the entire family. At the same time, health systems should train health workers to provide and manage emergencies. A specific implementation strategy could be the provision of birthing kits to the TBAs which will ensure access to this facility to those residing in remote areas. This will reduce mortality arising from delay in the provision of emergency medical aid during childbirth. In addition, government should subsidize health care costs and should introduce schemes such as conditional cash transfers particularly for places where access to health care facility is an issue. These remedies have also been found to be cost-effective157.\n\nWhile we were able to extract the important factors associated with maternal and newborn health care seeking, the review also faced some methodological challenges. First, the findings from the observational studies need to be interpreted with caution as included studies employed different inclusion criteria. Second, the studies used different statistical modelling to control for confounders and clustering therefore made it hard to compare the results. Third, the findings, particularly from the qualitative studies, were from different geographical settings and the barriers faced in one community may not exist or differ in another community. Therefore, strategies to improve health care seeking need to be context- and community-specific. Earlier review of experimental studies suggested that simple strategies such as community mobilization and home visitation via community health workers may improve health care seeking and perinatal survival10. Our findings from the observational and qualitative studies have identified the important barriers of health care seeking that need to be considered while developing strategies.\n\n\nConclusion\n\nDespite all the progress made towards improving maternal and newborn health in past few decades, many LMICs could not reach the MDGs. This review has identified several socio-economic, cultural and health services related factors that contribute to delays in health care seeking. Effective implementation of strategies after controlling for these factors of delays such as increasing women’s autonomy would lead to significant improvement in mortality, morbidity and care seeking outcomes.\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\nOpen Science Framework: Health care seeking. https://doi.org/10.17605/OSF.IO/5UT6X15. Supplementary Files contain information concerning characteristics of the studies included in this review.\n\nOpen Science Framework: PRISMA 2009 Checklist for this study. https://doi.org/10.17605/OSF.IO/5UT6X15.\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\nThis review was part of a doctoral thesis that was funded by the University of Adelaide, Australia.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nWHO: Trends in maternal mortality: 1990 to 2010. In: Geneva, Switzerland: WHO, UNICEF, UNFPA and The World Bank estimates; 2012. Reference Source\n\nLozano R, Wang H, Foreman KJ, et al.: Progress towards Millennium Development Goals 4 and 5 on maternal and child mortality: an updated systematic analysis. Lancet. 2011; 378(9797): 1139–1165. PubMed Abstract | Publisher Full Text\n\nWorld Health Organization: Trends in Maternal Mortality: 1990 to 2015. In: Estimates by WHO, UNICEF, UNFPA, The World Bank and the United Nations Population Division. 2015. Reference Source\n\nLevels & Trends in Child Mortality: Report 2014. Estimates Developed by the UN Inter-agency Group for Child Mortality Estimation. In.: UN Child Fund, WHo, The World Bank, UNPD; 2015. Reference Source\n\nUNICEF: THE STATE OF THE WORLD’S CHILDREN 2008 In. New York; 2008. Reference Source\n\nDarmstadt GL, Lee AC, Cousens S, et al.: 60 Million non-facility births: who can deliver in community settings to reduce intrapartum-related deaths? Int J Gynaecol Obstet. 2009; 107 Suppl 1: S89–S112. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrata N, Passano P, Rowen T, et al.: Where there are (few) skilled birth attendants. J Health Popul Nutr. 2011; 29(2): 81–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLawn JE, Kerber K, Enweronu-Laryea C, et al.: 3.6 million neonatal deaths--what is progressing and what is not? Semin Perinatol. 2010; 34(6): 371–386. PubMed Abstract | Publisher Full Text\n\nGabrysch S, Campbell OM: Still too far to walk: literature review of the determinants of delivery service use. BMC Pregnancy Childbirth. 2009; 9: 34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLassi ZS, Middleton PF, Bhutta ZA, et al.: Strategies for improving health care seeking for maternal and newborn illnesses in low- and middle-income countries: a systematic review and meta-analysis. Glob Health Action. 2016; 9: 31408. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThaddeus S, Maine D: Too far to walk: maternal mortality in context. Soc Sci Med. 1994; 38(8): 1091–1110. PubMed Abstract | Publisher Full Text\n\nJones ML: Application of systematic review methods to qualitative research: practical issues. J Adv Nurs. 2004; 48(3): 271–278. 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; 6(7): e1000097. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSTROBE Statement—Checklist of items that should be included in reports of cross-sectional studies. Available from: STROBE Statement—Checklist of items that should be included in reports of cross-sectional studies. Reference Source\n\nLassi Z: Health care seeking. 2019. http://www.doi.org/10.17605/OSF.IO/5UT6X\n\nAnson O: Utilization of maternal care in rural HeBei Province, the People's Republic of China: individual and structural characteristics. Health Policy. 2004; 70(2): 197–206. PubMed Abstract | Publisher Full Text\n\nAlexandre PK, Saint-Jean G, Crandall L, et al.: Prenatal care utilization in rural areas and urban areas of Haiti. Rev Panam Salud Publica. 2005; 18(2): 84–92. PubMed Abstract | Publisher Full Text\n\nAdhikari R, Sawangdee Y: Influence of women's autonomy on infant mortality in Nepal. Reprod Health. 2011; 8: 7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAnwar I, Sami M, Akhtar N, et al.: Inequity in maternal health-care services: evidence from home-based skilled-birth-attendant programmes in Bangladesh. Bull World Health Organ. 2008; 86(4): 252–259. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFosu GB: Childhood morbidity and health services utilization: cross-national comparisons of user-related factors from DHS data. Soc Sci Med. 1994; 38(9): 1209–1220. PubMed Abstract | Publisher Full Text\n\nLiang J, Mao M, Dai L, et al.: Neonatal mortality due to preterm birth at 28-36 weeks' gestation in China, 2003-2008. Paediatr Perinat Epidemiol. 2011; 25(6): 593–600. PubMed Abstract | Publisher Full Text\n\nMalqvist M, Nga NT, Eriksson L, et al.: Delivery care utilisation and care-seeking in the neonatal period: a population-based study in Vietnam. Ann Trop Paediatr. 2008; 28(3): 191–198. PubMed Abstract | Publisher Full Text\n\nMore NS, Alcock G, Das S, et al.: Spoilt for choice? Cross-sectional study of care-seeking for health problems during pregnancy in Mumbai slums. Glob Public Health. 2011; 6(7): 746–759. PubMed Abstract | Publisher Full Text | Free Full Text\n\nObermeyer CM, Potter JE: Maternal health care utilization in Jordan: a study of patterns and determinants. Stud Fam Plann. 1991; 22(3): 177–187. PubMed Abstract | Publisher Full Text\n\nRahman MH, Mosley WH, Ahmed S, et al.: Does service accessibility reduce socioeconomic differentials in maternity care seeking? Evidence from rural Bangladesh. J Biosoc Sci. 2008; 40(1): 19–33. PubMed Abstract | Publisher Full Text\n\nRai RK, Singh PK, Singh L: Utilization of maternal health care services among married adolescent women: insights from the Nigeria Demographic and Health Survey, 2008. Womens Health Issues. 2012; 22(4): e407–414. PubMed Abstract | Publisher Full Text\n\nRamos S, Karolinski A, Romero M, et al.: A comprehensive assessment of maternal deaths in Argentina: translating multicentre collaborative research into action. Bull World Health Organ. 2007; 85(8): 615–622. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRonsmans C, Chowdhury ME, Koblinsky M, et al.: Care seeking at time of childbirth, and maternal and perinatal mortality in Matlab, Bangladesh. Bull World Health Organ. 2010; 88(4): 289–296. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRööst M, Altamirano VC, Liljestrand J, et al.: Does antenatal care facilitate utilization of emergency obstetric care? A case-referent study of near-miss morbidity in Bolivia. Acta Obstet Gynecol Scand. 2010; 89(3): 335–342. PubMed Abstract | Publisher Full Text\n\nSaini AG, Bharti B, Gautam S: Healthcare behavior and expenditure in an urban slum in relation to birth experience and newborn care. J Trop Pediatr. 2012; 58(3): 214–219. PubMed Abstract | Publisher Full Text\n\nSein KK: Maternal health care utilization among ever married youths in Kyimyindaing Township, Myanmar. Matern Child Health J. 2012; 16(5): 1021–1030. PubMed Abstract | Publisher Full Text\n\nSharma SK, Vong-Ek P: Contextual influences on obstetric morbidity and related care seeking behaviour in Thailand. J Health Psychol. 2009; 14(1): 108–123. PubMed Abstract | Publisher Full Text\n\nBrenner JL, Kabakyenga J, Kyomuhangi T, et al.: Can volunteer community health workers decrease child morbidity and mortality in southwestern Uganda? An impact evaluation. PLoS One. 2011; 6(12): e27997. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVan der Stuyft P, Sorensen SC, Delgado E, et al.: Health seeking behaviour for child illness in rural Guatemala. Trop Med Int Health. 1996; 1(2): 161–170. PubMed Abstract | Publisher Full Text\n\nSutrisna B, Reingold A, Kresno S, et al.: Care-seeking for fatal illnesses in young children in Indramayu, west Java, Indonesia. Lancet. 1993; 342(8874): 787–789. PubMed Abstract | Publisher Full Text\n\nTeerawichitchainan B, Phillips JF: Ethnic differentials in parental health seeking for childhood illness in Vietnam. Soc Sci Med. 2008; 66(5): 1118–1130. PubMed Abstract | Publisher Full Text\n\nOgunlesi TA, Olanrewaju DM: Socio-demographic factors and appropriate health care-seeking behavior for childhood illnesses. J Trop Pediatr. 2010; 56(6): 379–385. PubMed Abstract | Publisher Full Text\n\nTitaley CR, Dibley MJ, Roberts CL: Factors associated with underutilization of antenatal care services in Indonesia: results of Indonesia Demographic and Health Survey 2002/2003 and 2007. BMC Public Health. 2010; 10: 485. PubMed Abstract | Publisher Full Text | Free Full Text\n\nToan NV, Hoa HT, Trong PV, et al.: Utilisation of reproductive health services in rural Vietnam; are there equal opportunities to plan and protect pregnancies? J Epidemiol Community Health. 1996; 50(4): 451–455. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTrinh LT, Dibley MJ, Byles J: Determinants of antenatal care utilization in three rural areas of Vietnam. Public Health Nurs. 2007; 24(4): 300–310. PubMed Abstract | Publisher Full Text\n\nYe Y, Yoshida Y, Harun-Or-Rashid M, et al.: Factors affecting the utilization of antenatal care services among women in Kham District, Xiengkhouang province, Lao PDR. Nagoya J Med Sci. 2010; 72(1–2): 23–33. PubMed Abstract | Publisher Full Text\n\nErci B: Barriers to utilization of prenatal care services in Turkey. J Nurs Scholarsh. 2003; 35(3): 269–273. PubMed Abstract | Publisher Full Text\n\nUzma A, Underwood P, Atkinson D, et al.: Postpartum health in a Dhaka slum. Soc Sci Med. 1999; 48(3): 313–320. PubMed Abstract | Publisher Full Text\n\nFotso JC, Ezeh A, Oronje R: Provision and use of maternal health services among urban poor women in Kenya: what do we know and what can we do? J Urban Health. 2008; 85(3): 428–442. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSunil TS, Rajaram S, Zottarelli LK: Do individual and program factors matter in the utilization of maternal care services in rural India?: a theoretical approach. Soc Sci Med. 2006; 62(8): 1943–1957. PubMed Abstract | Publisher Full Text\n\nSchillaci MA, Waitzkin H, Carson EA, et al.: Prenatal care utilization for mothers from low-income areas of New Mexico, 1989-1999. PLoS One. 2010; 5(9): pii: e12809. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbraham S, Joshi S, Kumar V, et al.: Indian experience of home based mothers card: ICMR task force study. Indian J Pediatr. 1991; 58(6): 795–804. PubMed Abstract | Publisher Full Text\n\nAhmed S, Sobhan F, Islam A, et al.: Neonatal morbidity and care-seeking behaviour in rural Bangladesh. J Trop Pediatr. 2001; 47(2): 98–105. PubMed Abstract | Publisher Full Text\n\nCarlson M, Smith Paintain L, Bruce J, et al.: Who attends antenatal care and expanded programme on immunization services in Chad, Mali and Niger? The implications for insecticide-treated net delivery. Malar J. 2011; 10: 341. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChamberlain J, Watt S, Mohide P, et al.: Women's perception of self-worth and access to health care. Int J Gynaecol Obstet. 2007; 98(1): 75–79. PubMed Abstract | Publisher Full Text\n\nCiceklioglu M, Soyer MT, Ocek ZA: Factors associated with the utilization and content of prenatal care in a western urban district of Turkey. Int J Qual Health Care. 2005; 17(6): 533–539. PubMed Abstract | Publisher Full Text\n\nEl-Kak F, Khawaja M, Salem M, et al.: Care-seeking behavior of women with reproductive health problems from low-income areas of Beirut. Int J Gynaecol Obstet. 2009; 104(1): 60–63. PubMed Abstract | Publisher Full Text\n\nElo IT: Utilization of maternal health-care services in Peru: the role of women's education. Health Transit Rev. 1992; 2(1): 49–69. PubMed Abstract\n\nOkafor CB: Availability and use of services for maternal and child health care in rural Nigeria. Int J Gynaecol Obstet. 1991; 34(4): 331–346. PubMed Abstract | Publisher Full Text\n\nHaque SE, Rahman M, Mostofa MG, et al.: Reproductive health care utilization among young mothers in Bangladesh: does autonomy matter? Womens Health Issues. 2012; 22(2): e171–180. PubMed Abstract | Publisher Full Text\n\nHabibov NN, Fan L: Modelling prenatal health care utilization in Tajikistan using a two-stage approach: implications for policy and research. Health Policy Plan. 2008; 23(6): 443–451. PubMed Abstract | Publisher Full Text\n\nvan den Heuvel OA, de Mey WG, Buddingh H, et al.: Use of maternal care in a rural area of Zimbabwe: a population-based study. Acta Obstet Gynecol Scand. 1999; 78(10): 838–846. PubMed Abstract | Publisher Full Text\n\nHirose A, Borchert M, Niksear H, et al.: Difficulties leaving home: a cross-sectional study of delays in seeking emergency obstetric care in Herat, Afghanistan. Soc Sci Med. 2011; 73(7): 1003–1013. PubMed Abstract | Publisher Full Text\n\nGombojav N, Manaseki-Holland S, Pollock J, et al.: The effects of social variables on symptom recognition and medical care seeking behaviour for acute respiratory infections in infants in urban Mongolia. Arch Dis Child. 2009; 94(11): 849–854. PubMed Abstract | Publisher Full Text\n\nHazarika I: Factors that determine the use of skilled care during delivery in India: implications for achievement of MDG-5 targets. Matern Child Health J. 2011; 15(8): 1381–1388. PubMed Abstract | Publisher Full Text\n\nIbrahim SA, Omer MI, Amin IK, et al.: The role of the village midwife in detection of high risk pregnancies and newborns. Int J Gynaecol Obstet. 1992; 39(2): 117–122. PubMed Abstract | Publisher Full Text\n\nMercer A, Haseen F, Huq NL, et al.: Risk factors for neonatal mortality in rural areas of Bangladesh served by a large NGO programme. Health Policy Plan. 2006; 21(6): 432–443. PubMed Abstract | Publisher Full Text\n\nOgunlesi TA, Ogunlesi FB: Family socio-demographic factors and maternal obstetric factors influencing appropriate health-care seeking behaviours for newborn jaundice in Sagamu, Nigeria. Matern Child Health J. 2012; 16(3): 677–684. PubMed Abstract | Publisher Full Text\n\nRani M, Bonu S: Rural Indian women's care-seeking behavior and choice of provider for gynecological symptoms. Stud Fam Plann. 2003; 34(3): 173–185. PubMed Abstract | Publisher Full Text\n\nNdugwa RP, Zulu EM: Child morbidity and care-seeking in Nairobi slum settlements: the role of environmental and socio-economic factors. J Child Health Care. 2008; 12(4): 314–328. PubMed Abstract | Publisher Full Text\n\nEnato EF, Mens PF, Okhamafe AO, et al.: Plasmodium falciparum malaria in pregnancy: prevalence of peripheral parasitaemia, anaemia and malaria care-seeking behaviour among pregnant women attending two antenatal clinics in Edo State, Nigeria. J Obstet Gynaecol. 2009; 29(4): 301–306. PubMed Abstract | Publisher Full Text\n\nRegassa N: Antenatal and postnatal care service utilization in southern Ethiopia: a population-based study. Afr Health Sci. 2011; 11(3): 390–397. PubMed Abstract | Free Full Text\n\nGalvan J, Woelk GB, Mahomed K, et al.: Prenatal care utilization and foetal outcomes at Harare Maternity Hospital, Zimbabwe. Cent Afr J Med. 2001; 47(4): 87–92. PubMed Abstract\n\nKhan AK: Obstetric complications: the health care seeking behaviour & cost pressure generated from it in rural Bangladesh. Mymensingh Med J. 2002; 11(2): 110–112. PubMed Abstract\n\nOkusanya BO, Okogbo FO, Momoh MM, et al.: Maternal mortality and delay: socio-demographic characteristics of maternal deaths with delay in Irrua, Nigeria. Niger J Med. 2007; 16(1): 38–41. PubMed Abstract | Publisher Full Text\n\nIyoke CA, Ifeadike CO, Nnebue CC, et al.: Perception and care-seekling behaviour for post partum morbidity among mothers in Enugu south east, Nigeria. Niger J Med. 2011; 20(2): 260–265. PubMed Abstract\n\nBloom SS, Wypij D, Das Gupta M: Dimensions of women's autonomy and the influence on maternal health care utilization in a north Indian city. Demography. 2001; 38(1): 67–78. PubMed Abstract | Publisher Full Text\n\nBazzano AN, Kirkwood BR, Tawiah-Agyemang C, et al.: Beyond symptom recognition: care-seeking for ill newborns in rural Ghana. Trop Med Int Health. 2008; 13(1): 123–128. PubMed Abstract | Publisher Full Text\n\nAlcock G, Das S, Shah More N, et al.: Examining inequalities in uptake of maternal health care and choice of provider in underserved urban areas of Mumbai, India: a mixed methods study. BMC Pregnancy Childbirth. 2015; 15: 231. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoldman N, Heuveline P: Health-seeking behaviour for child illness in Guatemala. Trop Med Int Health. 2000; 5(2): 145–155. PubMed Abstract | Publisher Full Text\n\nAmarasiri de Silva MW, Wijekoon A, Hornik R, et al.: Care seeking in Sri Lanka: one possible explanation for low childhood mortality. Soc Sci Med. 2001; 53(10): 1363–1372. PubMed Abstract | Publisher Full Text\n\nMash D, Aschenaki K, Kedamo T, et al.: Community and facility surveys illuminate the pathway to child survival in Liben Woreda, Ethiopia. East Afr Med J. 2003; 80(9): 463–469. PubMed Abstract | Publisher Full Text\n\nGupta N, Jain SK, Chawla U, et al.: An evaluation of diarrheal diseases and acute respiratory infections control programmes in a Delhi slum. Indian J Pediatr. 2007; 74(5): 471–476. PubMed Abstract | Publisher Full Text\n\nDelgado E, Sorensen SC, Van der Stuyft P: Health seeking behaviour and self-treatment for common childhood symptoms in rural Guatemala. Ann Soc Belg Med Trop. 1994; 74(2): 161–168. PubMed Abstract\n\nde Zoysa I, Bhandari N, Akhtari N, et al.: Careseeking for illness in young infants in an urban slum in India. Soc Sci Med. 1998; 47(12): 2101–2111. PubMed Abstract | Publisher Full Text\n\nAdair T, Pardosi JF, Rao C, et al.: Access to health services and early age mortality in Ende, Indonesia. Indian J Pediatr. 2012; 79(5): 612–618. PubMed Abstract | Publisher Full Text\n\nOkong P, Byamugisha J, Mirembe F, et al.: Audit of severe maternal morbidity in Uganda--implications for quality of obstetric care. Acta Obstet Gynecol Scand. 2006; 85(7): 797–804. PubMed Abstract | Publisher Full Text\n\nSafdar S, Inam SN, Omair A, et al.: Maternal health care in a rural area of Pakistan. J Pak Med Assoc. 2002; 52(7): 308–311. PubMed Abstract\n\nAllendorf K: The quality of family relationships and use of maternal health-care services in India. Stud Fam Plann. 2010; 41(4): 263–276. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKamiya Y: Women's autonomy and reproductive health care utilisation: empirical evidence from Tajikistan. Health Policy. 2011; 102(2–3): 304–313. PubMed Abstract | Publisher Full Text\n\nRahman MM, Kabir M, Shahidullah M: Adolescent self reported reproductive morbidity and health care seeking behaviour. J Ayub Med Coll Abbottabad. 2004; 16(2): 9–14. PubMed Abstract\n\nKumar V, Mohanty S, Kumar A, et al.: Effect of community-based behaviour change management on neonatal mortality in Shivgarh, Uttar Pradesh, India: a cluster-randomised controlled trial. Lancet. 2008; 372(9644): 1151–1162. PubMed Abstract | Publisher Full Text\n\nGranich R, Cantwell MF, Long K, et al.: Patterns of health seeking behavior during episodes of childhood diarrhea: a study of Tzotzil-speaking Mayans in the highlands of Chiapas, Mexico. Soc Sci Med. 1999; 48(4): 489–495. PubMed Abstract | Publisher Full Text\n\nKnobel HH, Yang WS, Ho MS: Urban-rural and regional differences in infant mortality in Taiwan. Soc Sci Med. 1994; 39(6): 815–822. PubMed Abstract | Publisher Full Text\n\nThapa DK, Niehof A: Women's autonomy and husbands' involvement in maternal health care in Nepal. Soc Sci Med. 2013; 93: 1–10. PubMed Abstract | Publisher Full Text\n\nSimkhada B, Porter MA, van Teijlingen ER: The role of mothers-in-law in antenatal care decision-making in Nepal: a qualitative study. BMC Pregnancy Childbirth. 2010; 10: 34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAwasthi S, Srivastava NM, Agarwal GG, et al.: Effect of behaviour change communication on qualified medical care-seeking for sick neonates among urban poor in Lucknow, northern India: a before and after intervention study. Trop Med Int Health. 2009; 14(10): 1199–1209. PubMed Abstract | Publisher Full Text\n\nCastro R, Campero L, Hernandez B, et al.: A study on maternal mortality in Mexico through a qualitative approach. J Womens Health Gend Based Med. 2000; 9(6): 679–690. PubMed Abstract | Publisher Full Text\n\nRabiu KA, Adewunmi AA, Akinlusi FM, et al.: Female reproductive tract infections: understandings and care seeking behaviour among women of reproductive age in Lagos, Nigeria. BMC Womens Health. 2010; 10: 8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMagoma M, Requejo J, Campbell OM, et al.: High ANC coverage and low skilled attendance in a rural Tanzanian district: a case for implementing a birth plan intervention. BMC Pregnancy Childbirth. 2010; 10: 13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMathole T, Lindmark G, Majoko F, et al.: A qualitative study of women's perspectives of antenatal care in a rural area of Zimbabwe. Midwifery. 2004; 20(2): 122–132. PubMed Abstract | Publisher Full Text\n\nRosato M, Mwansambo CW, Kazembe PN, et al.: Women's groups' perceptions of maternal health issues in rural Malawi. Lancet. 2006; 368(9542): 1180–1188. PubMed Abstract | Publisher Full Text\n\nSibley LM, Hruschka D, Kalim N, et al.: Cultural theories of postpartum bleeding in Matlab, Bangladesh: implications for community health intervention. J Health Popul Nutr. 2009; 27(3): 379–390. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWong GC, Li VC, Burris MA, et al.: Seeking women's voices: setting the context for women's health interventions in two rural counties in Yunnan, China. Soc Sci Med. 1995; 41(8): 1147–1157. PubMed Abstract | Publisher Full Text\n\nBarua A, Kurz K: Reproductive health-seeking by married adolescent girls in Maharashtra, India. Reprod Health Matters. 2001; 9(17): 53–62. PubMed Abstract | Publisher Full Text\n\nAbrahams N, Jewkes R, Mvo Z: Health care-seeking practices of pregnant women and the role of the midwife in Cape Town, South Africa. J Midwifery Womens Health. 2001; 46(4): 240–247. PubMed Abstract | Publisher Full Text\n\nGurung G: Practices on immediate care of newborn in the communities of Kailali district. Nepal Med Coll J. 2008; 10(1): 41–44. PubMed Abstract\n\nKalim N, Anwar I, Khan J, et al.: Postpartum haemorrhage and eclampsia: differences in knowledge and care-seeking behaviour in two districts of Bangladesh. J Health Popul Nutr. 2009; 27(2): 156–169. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSikder SS, Labrique AB, Ullah B, et al.: Accounts of severe acute obstetric complications in rural Bangladesh. BMC Pregnancy Childbirth. 2011; 11: 76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCommittee on Nutrition: Calcium Requirements of Infants, Children, and Adolescents. Pediatrics. 1999; 104(5): 1152–1157. Publisher Full Text\n\nOkyere E, Tawiah-Agyemang C, Manu A, et al.: Newborn care: the effect of a traditional illness, asram, in Ghana. Ann Trop Paediatr. 2010; 30(4): 321–328. PubMed Abstract | Publisher Full Text\n\nFikree FF, Ali T, Durocher JM, et al.: Health service utilization for perceived postpartum morbidity among poor women living in Karachi. Soc Sci Med. 2004; 59(4): 681–694. PubMed Abstract | Publisher Full Text\n\nWulandari LP, Klinken Whelan A: Beliefs, attitudes and behaviours of pregnant women in Bali. Midwifery. 2011; 27(6): 867–871. PubMed Abstract | Publisher Full Text\n\nYadav SP: A study of treatment seeking behaviour for malaria and its management in febrile children in rural part of desert, Rajasthan, India. J Vector Borne Dis. 2010; 47(4): 235–242. PubMed Abstract\n\nPatel Z, Kumar V, Singh P, et al.: Feasibility of community neonatal death audits in rural Uttar Pradesh, India. J Perinatol. 2007; 27(9): 556–564. PubMed Abstract | Publisher Full Text\n\nFarnes C, Beckstrand RL, Callister LC: Help-seeking behaviours in childbearing women in Ghana, West Africa. Int Nurs Rev. 2011; 58(4): 491–497. PubMed Abstract | Publisher Full Text\n\nYassin K, Laaser U, Kraemer A: Maternal morbidity in rural upper Egypt: levels, determinants, and care seeking. Health Care Women Int. 2003; 24(5): 452–467. PubMed Abstract | Publisher Full Text\n\nMoran AC, Winch PJ, Sultana N, et al.: Patterns of maternal care seeking behaviours in rural Bangladesh. Trop Med Int Health. 2007; 12(7): 823–832. PubMed Abstract | Publisher Full Text\n\nPettersson KO, Christensson K, de Freitas Eda G, et al.: Adaptation of health care seeking behavior during childbirth: focus group discussions with women living in the suburban areas of Luanda, Angola. Health Care Women Int. 2004; 25(3): 255–280. PubMed Abstract | Publisher Full Text\n\nKhadduri R, Marsh DR, Rasmussen B, et al.: Household knowledge and practices of newborn and maternal health in Haripur district, Pakistan. J Perinatol. 2008; 28(3): 182–187. PubMed Abstract | Publisher Full Text\n\nHead SK, Yount KM, Sibley LM: Delays in recognition of and care-seeking response to prolonged labor in Bangladesh. Soc Sci Med. 2011; 72(7): 1157–1168. PubMed Abstract | Publisher Full Text\n\nTitaley CR, Hunter CL, Dibley MJ, et al.: Why do some women still prefer traditional birth attendants and home delivery?: a qualitative study on delivery care services in West Java Province, Indonesia. BMC Pregnancy Childbirth. 2010; 10: 43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEssendi H, Mills S, Fotso JC: Barriers to formal emergency obstetric care services' utilization. J Urban Health. 2011; 88 Suppl 2: S356–369. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchooley J, Mundt C, Wagner P, et al.: Factors influencing health care-seeking behaviours among Mayan women in Guatemala. Midwifery. 2009; 25(4): 411–421. PubMed Abstract | Publisher Full Text\n\nWhite K, Small M, Frederic R, et al.: Health seeking behavior among pregnant women in rural Haiti. Health Care Women Int. 2006; 27(9): 822–838. PubMed Abstract | Publisher Full Text\n\nAtuyambe L, Mirembe F, Annika J, et al.: Seeking safety and empathy: adolescent health seeking behavior during pregnancy and early motherhood in central Uganda. J Adolesc. 2009; 32(4): 781–796. PubMed Abstract | Publisher Full Text\n\nLubbock LA, Stephenson RB: Utilization of maternal health care services in the department of Matagalpa, Nicaragua. Rev Panam Salud Publica. 2008; 24(2): 75–84. PubMed Abstract | Publisher Full Text\n\nPritham UA, Sammons LN: Korean women's attitudes toward pregnancy and prenatal care. Health Care Women Int. 1993; 14(2): 145–153. PubMed Abstract | Publisher Full Text\n\nBirungi H, Obare F, van der Kwaak A, et al.: Maternal health care utilization among HIV-positive female adolescents in Kenya. Int Perspect Sex Reprod Health. 2011; 37(3): 143–149. PubMed Abstract | Publisher Full Text\n\nWalraven G, Telfer M, Rowley J, et al.: Maternal mortality in rural Gambia: levels, causes and contributing factors. Bull World Health Organ. 2000; 78(5): 603–613. PubMed Abstract | Free Full Text\n\nMahdi SS, Habib OS: A study on preference and practices of women regarding place of delivery. East Mediterr Health J. 2010; 16(8): 874–878. PubMed Abstract | Publisher Full Text\n\nKillewo J, Anwar I, Bashir I, et al.: Perceived delay in healthcare-seeking for episodes of serious illness and its implications for safe motherhood interventions in rural Bangladesh. J Health Popul Nutr. 2006; 24(4): 403–412. PubMed Abstract | Free Full Text\n\nKoenig MA, Jamil K, Streatfield PK, et al.: Maternal health and care-seeking behavior in Bangladesh: findings from a national survey. Int Fam Plan Perspect. 2007; 33(2): 75–82. PubMed Abstract | Publisher Full Text\n\nIyengar K, Iyengar SD, Suhalka V, et al.: Pregnancy-related deaths in rural Rajasthan, India: exploring causes, context, and care-seeking through verbal autopsy. J Health Popul Nutr. 2009; 27(2): 293–302. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOwais A, Sultana S, Stein AD, et al.: Why do families of sick newborns accept hospital care? A community-based cohort study in Karachi, Pakistan. J Perinatol. 2011; 31(9): 586–592. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNabukera SK, Witte K, Muchunguzi C, et al.: Use of postpartum health services in rural Uganda: knowledge, attitudes, and barriers. J Community Health. 2006; 31(2): 84–93. PubMed Abstract | Publisher Full Text\n\nKambarami RA, Chirenje MZ, Rusakaniko S: Antenatal care patterns and factors associated with perinatal outcome in two rural districts in Zimbabwe. Cent Afr J Med. 1999; 45(11): 294–299. PubMed Abstract | Publisher Full Text\n\nMoronkola OA, Omonu JB, Iyayi DA, et al.: Perceived determinants of the utilization of maternal health-care services by rural women in Kogi State, Nigeria. Trop Doct. 2007; 37(2): 94–96. PubMed Abstract | Publisher Full Text\n\nMateria E, Mehari W, Mele A, et al.: A community survey on maternal and child health services utilization in rural Ethiopia. Eur J Epidemiol. 1993; 9(5): 511–516. PubMed Abstract | Publisher Full Text\n\nPeltzer K, Mosala T, Shisana O, et al.: Utilization of delivery services in the context of prevention of HIV from mother-to-child (PMTCT) in a rural community, South Africa. Curationis. 2006; 29(1): 54–61. PubMed Abstract | Publisher Full Text\n\nMoore BM, Alex-Hart BA, George IO: Utilization of health care services by pregnant mothers during delivery: a community based study in Nigeria. East Afr J Public Health. 2011; 8(1): 49–51. PubMed Abstract\n\nOkwaraji YB, Cousens S, Berhane Y, et al.: Effect of geographical access to health facilities on child mortality in rural Ethiopia: a community based cross sectional study. PLoS One. 2012; 7(3): e33564. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPhoxay C, Okumura J, Nakamura Y, et al.: Influence of women's knowledge on maternal health care utilization in southern Laos. Asia Pac J Public Health. 2001; 13(1): 13–19. PubMed Abstract | Publisher Full Text\n\nFaye A, Faye M, Bâ IO, et al.: [Factors determining the place of delivery in women who attended at least one antenatal consultation in a health facility (Senegal)]. Rev Epidemiol Sante Publique. 2010; 58(5): 323–329. PubMed Abstract | Publisher Full Text\n\nMambu Nyangi Mondo T, Malengreau M, Kalambayi P, et al.: [Delays in seeking and getting care, in seriously ill women of childbearing age in Kinshasa]. Rev Epidemiol Sante Publique. 2010; 58(3): 189–196. PubMed Abstract | Publisher Full Text\n\nMyer L, Harrison A: Why do women seek antenatal care late? Perspectives from rural South Africa. J Midwifery Womens Health. 2003; 48(4): 268–272. PubMed Abstract | Publisher Full Text\n\nAtuoye KN, Dixon J, Rishworth A, et al.: Can she make it? Transportation barriers to accessing maternal and child health care services in rural Ghana. BMC Health Serv Res. 2015; 15: 333. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSarfraz M, Tariq S, Hamid S, et al.: Social And Societal Barriers In Utilization Of Maternal Health Care Services In Rural Punjab, Pakistan. J Ayub Med Coll Abbottabad. 2015; 27(4): 843–9. PubMed Abstract\n\nMrisho M, Obrist B, Schellenberg JA, et al.: The use of antenatal and postnatal care: perspectives and experiences of women and health care providers in rural southern Tanzania. BMC Pregnancy Childbirth. 2009; 9: 10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUyirwoth GP, Itsweng MD, Mpai S, et al.: Obstetrics service utilisation by the community in Lebowa, northern Transvaal. East Afr Med J. 1996; 73(2): 91–94. PubMed Abstract\n\nLama S, Krishna AK: Barriers in Utilization of Maternal Health Care Services: Perceptions of Rural Women in Eastern Nepal. Kathmandu Univ Med J (KUMJ). 2014; 12(48): 253–258. PubMed Abstract | Publisher Full Text\n\nHunt LM, Glantz NM, Halperin DC: Childbirth care-seeking behavior in Chiapas. Health Care Women Int. 2002; 23(1): 98–118. PubMed Abstract | Publisher Full Text\n\nMacLeod J, Rhode R: Retrospective follow-up of maternal deaths and their associated risk factors in a rural district of Tanzania. Trop Med Int Health. 1998; 3(2): 130–137. PubMed Abstract | Publisher Full Text\n\nIyun F: An assessment of a rural health programme on child and maternal care: the Ogbomoso Community Health Care Programme (CHCP), Oyo State, Nigeria. Soc Sci Med. 1989; 29(8): 933–938. PubMed Abstract | Publisher Full Text\n\nMbonye AK, Neema S, Magnussen P: Malaria in pregnancy, risk perceptions and care seeking practices among adolescents in Mukono district Uganda. Int J Adolesc Med Health. 2006; 18(4): 561–573. PubMed Abstract | Publisher Full Text\n\nFoster J, Burgos R, Tejada C, et al.: A community-based participatory research approach to explore community perceptions of the quality of maternal-newborn health services in the Dominican Republic. Midwifery. 2010; 26(5): 504–511. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYakong VN, Rush KL, Bassett-Smith J, et al.: Women's experiences of seeking reproductive health care in rural Ghana: challenges for maternal health service utilization. J Adv Nurs. 2010; 66(11): 2431–2441. PubMed Abstract | Publisher Full Text\n\nTerra de Souza AC, Peterson KE, Andrade FM, et al.: Circumstances of post-neonatal deaths in Ceara, Northeast Brazil: mothers' health care-seeking behaviors during their infants' fatal illness. Soc Sci Med. 2000; 51(11): 1675–1693. PubMed Abstract | Publisher Full Text\n\nKongnyuy EJ, Hofman J, Mlava G, et al.: Availability, utilisation and quality of basic and comprehensive emergency obstetric care services in Malawi. Matern Child Health J. 2009; 13(5): 687–694. PubMed Abstract | Publisher Full Text\n\nMbonye AK: Prevalence of childhood illnesses and care-seeking practices in rural Uganda. ScientificWorldJournal. 2003; 3: 721–730. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKnight HE, Self A, Kennedy SH: Why are women dying when they reach hospital on time? A systematic review of the 'third delay'. PLoS One. 2013; 8(5): e63846. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMangham-Jefferies L, Pitt C, Cousens S, et al.: Cost-effectiveness of strategies to improve the utilization and provision of maternal and newborn health care in low-income and lower-middle-income countries: a systematic review. BMC Pregnancy Childbirth. 2014; 14: 243. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "45478",
"date": "25 Mar 2019",
"name": "Aastha Kant",
"expertise": [
"Reviewer Expertise Qualitative research methodology",
"reproductive",
"maternal",
"newborn and child 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\nThe present article is a systematic review of qualitative and observational research studies on care seeking for maternal and newborn illness in low and middle income countries. There has been a knowledge gap in synthesizing studies to understand barriers and facilitators in the context of the Three Delay Model and pathways of care seeking.\nThe literature search has been conducted in a robust manner. The authors could expand a little more on 'narrative synthesis' of the findings in the context of qualitative and observational studies.\n\nThe authors have included literature search up to September 2017. A special issue of the Journal of Health, Population and Nutrition (December 2017), focuses on illness recognition and care seeking pathways of maternal and newborn illness in seven countries. The authors might like to update the search to include more recent articles.\n\nThe authors have synthesized the findings of the studies, combining maternal and newborn illness. They may want to present the care seeking pathways for maternal and newborn illness independently to explore the effect of the timing of the onset of illness (during pregnancy, childbirth or post childbirth for maternal illness, for instance) on care seeking behavior. This will also allow the authors to explore how the different perceptions of the causes of illness- supernatural or medical - often influence care seeking pathways.\nOverall, this is a very relevant work that can add to the body of knowledge in the context of maternal and newborn health.\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? Not applicable\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": []
},
{
"id": "44643",
"date": "23 Apr 2019",
"name": "Alfonso Rosales",
"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\nThis systematic review of maternal and newborn care-seeking practices adds important information regarding maternal and newborn care seeking patterns and associated barriers.\n\nWith its methodology clear and sound, results provide important findings to support future health program design in service provision for mothers and newborn.\n\nOne important aspect not clearly addressed, in the results and discussion section, is related to the difference or lack of thereof between maternal and newborn care-seeking pathway. There seems to be published evidence on this subject, which seems to lead on the different care-seeking pathways, with consequences in future program design. Authors, may want to include a special issue on this theme published at the Journal of Health, Population and nutrition in December of 2017.\n\nThe conclusion, seems to be rather short and not conclusive enough for relevant findings. It leaves out important common factors such as the need to include community/user context for strategy/intervention design, for example.\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": []
}
] | 1
|
https://f1000research.com/articles/8-200
|
https://f1000research.com/articles/8-198/v1
|
18 Feb 19
|
{
"type": "Research Article",
"title": "Nuclear envelope impairment is facilitated by the herpes simplex virus 1 Us3 kinase",
"authors": [
"Peter Wild",
"Sabine Leisinger",
"Anna Paula de Oliveira",
"Jana Doehner",
"Elisabeth M. Schraner",
"Cornel Fraevel",
"Mathias Ackermann",
"Andres Kaech",
"Sabine Leisinger",
"Anna Paula de Oliveira",
"Jana Doehner",
"Elisabeth M. Schraner",
"Cornel Fraevel",
"Mathias Ackermann",
"Andres Kaech"
],
"abstract": "Background: Capsids of herpes simplex virus 1 (HSV-1) are assembled in the nucleus, translocated either to the perinuclear space by budding at the inner nuclear membrane acquiring tegument and envelope, or released to the cytosol in a “naked” state via impaired nuclear pores that finally results in impairment of the nuclear envelope. The Us3 gene encodes a protein acting as a kinase, which is responsible for phosphorylation of numerous viral and cellular substrates. The Us3 kinase plays a crucial role in nucleus to cytoplasm capsid translocation. We thus investigate the nuclear surface in order to evaluate the significance of Us3 in maintenance of the nuclear envelope during HSV-1 infection. Methods: To address alterations of the nuclear envelope and capsid nucleus to cytoplasm translocation related to the function of the Us3 kinase we investigated cells infected with wild type HSV-1 or the Us3 deletion mutant R7041(∆Us3) by transmission electron microscopy, focused ion-beam electron scanning microscopy, cryo-field emission scanning electron microscopy, confocal super resolution light microscopy, and polyacrylamide gel electrophoresis. Results: Confocal super resolution microscopy and cryo-field emission scanning electron microscopy revealed decrement in pore numbers in infected cells. Number and degree of pore impairment was significantly reduced after infection with R7041(∆Us3) compared to infection with wild type HSV-1. The nuclear surface was significantly enlarged in cells infected with any of the viruses. Morphometric analysis revealed that additional nuclear membranes were produced forming multiple folds and caveolae, in which virions accumulated as documented by three-dimensional reconstruction after ion-beam scanning electron microscopy. Finally, significantly more R7041(∆Us3) capsids were retained in the nucleus than wild-type capsids whereas the number of R7041(∆Us3) capsids in the cytosol was significantly lower. Conclusions: The data indicate that Us3 kinase is involved in facilitation of nuclear pore impairment and, concomitantly, in capsid release through impaired nuclear envelope.",
"keywords": [
"HSV-1 egress",
"nuclear pores",
"nuclear envelope breakdown",
"intraluminal transport",
"budding",
"fusion"
],
"content": "Introduction\n\nCapsids of herpes simplex virus 1 (HSV-1) assemble in replication centers (RCs) in host cell nuclei (Quinlan et al., 1984). From there, they are transported to the nuclear periphery and are translocated to the cytoplasm via two diverse routes (Roizman et al., 2014). In one route, the nucleocytoplasmic barrier is overcome by budding of capsids at the inner nuclear membrane (INM). During budding, tegument and viral envelope are acquired (Granzow et al., 2001; Leuzinger et al., 2005). The result is a fully enveloped virion located in the perinuclear space (PNS) delineated by the INM and outer nuclear membrane (ONM) that are part of the endoplasmic reticulum (ER). Virions in the PNS have been proposed for 5 decades to de-envelope by fusion of the viral envelope with the ONM releasing capsid and tegument into the cytoplasmic matrix (Skepper et al., 2001; Stackpole, 1969) for secondary envelopment at the trans Golgi network (TGN). Envelopment at the INM, de-envelopment at the ONM and re-envelopment at the TGN have been proposed to be essential for production of infectious progeny virus e.g. (Mettenleiter et al., 2006). The cruxes of the de-envelopment theory are i) that the viral envelope of Us3 deletion mutants cannot fuse with the ONM (Wisner et al., 2009), and hence, their capsids cannot be released into the cytoplasmic matrix. Consequently, they cannot be re-enveloped. Instead, virions of Us3 deletion mutants accumulate in the PNS. Despite of the inability of de- and re-envelopment Us3 deletion mutants are fully infective (Reynolds et al., 2002; Wild et al., 2015; Wisner et al., 2009). ii) The process taking place at the ONM exhibits all characteristics of budding shown for the first time 50 years ago (Darlington & Moss, 1968). The process at the ONM also takes place in the absence of the fusion glycoproteins gB and gH leading to accumulation of virions in the PNS-ER compartment (Farnsworth et al., 2007). Therefore, the virus-membrane interaction taking place at the ONM is budding, indeed, not fusion as discussed in detail (Wild et al., 2018). Virions are transported out of the PNS into adjacent ER cisternae (Gilbert et al., 1994; Granzow et al., 1997; Maric et al., 2011; Radsak et al., 1996; Schwartz & Roizman, 1969; Stannard et al., 1996; Sutter et al., 2012; Whealy et al., 1991; Wild et al., 2002). ER membranes connect to Golgi membranes forming a PNS-ER-Golgi continuum that is considered very likely to function as a direct intraluminal transportation route for virions from the PNS into Golgi cisternae (Wild et al., 2018).\n\nTherefore, the question remains how naked capsids gain access to the cytoplasmic matrix if the viral envelope does not fuse with the ONM, and, consequently, de-envelopment does not take place. In cells infected with the monkey herpes pathogen simian agent 8 (Borchers & Ozel, 1993), capsids gained access to the cytoplasmic matrix via impaired nuclear envelope (NE). It was clearly shown that the ONM turned into the INM at the sites of NE breakdown indicating that the NE breakdown was rather a result of nuclear pore impairment than a rupture of nuclear membranes. In bovine herpes virus 1 (BoHV-1) infected MDBK cells (Wild et al., 2005) and in HSV-1 infected Vero cells (Leuzinger et al., 2005; Wild et al., 2009), impaired nuclear pores measured from about 150 nm to 300 nm. Large areas of impaired nuclear surface measuring several micrometers clearly exhibited intact transformation of the INM into the ONM indicating that NE impairment started by nuclear pore impairment. Impaired NE was also shown in cells infected with pseudorabies virus (PrV) UL31 and UL34-null recombinants (Grimm et al., 2012; Klupp et al., 2011; Schulz et al., 2015) as well as after HSV-1 infection of embryonic mouse fibroblasts (Maric et al., 2014). Capsids of HSV-1 and BoHV-1 were present in the nuclear matrix, which protruded through impaired nuclear pores into the cytoplasmic matrix, indicating that capsids are released via impaired NE. Capsids were also shown – though unrecognized – in impaired nuclear pores in HSV-1 infected mouse fibroblasts (Maric et al., 2014).\n\nUs3 is a multifunctional protein that plays various roles in the viral life cycle by phosphorylating more than 20 viral and cellular substrates (Kato & Kawaguchi, 2018). Phosphorylation of gB by Us3 was reported to be crucial for proper regulation of gB intracellular transport and in viral replication (Imai et al., 2011; Imai et al., 2010). Us3 is involved in blocking apoptosis induced by HSV-1 (Benetti et al., 2003; Deruelle et al., 2010; Jerome et al., 1999; Leopardi et al., 1997; Munger & Roizman, 2001; Ogg et al., 2004), bovine herpes virus 1 (Brzozowska et al., 2018) and PrV (Deruelle et al., 2010). Us3 kinase is supposed to play a crucial role in capsid nucleus to cytoplasm translocation in association with phosphorylation of viral proteins including glycoprotein B (Kato et al., 2009; Wisner et al., 2009), UL31 (Mou et al., 2009) and UL34 (Ryckman & Roller, 2004). The 3 proteins facilitate translocation of virions out of the PNS (Poon et al., 2006; Reynolds et al., 2004; Reynolds et al., 2001; Reynolds et al., 2002; Wisner et al., 2009). In contrast, inhibited nucleus to cytoplasm translocation was suggested to be independent of phosphorylation of UL34 by Us3 in PrV infected cells (Klupp et al., 2001). UL31 and UL34 also promote the late maturation of viral replication compartments at the periphery (Simpson-Holley et al., 2004), and are involved in nuclear expansion during HSV-1 infection (Simpson-Holley et al., 2005). Us3 kinase also phosphorylates the nuclear lamin A/C (Mou et al., 2007) and is involved in disrupting the nuclear lamina together with UL34 (Bjerke & Roller, 2006) possibly in association with phosphorylation of emerin (Leach et al., 2007). Recently, it was shown that UL31 and UL34 are responsible for budding of capsids at the INM (Bigalke & Heldwein, 2015; Bigalke & Heldwein, 2016; Hagen et al., 2015) and that the endosomal sorting complex required for transport-III (ESCRT III) is responsible for scission of the viral envelope from the INM (Arii et al., 2018). Us3 kinase down-regulates phospholipid biosynthesis (Wild et al., 2012a) induced by HSV-1 (Sutter et al., 2012) to maintain nuclear membrane integrity upon nuclear expansion and budding of capsids. Us3 kinase was suggested to inhibit breakdown of the NE (Maric et al., 2014).\n\nBased on the proposed effects of Us3 kinase on the NE and nucleus to cytoplasm translocation we investigated the nucleus and the nuclear periphery in Vero cells infected with wild type (wt) HSV-1, the Us3 deletion mutant R7041(∆Us3) (Purves et al., 1987; Purves et al., 1991) and its repair mutant R2641 by cryo-field emission scanning electron microscopy (cryo-FESEM) of cells after freezing and freeze-fracturing, by transmission electron microscopy (TEM) prepared by high pressure-freezing followed by freeze-substitution, and by super resolution light microscopy using the stimulated emission depletion (STED) principle. The cryo-techniques enable visualization of structures in great detail, and, even more important, in a state that is closest to the situation in living cells (Harreveld & Fifkova, 1975). The data suggest that Us3 kinase is involved in facilitation of nuclear pore impairment as well as in intranuclear capsid transportation and capsid release via impaired nuclear pores.\n\n\nMethods\n\nVero cells (European Collection of Cell Cultures, ECACC, 84113001) were grown in Dulbecco’s modified minimal essential medium (DMEM, 31885-023; Gibco, Bethesda, MD, USA) supplemented with penicillin (100 U/ml), streptomycin (100 μg/ml) (Anti-Anti, 15240-062, Gibco) and 10% fetal bovine serum (FBS; 2-01F10-I, Bio Concept, Allschwil, Switzerland). Wild-type (wt) strain F (Ejercito et al., 1968), the Us3 deletion mutant R7041(∆Us3) and the repair mutant R2641 (kindly provided by B. Roizman, The Marjorie B. Kovler Viral Oncology Laboratories, University of Chicago, Illinois, USA). Wt HSV-1 were propagated in Vero cells. Virus yields were determined by plaque titration. For infection, cells were washed with DMEM without FBS, inoculated with virus diluted in DMEM without FBS, and kept for 1 h at 37°C. Then, cells were quickly washed with PBS, and incubated at 37°C in the presence of DMEM supplemented with 2%FBS. For controls, cells were mock infected by the same procedure replacing virus suspension with DMEM without FBS.\n\n50 mm thick sapphire disks (100.00174, Bruegger, Minusio, Switzerland) measuring 3 mm in diameter were coated with 8–10 nm carbon obtained by evaporation under high vacuum conditions to enhance cell growth. Vero cells were grown for 2 days on sapphire disks placed in 6 well plates. Cells were inoculated with R7041(∆Us3), the repair mutant R2641 or wt HSV-1 at a multiplicity of infection (MOI) of 5, incubated at 37°C, and fixed at 9, 12, 16, 20 and 24 hours post infection (hpi) by adding 0.25% glutaraldehyde to the medium prior to freezing in a high-pressure freezing unit (HPM010; BAL-TEC, Balzers, Lichtenstein) and processed as described in detail (Wild, 2008). In brief, the frozen water was substituted with acetone in a freeze-substitution unit (FS 7500; Boeckeler Instruments, Tucson, AZ, USA) at -88°C, and subsequently fixed with 0.25% glutaraldehyde and 0.5% osmium tetroxide (in water) raising the temperature gradually to +2°C to achieve good contrast of membranes (Wild et al., 2001), and embedded in epon prepared by mixing 61g Epon 812 (45345, Merck, Darmstadt, Germany), 40g Dodecenylsuccinic anhydride (DDSA, 45346, Merck), 27g methyl nadic anhydride (MNA, 45347, Merck) and 1.92 ml 2,4,6-Tris(dimethylaminomethyl)phenol (DMP30, 45348, Merck) at 4°C followed by polymerization at 60°C for 2.5 days. Serial sections of 60 to 90 nm thickness were analyzed in a transmission electron microscope (CM12; FEI, Eindhoven, The Netherlands) equipped with a CCD camera (Ultrascan 1000; Gatan, Pleasanton, CA, USA) at an acceleration voltage of 100 kV.\n\nFor 3D reconstruction, a trimmed epon block was mounted on a regular SEM stub using conductive carbon and coated with 10 nm of carbon by electron beam evaporation to render the sample conductive. Ion milling and image acquisition was performed simultaneously in an Auriga 40 Crossbeam system (Zeiss, Oberkochen, Germany) using the FIBICS Nanopatterning engine (NPVE v4.6, Fibics Inc., Ottawa, Canada). A large trench was milled at a current of 16 nA and 30 kV, followed by fine milling at 240 pA and 30 kV during image acquisition with an advance of 5 nm per image. Prior to starting the fine milling and imaging, a protective Platinum layer of approximately 300 nm was applied on top of the surface of the area of interest using the single gas injection system at the FIB-SEM. Images were acquired at 1.9 kV (30 µm aperture) using an in-lens energy selective backscattered electron detector (ESB) with a grid voltage of 500 V, and a dwell time of 1 μs and a line averaging of 50 lines. The pixel size was set to 5 nm and tilt-corrected to obtain isotropic voxels. The final image stack was registered and cropped to the area of interest for segmentation using the TrakEM2 plug-in (version 1.0i) for Fiji image-processing package v1.51f.\n\nVero cells were grown in 25 cm2 cell culture flasks for 2 days prior to inoculation with R7041(∆Us3), wt HSV-1 or R2641 at MOI of 5. Cells were harvested at 16 hpi by trypsinization followed by centrifugation at 150 × g for 8 min. The pellet was resuspended in 1 ml fresh medium, collected in Eppendorf tubes and fixed by adding 0.25% glutaraldehyde to the medium. The suspension was kept in the tubes at 4°C until cells were sedimented. After removal of the supernatant cells were frozen in a high-pressure freezing machine EM HPM100 (Leica Microsystems, Vienna, Austria) as described in detail previously (Wild et al., 2012b; Wild et al., 2009). Cells were fractured at -120°C in a freeze-fracturing device BAF 060 (Leica Microsystems) in a vacuum of 10-7 mbar. The fractured surfaces were partially freeze-dried (“etched”) at -105°C for 2 min, and coated with 2.5 nm platinum/carbon by electron beam evaporation at an angle of 45°. Some specimens were coated additionally with 4 nm of carbon to reduce electron beam damage during imaging at high magnifications. Specimens were imaged in an Auriga 40 Cross Beam system equipped with a cryo-stage (Zeiss, Oberkochen, Germany) at -115°C and an acceleration voltage of 5 kV using the inlens secondary electron detector.\n\nCells were grown for 2 days on 0.17 mm thick cover slips measuring 12 mm in diameter (Hecht-Assistent, Sondheim, Germany) and inoculated with R7041(∆Us3), wt HSV-1 or R2641 at a MOI of 5 and incubated at 37°C. After fixation with 2% formaldehyde for 25 min at room temperature, cells were permeabilized with 0.1% Triton-X-100 at room temperature for 7 min and blocked with 3% bovine serum albumin in phosphate-buffered saline containing 0.05% Tween 20 (PBST20). To identify nuclear pore complexes, cells incubated for 16 h were processed as described (Wild et al., 2009) using mouse monoclonal antibodies Mab414 (MMS-120P, Covance, Princeton, NJ, USA), and Alexa 488-conjugated secondary antibodies (goat anti-mouse, A32723, Thermo Fisher, Rockford, IL, USA). To identify infectivity, cells were labeled with polyclonal antibodies (1:1000) raised in rabbits against the tegument protein VP16 (gift from B. Roizman), and with Alexa 594-conjugated secondary antibodies, diluted 1:500, (goat anti-rabbit, A11037, Thermo Fisher Scientific). For measuring nuclear diameters, nuclei were stained with 4’,6’-diamidino-2-phenylindole (DAPI). Cells were embedded in glycergel mounting media (C0563, Dako North America, Carpinteria, CA, USA) and 25 mg/ml DABCO (1,4-diazabicyclo [2.2.2] octane; 33480, Fluka, Buchs, Switzerland). Specimens were analyzed using a confocal laser scanning microscope (SP2, Leica, Microsystems, Wetzlar, Germany).\n\nFor super-resolution imaging, DAPI staining was avoided, Alexa 532-conjugated antibodies (goat anti mouse, 1:500) were used as secondary antibodies for Mab414, and Alexa 488 as secondary antibodies for VP16. Cells were mounted with ProLong Gold Antifade Reagent (P36930 Thermo Fisher). Images were acquired with a TCS SP8 gSTED 3X microscope (Leica Microsystems, Wetzlar, Germany), which allows, in addition to standard confocal microscopy, the use of the gated STED (gSTED) principle to perform imaging beyond the diffraction limit. An HC PL Apo STED White 100x/1.4NA oil objective was used to obtain super resolved images with a final pixel size of 20 nm. The nuclear pores were excited using a super continuum white light laser (WLL) at a wavelength of 532 nm, depleted with a STED laser beam at 660 nm and detected with hybrid detectors adapted for time gated imaging (applied time gate: 1.5 – 7 ns). For analysis, the images were deconvolved employing the deconvolution algorithm of the program suite Huygens Professional version 18.04 (SVI, Hilversum, The Netherlands).\n\nNuclei of Vero cells are triaxial ellipsoids. Therefore, the mean nuclear volume (Vn) and mean nuclear surface area (Sn) were calculated on the basis of the half axes (a, b, c) measured on 25 deconvolved confocal images of DAPI stained nuclei as described in detail recently (Sutter et al., 2012). Capsids within nuclei were counted on TEM images selected at random at 16 hpi). Then the nuclear area was estimated by point counting applying a multipurpose test system (Weibel, 1979). The mean nuclear area (An) was calculated using the equation An = Pn·d2, whereby Pn are points hitting the nuclei and d the test line length. Capsids were counted on nuclear profiles. From the number of capsids (c) and the nuclear area, the numerical density NVc = c/(An)/D can be calculated, whereby D is the mean particle diameter: D =125 nm for capsids (Zhou et al., 1998). Then, the total number of capsids (Nc) per mean nuclear volume can be calculated: Nc= NVc·Vn. The mean number of RCs was expressed per nuclear profile because the true size of RCs cannot be measured accurately. Diameters of nuclear pores visualized by cryo-FESEM imaging were measured using the AnalySIS (version 5) Five software (Olympus, Hamburg, Germany). The number of nuclear pores were counted and expressed per 1 µm2 nuclear area and calculated per the mean nuclear surface obtained from confocal images. The number of nuclear pore complexes (NPC) was determined on Mab414 stained nuclei using AnalySIS Five (Olympus).\n\nTo determine changes in nuclear membranes arising during R7041(∆Us3) infection, and the amount of membranes used for envelopment during budding, images were collected at a final magnification of 87500x. On these images the surface density of membrane folds (Svf) and of the viral envelope in the PNS (Sve) were estimated using the equations Svf,,e = 4If,e/d·Pn, whereby If,e are the number of intersections of the test lines d with membrane folds and viral envelope, respectively. From the surface density, the area of membrane folds and viral envelope were calculated per mean nuclear volume: Sf= Svf·Vn and Se= Sve·Vn. Mean and variance of nuclear pore diameters were compared by the Welch-Test, Mean and standard deviation of all data by a multiple t-test using GraphPad Prism version 8.\n\nVero cells were grown in 25 cm2 cell culture flasks. Cells were inoculated with R7041(∆Us3) or wt HSV-1 at a MOI of 5 and incubated at 37°C for 24 h. The protein extraction was accomplished as following. After washing with PBS protein lysis buffer (0.5 M Tris-HCl pH 6.8, 4.4% SDS, 1% β-mercaptoethanol, 20% glycerol, 1% bromphenol blue, H2O) was added, and the samples were boiled for 5 min. 10 µl protein of each sample were separated on 7% SDS-polyacrylamid gel. After electrophoresis at 100 V for 2 h, the proteins were blotted onto a nitrocellulose membrane (10600002, Amersham Biosciences Europe, Freiburg, Germany). Blots were blocked with 5% low-fat milk in PBST20 (50 mM sodium phosphate buffer containing 155 mM NaCl and 0.3% Tween 20) over night. Subsequently, blots were probed with monoclonal mouse antibodies against capsid protein ICP5 (ab6508, Abcam, Cambridge, UK), diluted 1:3200, and polyclonal antibodies against tegument proteins VP16 and VP22 raised in rabbits (gift from B. Roizman), diluted in PBST20 (1:3000 to 1:5000). After two washing steps with PBST20, blots were incubated with horse radish peroxidase-conjugated anti-mouse, diluted 1:10000, (AP124P, Sigma-Aldrich, Buchs, Switzerland) or anti-rabbit secondary antibodies, diluted 1:1000, (GERPN4301, Sigma-Aldrich). Protein bands were visualized on X-ray films using chemiluminescence. For loading control, antibodies were stripped out of the membranes with Restore Western blot stripping buffer (21059, Thermo Fisher Scientific) according to manufacturer instructions. Membranes were probed with monoclonal anti-beta actin antibodies produced in mouse, diluted 1:1000 (SAB1305567, Sigma-Aldrich).\n\n\nResults\n\nTo visualize the nuclear surface in the frozen hydrated state, frozen cells need to be fractured. Fracturing of frozen hydrated cells does not create completely arbitrary surfaces. Rather, fracture planes run preferentially along the hydrophobic center of the lipid bilayer of cell membranes, e.g. along the center of the INM or ONM (Severs, 2007). In cryo-FESEM images, intact nuclear pores appear basically as flat button-like structures at the INM, and as small indentations at the ONM (Figure 1) as described in detail (Wild et al., 2012b) and by many other authors from the early days of introducing the freeze-fracture technique, e.g. (Haggis, 1989; Nicolini et al., 1984; Teigler & Baerwald, 1972). Nuclear pore diameter measures 125 nm in negatively stained frog oocytes (Pante & Aebi, 1996). The diameter of nuclear pores in mock infected Vero cells imaged by cryo-FESEM varies due to changes taking place during preparation and imaging. The NPC can be removed together with the ONM during cryo-fracturing leading to small depressions at the INM. Alternatively, the NPC may slightly protrude into the cytoplasm (Wild et al., 2012b). The average diameter of these small protrusions was 120 nm. Distribution of nuclear pores was irregular (Figure 1C). In wt HSV-1 infected cells, large areas of the nuclear surface were devoid of nuclear pores and of nuclear membrane proteins (Orci & Perrelet, 1975). This was also shown in herpes virus infected BHK-21 cells employing the freeze-fracture technique (Haines & Baerwald, 1976). Most of the nuclear pores appeared similar as in mock infected cells (Figure 2). However, there were large clearly confined holes. Many of the holes contained material protruding into the cytoplasm. TEM analysis revealed that most of these holes were confined by an intact INM turning into the ONM (Figure 3A and B), and that nuclear material containing capsids protruded through the holes into the cytoplasm. In a sole case, the nuclear membranes were obviously disrupted (Figure 3C). We thus conclude that the clearly confined holes are dilated nuclear pores.\n\n(A) Detail of inner nuclear membrane (INM) (i), (B) detail of outer nuclear membrane (ONM) (o) showing nuclear pore complex (NPC) anchored within the nuclear pore. (C) Overview showing in addition to nuclear pores with anchored NPC, pores of which the NPC has been removed (d) as well as pores with protruding NPC (p). Bars 500 nm.\n\nThe nuclear surface of Vero cells imaged 14 hpi with wt HSV-1 by cryo-FESEM showing in panel (A) large areas devoid of nuclear pores (asterisks) at the inner nuclear membrane (INM) (i) and outer nuclear membrane (ONM) (o), holes with (B) or without (C) protruding material, large depressions at the INM (D) and at the ONM (E). Bars 200 nm.\n\nTransmission electron microscopy (TEM) images of dilated nuclear pores at 16 hpi (A to D) and at 12 hpi (E) with wt HSV-1. The outer nuclear membrane (ONM) (o) continues into the inner nuclear membrane (INM) (i) clearly visible (arrows) at least at one side (A, B and D). The nuclear material protruding into the cytoplasm contains capsids (c) indicating that capsids are released into the cytoplasm. Possible breakdown of nuclear membrane (C). The INM is disrupted (di), and the ONM runs towards the cytoplasm just beside an intact nuclear pore (np). Nuclear pore dilated to 170 nm distinctly showing the continuum (arrows) between ONM and INM (E). Bars 200 nm (A, B, C), 100 nm (D, E).\n\nIn cells infected with the deletion mutant R7041(∆Us3), the most striking feature was the irregular nuclear surface showing folds and invaginations (Figure 4). Nuclear pores appeared similar as in mock-infected cells. The number of dilated pores was low. To address frequency and size of pore dilation, we measured nuclear pores on 10 nuclei harvested at 16 h post inoculation (hpi). In mock infected cells, pore diameter ranged between 90 and 140 nm with a few exceptions (Figure 5A). In cells infected with R7041(∆Us3), nuclear pores measured up to 180 nm, and in wt HSV-1 or the Us3 repair mutant R2641 up to 400 nm. The mean pore diameter was significantly larger (p<0.0001) in wt HSV-1 and R2641 infected cells compared to mock or R7041(∆Us3) infected cells whereas it did not differ significantly between R7041(∆Us3) and mock infection. The variance of pore diameter was significantly different (p<0.0001) between all groups except between wt HSV-1 and R2641 infected cells. NPCs disintegrate and, subsequently, nuclear pores dilate in the course of NE breakdown during mitosis (Georgatos et al., 1997; Terasaki et al., 2001). HSV-1 arrests the cell cycle in the G1/S phase and S phase (de Bruyn Kops & Knipe, 1988; Ehmann et al., 2000). Mitotic activity almost completely declines by 6 hpi with wt HSV-1 at a MOI of 5 (Sutter et al., 2012). Therefore, pore dilation in HSV-1 infected cells is likely not related to mitosis. In measuring pore diameter, we only considered nuclei with clear indications of infection such as budding capsids. We thus conclude that i) nuclear pores dilate during HSV-1 infection, and ii) dilation of nuclear pores is facilitated by the Us3 kinase.\n\nThe nuclear surface is folded and invaginated. It contains several nuclear pores of which nuclear pore complex (NPC) has been removed during fracturing (d) or with slightly protruding NPC (p) at the inner nuclear membrane (INM) (i). The outer nuclear membrane (ONM) (o) has been largely removed. Bar 500 nm.\n\n(A) Range of nuclear pore diameter, as well as mean and SD. Level of significance a) p<0.0001 for variance and mean diameter compared to R7041(∆Us3) or mock, b) p=0.89 for the mean diameter compared to mock, n=10. (B) Mean nuclear surface area. (C) Surface area of membrane folds and of the total number of virions present in the perinuclear space (PNS) at 16 hpi with R7041(∆Us3) and HSV-1 that was close to zero. (D) Number of nuclear pores counted on cryo-field emission scanning electron microscopy (cryo-FESEM) images and calculated per mean nuclear surface area or determined on confocal microscopic (CLS) or stimulated emission depletion (STED) images after labeling of pore complexes with Mab414. Level of significance **p<0.01, ***p<0.001 compared to mock; n=10 (STED: n=3).\n\nThe nuclear membranes expand during infection with HSV-1. Expansion of nuclear membranes is out of control in cells infected with Us3 deletion mutants leading to formation of folds and invagination that contain virions (Reynolds et al., 2002; Wild et al., 2012a; Wild et al., 2015; Wisner et al., 2009). We calculated the excessively produced membranes by morphometric analysis. The membrane folds cover an area equaling about 30% of the total nuclear surface area (Figure 5C) at16 hpi with R7041(∆Us3). R7041(∆Us3) virions accumulate in the PNS. The total surface area of membranes needed for envelopment was about 100 µm2 which equals about 20% of the nuclear surface. Nuclear membrane folds form complicated structures. Therefore, we imaged an area of folds of the INM on serial sections employing the FIB-SEM technology. 3D reconstruction revealed that the INM folds to complicated structures enclosing cavities that contain virions (Figure 6). We conclude that excessive production of membrane may counteract the dilation of nuclear pores.\n\nBar 500 nm.\n\nThe number of nuclear pores was counted on cryo-FESEM images of 10 nuclei harvested at 16 hpi. Then, the mean number of nuclear pores was expressed per mean nuclear surface area (Figure 5B) calculated from the three axes measured on confocal images. The total pore number was significantly lower (p<0.001) in cells infected with any virus compared to mock-infected cells (Figure 5C). This was probably due to areas devoid of nuclear pores after infection (Figure 2 and Figure 8). Confocal microscopy of Mab414 stained nucleoporins revealed also statistically significant lower numbers of (p<0.001) NPCs after infection with R7041(∆Us3), wt HSV-1 or R2641 compared to the pore number in mock infected cells (Figure 5C). To ascertain whether resolution power of confocal microscopy was sufficient for accurate determination of pore numbers (Wild et al., 2009), we visualized nuclear pore distribution by gSTED (Figure 7). Quantitation of nuclear pores on 3 nuclei per group revealed that the mean total number of nuclear pores per mean nuclear surface was almost equal after R7041(∆Us3) infection, 12% lower after wt HSV-1 infection but 5% higher in mock infected cells (Figure 5C). We assume that pore numbers were overestimated to some extent in SEM images. However, determination of pore numbers in confocal images and STED images is considered very likely to result in some underestimation. Minimal interpore distance was less than 30 nm after wt HSV-1 infection, and less than 7 nm in mock infected cells (Wild et al., 2009). Lateral resolution of STED is 50 nm. Nuclei and, consequently, the nuclear surface area expand during HSV-1 infection (Simpson-Holley et al., 2005; Sutter et al., 2012). Phospholipid biosynthesis is induced by HSV-1 contributing to nuclear membrane enlargement (Sutter et al., 2012). From the nuclear surface devoid of pores and nuclear membrane proteins we conclude that nuclear membranes enlarge upon HSV-1 infection, but pore formation is delayed or inhibited, and insertion of host cell membrane proteins ceased. Instead viral proteins are inserted leading to fundamental changes of nuclear membranes protein composition (Johnson & Baines, 2011).\n\nStimulated emission depletion (STED) images of nuclei at 16 hpi with wt HSV-1 (A), R7041(∆Us3) (B) or mock (C). The same is displayed at higher magnification in panel D. Nuclear pores (green), imaged by STED, are labeled with Mab414 and Alexa 532 as secondary antibody. The viral protein VP16 (red) was labeled with a polyclonal antibody and Alexa 488 as secondary antibody. In the mock infected cells (C) a region devoid of pores, due to the uneven surface of a nucleus, can be observed (asterisk). Note the focal distribution of VP16 after R7041(∆Us3) infection. Bar 1µm.\n\nCapsids overcome the nucleocytoplasmic barrier by budding at the INM acquiring tegument and envelope. The result is a fully enveloped virion in the PNS (Figure 8 and Figure 10). In cryo-FESEM images, budding capsids appear as spheres covered with bright dots (probably representing spikes) at the INM or in the PNS. They look like bulges when they are covered by the ONM, as clearly apparent when the ONM is partially removed. During budding, the capsid pushes the INM into the PNS whilst at the periphery the INM is pulled behind the budding capsid for fission to give rise of a virion with an electron dense envelope located in a deep indentation (Figure 8) readily seen on the INM in cryo-FESEM images. The indentations at the ONM could be interpreted as late stages of fusion of the viral envelope with the ONM after release of capsid and tegument into the cytoplasm (Skepper et al., 2001). However, the process at the ONM takes place even in the absence of fusion proteins gB/gH (Farnsworth et al., 2007) discussed in detail by (Wild et al., 2018). Therefore, this process is budding rather than fusion, and hence, the indentations represent initial stages of budding capsids from the cytoplasm into the PNS. The phenotypes of the virus translocation across the ONM shows all characteristics of budding (Figure 9) as discussed in detail (Leuzinger et al., 2005; Wild et al., 2005; Wild et al., 2012b). The budding process at the ONM was described for the first time in baby hamster kidney cells infected with herpes simplex virus strain H4 (Darlington & Moss, 1968).\n\nCryo-field emission scanning electron microscopy (cryo-FESEM) images of the nuclear surface at 12 hpi with wt HSV-1 demonstrating in (A) budding capsids (bo) under the outer nuclear membrane (ONM) (o) and at two sites where the ONM has been focally removed (b) during fracturing, as well as dilated nuclear pores (d) with or without protruding material. (B) shows a virion (v) in the Perinuclear space (PNS), and a dilated nuclear pore (d) at the inner nuclear membrane (INM) (i) and at the ONM both occupied by protruding material. Bars 200 nm.\n\n(A) Scanning electron microscopy (SEM) image of 2 budding capsids at the outer nuclear membrane (ONM) (o) close to normal (p) and dilated (d) pores. The ONM forms folds (arrows) that derive by the capsids being pushed towards the perinuclear space (PNS) whilst the ONM is forced to cover the capsids. The space between ONM and capsid is filled with tegument (t) (B) Transmission electron microscopy (TEM) image of budding capsid at the ONM (o). Note the curvature (thick arrow) that is typical for budding, the tegument (t), which starts to be deposited between the budding front and the capsid, and the sharp bending of the ONM, which turns into the viral envelope that contains the budding proteins UL31 and UL34. Bars 100 nm.\n\n(A) RCs (replication centres) contain hundreds of B-capsids (with scaffold) and C-capsids (with DNA) as well as a few A-capsids (empty). Virions (v) accumulate in invaginations of the inner nuclear membrane (INM) (i) which had formed multiple folds. (B) A-, B- and C-capsids are scattered throughout the nucleus. Two C-capsids (bC) bud at the membrane of an invagination. Bars 500 nm.\n\nFor release into the cytoplasm, capsids need to be transported from the RCs to the nuclear periphery. UL31 and UL34 have been shown to be responsible for intranuclear capsid transportation (Simpson-Holley et al., 2004). Function of UL31 and UL34 depends on phosphorylation by the Us3 kinase (Mou et al., 2009; Ryckman & Roller, 2004). Therefore, we compared the number of RCs and of capsids (Figure 10), including A-capsids (empty capsids), B-capsids (scaffold containing capsids), and C capsids (DNA containing capsids) (Tandon et al., 2015) in 10 randomly selected Vero cells infected with R7041(∆Us3) or wt HSV-1 from 5 independent experiments. In R7041(∆Us3) infected cells, the mean number of RCs per nuclear profile was 0.3 (±0.3) at 9 hpi, and 4.5 (±0.8) at 24 hpi (Figure 11). In contrast, the number of RCs was lower than 1 per nuclear profile at any time point after inoculation with wt HSV-1 or the repair mutant R2641. The number of intranuclear capsids dispersed throughout the nucleus was significantly higher at any time point after infection with R7041(∆Us3) compared to wt HSV-1 or the repair mutant R2641 (Figure 12A), reaching a maximum of 10,500 (±2250) per mean nuclear volume at 24 hpi with R7041(∆Us3). Interestingly, the number of wt HSV-1 capsids increased up to 3,500 (±800) by 16 hpi, and remained constant thereafter. The higher number of RCs and capsids in R7041(∆Us3) infected cells may be due to enhanced assembly or inhibited release into the cytoplasm. Therefore, we harvested virus particles from cell cultures at 24 hpi for immunoblotting. Western blots probed with monoclonal antibodies against the capsid protein ICP5 did not reveal any obvious differences between R7041(∆Us3) and wt HSV-1 (Figure 13). We thus assume that the higher number of RCs and of capsids is more likely the result of impeded release than of enhanced synthesis and assembly.\n\nLevel of significance: **p<0.001, ***p<0.0001 compared to wt HSV-1, n=5.\n\nMean number and standard deviation of capsids in randomly selected nuclei in R7041(∆Us3), R2641 or wt HSV-1 infected cells calculated per mean nuclear volume (A), and of capsids in the cytoplasm including capsids free in the cytoplasmic matrix and capsids budding at Golgi membranes. Level of significance: **p<0.001, ***p<0.0001 compared to wt HSV-1 or repair mutant R2641, n=5.\n\nBeta actin staining served as loading control.\n\nCapsids gain access to the cytoplasmic matrix via impaired NE (Borchers & Ozel, 1993; Grimm et al., 2012; Klupp et al., 2011; Leuzinger et al., 2005; Maric et al., 2014; Schulz et al., 2015; Wild et al., 2005; Wild et al., 2012b) that starts by impairment of nuclear pores. Impairment of nuclear pore is shown in Figure 3. Capsids are not released from the PNS into the cytoplasmic matrix in the absence of Us3 (Reynolds et al., 2002; Wisner et al., 2009). Nonetheless, quantitative electron microscopic analysis revealed R7041(∆Us3) capsids in the cytoplasm though at a reduced number compared to wt HSV-1 (Wild et al., 2015). As shown in Figure 12B, the number of R7041(∆Us3) capsids including those in the cytoplasmic matrix and in the process of budding at membranes was significantly lower compared to wt HSV-1 capsids at any time after 12 hpi. It was postulated that gB is not phosphorylated in the absence of the Us3 gene, and hence, the viral envelope cannot fuse with the ONM. (Wisner et al., 2009). More important, the viral envelope does not fuse with the ONM at all as discussed above. Therefore, we conclude that capsids gain access to the cytoplasm via impaired nuclear pores, and that R7041(∆Us3) capsid release declines due to reduced nuclear pore impairment.\n\n\nDiscussion\n\nHSV-1 replicates in the nucleus and radically alters nuclear architecture including formation of RCs, nuclear expansion and disruption of the nuclear lamina (Simpson-Holley et al., 2005). The nuclear surface expands from ~400 µm2 to ~500 µm2, so that ~100 µm2 of membrane area must be inserted into the INM and the same amount into the ONM within the first 9 h of infection (Sutter et al., 2012). The required phospholipids are supplied by de novo biosynthesis. Translocation of capsids from the nucleus into the cytoplasm starts at about 8 hpi. Release of capsids by budding at the INM requires additional membranes. R7041(∆Us3) capsids bud at the INM, the resulting virions, however, cannot be transported out of the PNS (Figure 14). Morphometric analysis revealed that ~2400 virions accumulate in the PNS of a single cell by 24 hpi (Wild et al., 2012a; Wild et al., 2015) that means 98% of all enveloped R7041(∆Us3) virions produced. The diameter of a virion is 200 nm (Grünewald et al., 2003). Therefore, the surface area of 2400 virions equal an area of ~300 µm2. If the idea of de-envelopment by fusion of the viral envelope with the ONM was correct (Skepper et al., 2001) the same amount of membranes used for budding at the INM would have to be inserted into the ONM. The crux is that close to 2 capsids bud simultaneously per 1 µm2 nuclear surface at 10 hpi (Wild et al., 2009) demanding high dynamics in maintenance of nuclear membrane integrity. Us3 kinase down regulates phospholipid biosynthesis (Wild et al., 2012a). In the absence of the Us3 gene, the INM forms multiple folds, invaginations and evaginations due to excess biosynthesis of phospholipids. Therefore, we speculate that dilation of nuclear pores is provoked by the high demand of membranes for envelopment of budding wt HSV-1 capsids, and that pore dilation is largely prevented when membranes are over produced in the absence of the Us3 gene.\n\n(A) In pathway 1, wt HSV-1 virions derived by budding at the INM are intraluminally transported from the perinuclear space (PNS) via ER-to-Golgi transitions or via ER-Golgi intermediate compartments (ERGIC), the kiss and run mechanism, into Golgi cisternae for packaging into transport vacuoles that delivers virions to the plasma membrane for exocytotic release into the extracellular space. In pathway 2, capsids released via impaired nuclear pores (NP) either bud at Golgi or vacuolar membranes into Golgi cisternae or vacuoles or are wrapped by Golgi membranes or endosomal membranes. Wrapping means budding at membranes concomitantly forming the viral envelope and the vacuolar membrane. The result is a concentric vacuole containing a single virion. (B) In the absence of the Us3 gene, virions cannot be released from the PNS possibly because the intraluminal transportation route is impaired. Nevertheless, Us3 deletion mutants are infective. Nucleus to cytoplasm capsid translocation via impaired nuclear pores is inhibited. Budding into Golgi cisternae and wrapping at Golgi membranes is inhibited. 98% of enveloped virions have been shown to locate in the PNS (Wild et al., 2015).\n\nRecently, it was reported that the endosomal sorting complexes required for transport III (ESCRT-III) is responsible for scission of the viral envelope form the INM (Arii et al., 2018). ESCRT-III also is involved in maintaining INM integrity by downregulating excess INM. Interestingly, the depletion of ESCRT-III proteins induced aberrant INM proliferation in uninfected cells. In HSV-1 infected cells, virions accumulated between nuclear membranes in a similar fashion as Us3 deletion mutants. However, it has to be borne in mind that virions may accumulate in the PNS per se late in infection (Leuzinger et al., 2005). Knockdown of CD98 heavy chain and its binding partner β integrin induced invaginations of the INM that contained HSV-1 virions (Hirohata et al., 2015) resembling the phenotype of infection with Us3 deletion mutants. The question thus arises whether Us3 exerts its regulatory effect on phospholipid biosynthesis via CD98 heavy chain and/or its binding partner β integrin and/or ESCRT-III.\n\nBreakdown of the NE in the course of herpes virus infection was reported in cells infected with Simian agent 8 (Borchers & Ozel, 1993), PrV UL31 and UL34 recombinants (Grimm et al., 2012; Klupp et al., 2011; Schulz et al., 2015) and HSV-1 (Maric et al., 2014). Breakdown of the NE was considered likely to be related to dilation of nuclear pores after infection with bovine herpes virus 1 (Wild et al., 2005) and HSV-1 (Leuzinger et al., 2005). Nuclear pore dilation and nuclear membrane breakdown requires careful examination, preferably on serial sections through cells prepared by rapid freezing followed by freeze-substitution to keep membranes in place (Wild et al., 1997), to prevent loss of lipids (Weibull et al., 1984) and to improve both temporal and spatial resolution (Mueller, 1992). Images taken from such prepared cells clearly demonstrate the difference between pore dilation and membrane breakdown (Figure 3). Interestingly, true membrane rupture was found to be restricted to the INM. Another technique to visualize nuclei in the closest natural state is microscopy of cells in the frozen hydrated state. Cryo-FESEM revealed that the holes in the nuclear surface were clearly demarcated indicating that these holes are not the result of accidental ruptures of the INM and ONM. They are very likely dilated nuclear pores that can enlarge leading to large impaired areas of the nuclear envelope as shown for HSV-1 and BoHV-1 (Leuzinger et al., 2005; Wild et al., 2005). This is in line with the statement that the initial microscopically visible event in breakdown of the nuclear envelope is dilation of nuclear pores in cells undergoing meiosis (Terasaki et al., 2001).\n\nIt was reported that disassembly of the nuclear lamina is required, and that membrane disruption is driven by microtubules (Georgatos et al., 1997). Disassembly of the lamina is induced by HSV-1 enabling successful budding of capsids at the INM (Scott & O’Hare, 2001; Simpson-Holley et al., 2005). Lamina disassembly depends on Us3 (Bjerke & Roller, 2006; Mou et al., 2007), UL31 and UL34 (Reynolds et al., 2004; Simpson-Holley et al., 2005) suggesting that the Us3 kinase might be involved in NE breakdown. However, lamina disassembly is not established to cause disruption of the NE (Prunuske et al., 2006). Alternatively, breakdown of the NE has been proposed to start by disassembly of NPCs. As a consequence, nuclear pores are destabilized and expand (Terasaki et al., 2001). Therefore, we postulated that nuclear pores dilate leading to disruption when infection proceeds.\n\nNuclear pores are formed and NPCs are assembled during mitosis as well as in the interphase (Doucet & Hetzer, 2010). Despite nuclear expansion in the interphase, nuclear pore number remains constant (Doucet & Hetzer, 2010). However, the number of nuclear pores was reduced after both wt HSV-1 (Wild et al., 2009) and R7041(∆Us3) infection. The nucleoporin Nup153 was shown to be down regulated in HSV-1 infected cells (Ray & Enquist, 2004) whereas the cellular levels of major nucleoporins remained unchanged (Hofemeister & O’Hare, 2008; Wild et al., 2009). Nuclei expand during HSV-1 infection (Simpson-Holley et al., 2005; Sutter et al., 2012) and during R7041(∆Us3) infection (Wild et al., 2012a). Expanding nuclei require membrane constituents for enlargement of the NE. HSV-1 has been shown to induce biosynthesis of phospholipids which are incorporated into nuclear membranes. The nuclear surface area increases by about 100 µm2 within 12 hpi with wt HSV-1 (Sutter et al., 2012). The NE is a double coat. The total area of newly produce membranes equals 200 µm2. The INM also provides membranes for envelopment by budding. The total requirement of membrane constituents in HSV-1 infection is reflected by the incorporation of [3H]-choline, which was twice as high by 12 hpi compared to controls. Cellular proteins are embedded in nuclear membranes. These proteins are readily visible in fracture planes of the INM (Orci & Perrelet, 1975). Protein composition is drastically altered after HSV-1 infection, and cellular proteins are largely replaced by viral proteins (Johnson & Baines, 2011). After infection with any of the viruses, large areas of the INM were devoid of proteins as was described also in another study employing freeze-fracture technique (Haines & Baerwald, 1976). We thus conclude that the parts of the INM devoid of cellular proteins are the result of de novo synthesized phospholipids induced by HSV-1. Areas of the NE devoid of cellular proteins were also devoid of nuclear pores. The mean interpore area was almost as twice as large after HSV-1 infection. The maximal interpore area was even 10 times larger after HSV-1 infection compared to mock infection (Wild et al., 2009). Nuclear pore formation is induced by nucleoporins in the course of NPC assembly (Fichtman et al., 2010; Prunuske & Ullman, 2006) in cells undergoing mitosis. Areas of the NE devoid of cellular proteins and nuclear pores, hence, suggest that formation of nuclear pores ceased after infection with HSV-1.\n\nR7041(∆Us3) capsids are retained within the nucleus to a much larger extent than wt HSV-1 capsids. RCs, the site of capsid assembly, are surrounded by a chromatin layer, which becomes reorganized in wt HSV-1 infection, enabling spread of capsids to the nuclear periphery. There they gain direct access to the INM because the nuclear lamina underlying the INM is disrupted (Scott et al., 2001; Simpson-Holley et al., 2004). In the absence of UL31 and UL34, reorganization of the chromatin layer around RCs and disruption of the nuclear lamina does not take place. Us3 kinase functions in association with phosphorylation of UL31/UL34 (Poon et al., 2006; Reynolds et al., 2001; Reynolds et al., 2002; Simpson-Holley et al., 2004) that led to the suggestion that in the absence of Us3 translocation of capsids to the cytoplasm is impeded. Recently, it was shown that UL31 and UL34 are the proteins responsible for budding of capsids at the INM (Bigalke & Heldwein, 2015; Bigalke & Heldwein, 2016; Hagen et al., 2015). This raises the question, whether the idea that the inability of phosphorylation of UL31/UL34 (Reynolds et al., 2001; Reynolds et al., 2002) is the cause for inhibited nucleus to cytoplasm capsid translocation. The discrepancy is that capsids of UL31/UL34 deletion mutants cannot bud whereas virions of Us3 deletion mutants accumulate in the PNS. Phosphorylation of gB was also claimed to be responsible for release of HSV-1 virions out of the PNS via de-envelopment because it enables gB to act as fusion protein (Wisner et al., 2009). On the other hand, gB deletion mutants are not retained in the PNS indicating that gB is not important for virion release at all (Farnsworth et al., 2007; Klupp et al., 2008). These conflicts can be explained by the erroneous interpretation of the virus transportation across the ONM to be fusion e.g. (Mettenleiter et al., 2013) ignoring the fundamentals of membrane bound transportation (Bonifacino & Glick, 2004; Hughson, 1999; Imai et al., 2006; Jahn et al., 2003; Kanaseki et al., 1997; Leabu, 2006; May, 2002; Mayer, 2002; Orci et al., 1981; Peters et al., 2004; White, 1992). The process shows all characteristics of budding. It takes place even in the absence of the fusion proteins gB/gH as obvious in Figure 2 in (Farnsworth et al., 2007) leading to accumulation of virions in the PNS. Therefore, phosphorylation of gB, UL31 and UL34 does not play any role either in budding of capsids at the INM nor in release of virions via interaction with the ONM as was suggested for PrV UL34 (Klupp et al., 2001). Us3 rather plays a significant role in intraluminal transportation (Figure 13) of virions from the PNS into the ER (Schwartz & Roizman, 1969) and finally into Golgi cisternae (Leuzinger et al., 2005; Wild et al., 2018) in addition to its function in regulation of phospholipid-biosynthesis (Wild et al., 2012a) and apoptosis (Benetti & Roizman, 2004; Benetti & Roizman, 2007).\n\n\nConclusion\n\nHSV-1 induces severe alterations in the nuclear architecture and at the nuclear periphery enabling capsid release via budding at the INM or via distortion of nuclear pores leading to breakdown of the nuclear envelope. Us3 kinase plays a significant role in alterations of the NE considering regulation of biosynthesis of phospholipids induced by HSV-1. Further investigations are needed to elucidate mechanisms leading to alterations of the NE, to understand their impact on HSV-1 envelopment, and possibly on diverse cellular functions since the NE plays other crucial roles (Wilson & Berk, 2010) in addition to maintaining the nucleocytoplasmic barrier for controlling nuclear import and export.\n\n\nData availability\n\nUnderlying data is available from Figshare\n\nFigshare: Dataset 1. Nuclear envelope impairment is facilitated by the herpes simplex virus 1 Us3 kinase, https://doi.org/10.6084/m9.figshare.7586153 (Wild et al., 2019)\n\nLicense: CC BY 4.0",
"appendix": "Grant information\n\nThis study was supported by the Foundation for Scientific Research at the University of Zürich, Switzerland.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors thank B. Roizman (The Marjorie B. Kovler Viral Oncology Laboratories, University of Chicago, Illinois, USA) for providing R7041(∆Us3), the repair mutant R2641 and antibodies against VP16 and VP22.\n\n\nReferences\n\nArii J, Watanabe M, Maeda F, et al.: ESCRT-III mediates budding across the inner nuclear membrane and regulates its integrity. Nat Commun. 2018; 9(1): 3379. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenetti L, Munger J, Roizman B: The herpes simplex virus 1 US3 protein kinase blocks caspase-dependent double cleavage and activation of the proapoptotic protein BAD. J Virol. 2003; 77(11): 6567–6573. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenetti L, Roizman B: Herpes simplex virus protein kinase US3 activates and functionally overlaps protein kinase A to block apoptosis. Proc Natl Acad Sci U S A. 2004; 101(25): 9411–9416. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenetti L, Roizman B: In transduced cells, the US3 protein kinase of herpes simplex virus 1 precludes activation and induction of apoptosis by transfected procaspase 3. J Virol. 2007; 81(19): 10242–10248. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBigalke JM, Heldwein EE: Structural basis of membrane budding by the nuclear egress complex of herpesviruses. EMBO J. 2015; 34(23): 2921–2936. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBigalke JM, Heldwein EE: Nuclear Exodus: Herpesviruses Lead the Way. Annu Rev Virol. 2016; 3(1): 387–409. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBjerke SL, Roller RJ: Roles for herpes simplex virus type 1 UL34 and US3 proteins in disrupting the nuclear lamina during herpes simplex virus type 1 egress. Virology. 2006; 347(2): 261–276. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBonifacino JS, Glick BS: The mechanisms of vesicle budding and fusion. Cell. 2004; 116(2): 153–166. PubMed Abstract | Publisher Full Text\n\nBorchers K, Ozel M: Simian agent 8 (SA8): morphogenesis and ultrastructure. Zentralbl Bakteriol. 1993; 279(4): 526–536. PubMed Abstract | Publisher Full Text\n\nBrzozowska A, Lipińska AD, Derewońko N, et al.: Inhibition of apoptosis in BHV-1-infected cells depends on Us3 serine/threonine kinase and its enzymatic activity. Virology. 2018; 513: 136–145. PubMed Abstract | Publisher Full Text\n\nDarlington RW, Moss LH 3rd: Herpesvirus envelopment. J Virol. 1968; 2(1): 48–55. PubMed Abstract | Free Full Text\n\nde Bruyn Kops A, Knipe DM: Formation of DNA replication structures in herpes virus-infected cells requires a viral DNA binding protein. Cell. 1988; 55(5): 857–868. PubMed Abstract | Publisher Full Text\n\nDeruelle MJ, De Corte N, Englebienne J, et al.: Pseudorabies virus US3-mediated inhibition of apoptosis does not affect infectious virus production. J Gen Virol. 2010; 91(Pt 5): 1127–1132. PubMed Abstract | Publisher Full Text\n\nDoucet CM, Hetzer MW: Nuclear pore biogenesis into an intact nuclear envelope. Chromosoma. 2010; 119(5): 469–477. PubMed Abstract | Publisher Full Text\n\nEhmann GL, McLean TI, Bachenheimer SL: Herpes simplex virus type 1 infection imposes a G1/S block in asynchronously growing cells and prevents G1 entry in quiescent cells. Virology. 2000; 267(2): 335–349. PubMed Abstract | Publisher Full Text\n\nEjercito PM, Kieff ED, Roizman B: Characterization of herpes simplex virus strains differing in their effects on social behaviour of infected cells. J Gen Virol. 1968; 2(3): 357–364. PubMed Abstract | Publisher Full Text\n\nFarnsworth A, Wisner TW, Webb M, et al.: Herpes simplex virus glycoproteins gB and gH function in fusion between the virion envelope and the outer nuclear membrane. Proc Natl Acad Sci U S A. 2007; 104(24): 10187–10192. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFichtman B, Ramos C, Rasala B, et al.: Inner/Outer nuclear membrane fusion in nuclear pore assembly: biochemical demonstration and molecular analysis. Mol Biol Cell. 2010; 21(23): 4197–4211. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGeorgatos SD, Pyrpasopoulou A, Theodoropoulos PA: Nuclear envelope breakdown in mammalian cells involves stepwise lamina disassembly and microtubule-drive deformation of the nuclear membrane. J Cell Sci. 1997; 110(Pt 17): 2129–2140. PubMed Abstract\n\nGilbert R, Ghosh K, Rasile L, et al.: Membrane anchoring domain of herpes simplex virus glycoprotein gB is sufficient for nuclear envelope localization. J Virol. 1994; 68(4): 2272–2285. PubMed Abstract | Free Full Text\n\nGranzow H, Klupp BG, Fuchs W, et al.: Egress of alphaherpesviruses: comparative ultrastructural study. J Virol. 2001; 75(8): 3675–3684. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGranzow H, Weiland F, Jöns A, et al.: Ultrastructural analysis of the replication cycle of pseudorabies virus in cell culture: a reassessment. J Virol. 1997; 71(3): 2072–2082. PubMed Abstract | Free Full Text\n\nGrimm KS, Klupp BG, Granzow H, et al.: Analysis of viral and cellular factors influencing herpesvirus-induced nuclear envelope breakdown. J Virol. 2012; 86(12): 6512–6521. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrünewald K, Desai P, Winkler DC, et al.: Three-dimensional structure of herpes simplex virus from cryo-electron tomography. Science. 2003; 302(5649): 1396–1398. PubMed Abstract | Publisher Full Text\n\nHagen C, Dent KC, Zeev-Ben-Mordehai T, et al.: Structural Basis of Vesicle Formation at the Inner Nuclear Membrane. Cell. 2015; 163(7): 1692–1701. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHaggis GH: Freeze-fracture of 3T3 and HeLa cells for high-resolution SEM and deep-etch replicas: structure of the interphase nucleus. Prog Clin Biol Res. 1989; 295: 31–37. PubMed Abstract\n\nHaines H, Baerwald RJ: Nuclear membrane changes in herpes simplex virus-infected BHK-21 cells as seen by freeze-fracture. J Virol. 1976; 17(3): 1038–1042. PubMed Abstract | Free Full Text\n\nHarreveld AV, Fifkova E: Rapid freezing of deep cerebral structures for electron microscopy. Anat Rec. 1975; 182(3): 377–385. PubMed Abstract | Publisher Full Text\n\nHirohata Y, Arii J, Liu Z, et al.: Herpes Simplex Virus 1 Recruits CD98 Heavy Chain and β1 Integrin to the Nuclear Membrane for Viral De-Envelopment. J Virol. 2015; 89(15): 7799–7812. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHofemeister H, O'Hare P: Nuclear pore composition and gating in herpes simplex virus-infected cells. J Virol. 2008; 82(17): 8392–8399. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHughson FM: Membrane fusion: structure snared at last. Curr Biol. 1999; 9(2): R49–52. PubMed Abstract | Publisher Full Text\n\nImai M, Mizuno T, Kawasaki K: Membrane fusion by single influenza hemagglutinin trimers. Kinetic evidence from image analysis of hemagglutinin-reconstituted vesicles. J Biol Chem. 2006; 281(18): 12729–12735. PubMed Abstract | Publisher Full Text\n\nImai T, Arii J, Minowa A, et al.: Role of the herpes simplex virus 1 Us3 kinase phosphorylation site and endocytosis motifs in the intracellular transport and neurovirulence of envelope glycoprotein B. J Virol. 2011; 85(10): 5003–5015. PubMed Abstract | Publisher Full Text | Free Full Text\n\nImai T, Sagou K, Arii J, et al.: Effects of phosphorylation of herpes simplex virus 1 envelope glycoprotein B by Us3 kinase in vivo and in vitro. J Virol. 2010; 84(1): 153–162. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJahn R, Lang T, Südhof TC: Membrane fusion. Cell. 2003; 112(4): 519–533. PubMed Abstract | Publisher Full Text\n\nJerome KR, Fox R, Chen Z, et al.: Herpes simplex virus inhibits apoptosis through the action of two genes, Us5 and Us3. J Virol. 1999; 73(11): 8950–8957. PubMed Abstract | Free Full Text\n\nJohnson DC, Baines JD: Herpesviruses remodel host membranes for virus egress. Nat Rev Microbiol. 2011; 9(5): 382–394. PubMed Abstract | Publisher Full Text\n\nKanaseki T, Kawasaki K, Murata M, et al.: Structural features of membrane fusion between influenza virus and liposome as revealed by quick-freezing electron microscopy. J Cell Biol. 1997; 137(5): 1041–1056. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKato A, Arii J, Shiratori I, et al.: Herpes simplex virus 1 protein kinase Us3 phosphorylates viral envelope glycoprotein B and regulates its expression on the cell surface. J Virol. 2009; 83(1): 250–261. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKato A, Kawaguchi Y: Us3 Protein Kinase Encoded by HSV: The Precise Function and Mechanism on Viral Life Cycle. Adv Exp Med Biol. 2018; 1045: 45–62. PubMed Abstract | Publisher Full Text\n\nKlupp B, Altenschmidt J, Granzow H, et al.: Glycoproteins required for entry are not necessary for egress of pseudorabies virus. J Virol. 2008; 82(13): 6299–6309. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKlupp BG, Granzow H, Mettenleiter TC: Effect of the pseudorabies virus US3 protein on nuclear membrane localization of the UL34 protein and virus egress from the nucleus. J Gen Virol. 2001; 82(Pt 10): 2363–2371. PubMed Abstract | Publisher Full Text\n\nKlupp BG, Granzow H, Mettenleiter TC: Nuclear envelope breakdown can substitute for primary envelopment-mediated nuclear egress of herpesviruses. J Virol. 2011; 85(16): 8285–8292. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeabu M: Membrane fusion in cells: molecular machinery and mechanisms. J Cell Mol Med. 2006; 10(2): 423–427. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeach N, Bjerke SL, Christensen DK, et al.: Emerin is hyperphosphorylated and redistributed in herpes simplex virus type 1-infected cells in a manner dependent on both UL34 and US3. J Virol. 2007; 81(19): 10792–10803. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeopardi R, Van Sant C, Roizman B: The herpes simplex virus 1 protein kinase US3 is required for protection from apoptosis induced by the virus. Proc Natl Acad Sci U S A. 1997; 94(15): 7891–7896. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLeuzinger H, Ziegler U, Schraner EM, et al.: Herpes simplex virus 1 envelopment follows two diverse pathways. J Virol. 2005; 79(20): 13047–13059. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaric M, Haugo AC, Dauer W, et al.: Nuclear envelope breakdown induced by herpes simplex virus type 1 involves the activity of viral fusion proteins. Virology. 2014; 460–461: 128–137. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaric M, Shao J, Ryan RJ, et al.: A functional role for TorsinA in herpes simplex virus 1 nuclear egress. J Virol. 2011; 85(19): 9667–9679. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMay S: Structure and energy of fusion stalks: the role of membrane edges. Biophys J. 2002; 83(6): 2969–2980. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMayer A: Membrane fusion in eukaryotic cells. Annu Rev Cell Dev Biol. 2002; 18: 289–314. PubMed Abstract | Publisher Full Text\n\nMettenleiter TC, Klupp BG, Granzow H: Herpesvirus assembly: a tale of two membranes. Curr Opin Microbiol. 2006; 9(4): 423–429. PubMed Abstract | Publisher Full Text\n\nMettenleiter TC, Muller F, Granzow H, et al.: The way out: what we know and do not know about herpesvirus nuclear egress. Cell Microbiol. 2013; 15(2): 170–8. PubMed Abstract | Publisher Full Text\n\nMou F, Forest T, Baines JD: US3 of herpes simplex virus type 1 encodes a promiscuous protein kinase that phosphorylates and alters localization of lamin A/C in infected cells. J Virol. 2007; 81(12): 6459–6470. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMou F, Wills E, Baines JD: Phosphorylation of the UL31 protein of herpes simplex virus 1 by the US3-encoded kinase regulates localization of the nuclear envelopment complex and egress of nucleocapsids. J Virol. 2009; 83(10): 5181–5191. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMueller M: The integrating power of cryofixation-based electron microscopy in biology. Acta Microscopica. 1992; 1: 37–46.\n\nMunger J, Roizman B: The US3 protein kinase of herpes simplex virus 1 mediates the posttranslational modification of BAD and prevents BAD-induced programmed cell death in the absence of other viral proteins. Proc Natl Acad Sci U S A. 2001; 98(18): 10410–10415. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNicolini C, Vernazza G, Chiabrera A, et al.: Nuclear pores and interphase chromatin: high-resolution image analysis and freeze etching. J Cell Sci. 1984; 72: 75–87. PubMed Abstract\n\nOgg PD, McDonell PJ, Ryckman BJ, et al.: The HSV-1 Us3 protein kinase is sufficient to block apoptosis induced by overexpression of a variety of Bcl-2 family members. Virology. 2004; 319(2): 212–224. PubMed Abstract | Publisher Full Text\n\nOrci L, Montesano R, Perrelet A: Exocytosis-endocytosis as seen with morphological probes of membrane organization. Methods Cell Biol. 1981; 23: 283–300. PubMed Abstract | Publisher Full Text\n\nOrci L, Perrelet A: Freeze-Etch Histology. Springer-Verlag, New York. 1975. Publisher Full Text\n\nPanté N, Aebi U: Molecular dissection of the nuclear pore complex. Crit Rev Biochem Mol Biol. 1996; 31(2): 153–199. PubMed Abstract | Publisher Full Text\n\nPeters C, Baars TL, Bühler S, et al.: Mutual control of membrane fission and fusion proteins. Cell. 2004; 119(5): 667–678. PubMed Abstract | Publisher Full Text\n\nPoon AP, Benetti L, Roizman B: US3 and US3.5 protein kinases of herpes simplex virus 1 differ with respect to their functions in blocking apoptosis and in virion maturation and egress. J Virol. 2006; 80(8): 3752–3764. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrunuske AJ, Liu J, Elgort S, et al.: Nuclear envelope breakdown is coordinated by both Nup358/RanBP2 and Nup153, two nucleoporins with zinc finger modules. Mol Biol Cell. 2006; 17(2): 760–769. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrunuske AJ, Ullman KS: The nuclear envelope: form and reformation. Curr Opin Cell Biol. 2006; 18(1): 108–116. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPurves FC, Longnecker RM, Leader DP, et al.: Herpes simplex virus 1 protein kinase is encoded by open reading frame US3 which is not essential for virus growth in cell culture. J Virol. 1987; 61(9): 2896–2901. PubMed Abstract | Free Full Text\n\nPurves FC, Spector D, Roizman B: The herpes simplex virus 1 protein kinase encoded by the US3 gene mediates posttranslational modification of the phosphoprotein encoded by the UL34 gene. J Virol. 1991; 65(11): 5757–5764. PubMed Abstract | Free Full Text\n\nQuinlan MP, Chen LB, Knipe DM: The intranuclear location of a herpes simplex virus DNA-binding protein is determined by the status of viral DNA replication. Cell. 1984; 36(4): 857–868. PubMed Abstract | Publisher Full Text\n\nRadsak K, Eickmann M, Mockenhaupt T, et al.: Retrieval of human cytomegalovirus glycoprotein B from the infected cell surface for virus envelopment. Arch Virol. 1996; 141(3–4): 557–572. PubMed Abstract | Publisher Full Text\n\nRay N, Enquist LW: Transcriptional response of a common permissive cell type to infection by two diverse alphaherpesviruses. J Virol. 2004; 78(7): 3489–3501. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReynolds AE, Liang L, Baines JD: Conformational changes in the nuclear lamina induced by herpes simplex virus type 1 require genes UL31 and UL34. J Virol. 2004; 78(11): 5564–5575. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReynolds AE, Ryckman BJ, Baines JD, et al.: UL31 and UL34 proteins of herpes simplex virus type 1 form a complex that accumulates at the nuclear rim and is required for envelopment of nucleocapsids. J Virol. 2001; 75(18): 8803–8817. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReynolds AE, Wills EG, Roller RJ, et al.: Ultrastructural localization of the herpes simplex virus type 1 UL31, UL34, and US3 proteins suggests specific roles in primary envelopment and egress of nucleocapsids. J Virol. 2002; 76(17): 8939–8952. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoizman B, Knipe DM, Whitley RJ: Herpes simplex viruses. In: D M, K., P M, Howley (Ed.), Fields Virology. 6th ed. Wolters Kluver/Lipincott Wiliams & Wilkins, Philadelphia, 2014; 1823–1897.\n\nRyckman BJ, Roller RJ: Herpes simplex virus type 1 primary envelopment: UL34 protein modification and the US3-UL34 catalytic relationship. J Virol. 2004; 78(1): 399–412. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchulz KS, Klupp BG, Granzow H, et al.: Herpesvirus nuclear egress: Pseudorabies Virus can simultaneously induce nuclear envelope breakdown and exit the nucleus via the envelopment-deenvelopment-pathway. Virus Res. 2015; 209: 76–86. PubMed Abstract | Publisher Full Text\n\nSchwartz J, Roizman B: Concerning the egress of herpes simplex virus from infected cells: electron and light microscope observations. Virology. 1969; 38(1): 42–49. PubMed Abstract | Publisher Full Text\n\nScott ES, Malcomber S, O'Hare P: Nuclear translocation and activation of the transcription factor NFAT is blocked by herpes simplex virus infection. J Virol. 2001; 75(20): 9955–9965. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScott ES, O'Hare P: Fate of the inner nuclear membrane protein lamin B receptor and nuclear lamins in herpes simplex virus type 1 infection. J Virol. 2001; 75(18): 8818–8830. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSevers NJ: Freeze-fracture electron microscopy. Nat Protoc. 2007; 2(3): 547–576. PubMed Abstract | Publisher Full Text\n\nSimpson-Holley M, Baines J, Roller R, et al.: Herpes simplex virus 1 UL31 and UL34 gene products promote the late maturation of viral replication compartments to the nuclear periphery. J Virol. 2004; 78(11): 5591–5600. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSimpson-Holley M, Colgrove RC, Nalepa G, et al.: Identification and functional evaluation of cellular and viral factors involved in the alteration of nuclear architecture during herpes simplex virus 1 infection. J Virol. 2005; 79(20): 12840–12851. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSkepper JN, Whiteley A, Browne H, et al.: Herpes simplex virus nucleocapsids mature to progeny virions by an envelopment --> deenvelopment --> reenvelopment pathway. J Virol. 2001; 75(12): 5697–5702. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStackpole CW: Herpes-type virus of the frog renal adenocarcinoma. I. Virus development in tumor transplants maintained at low temperature. J Virol. 1969; 4(1): 75–93. PubMed Abstract | Free Full Text\n\nStannard LM, Himmelhoch S, Wynchank S: Intra-nuclear localization of two envelope proteins, gB and gD, of herpes simplex virus. Arch Virol. 1996; 141(3–4): 505–524. PubMed Abstract | Publisher Full Text\n\nSutter E, de Oliveira AP, Tobler K, et al.: Herpes simplex virus 1 induces de novo phospholipid synthesis. Virology. 2012; 429(2): 124–135. PubMed Abstract | Publisher Full Text\n\nTandon R, Mocarski ES, Conway JF: The A, B, Cs of herpesvirus capsids. Viruses. 2015; 7(3): 899–914. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTeigler DJ, Baerwald RJ: A freeze-etch study of clustered nuclear pores. Tissue Cell. 1972; 4(3): 447–456. PubMed Abstract | Publisher Full Text\n\nTerasaki M, Campagnola P, Rolls MM, et al.: A new model for nuclear envelope breakdown. Mol Biol Cell. 2001; 12(2): 503–510. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWeibel E: Stereological methods, Vol.I. Practical methods for biological morphometry. Academic Press, London. 1979. Reference Source\n\nWeibull C, Villiger W, Carlemalm E: Extraction of lipids during freeze-substitution of Acholeplasma laidlawii-cells for electron microscopy. J Microsc. 1984; 134(Pt 2): 213–216. PubMed Abstract | Publisher Full Text\n\nWhealy ME, Card JP, Meade RP, et al.: Effect of brefeldin A on alphaherpesvirus membrane protein glycosylation and virus egress. J Virol. 1991; 65(3): 1066–1081. PubMed Abstract | Free Full Text\n\nWhite JM: Membrane fusion. Science. 1992; 258(5084): 917–924. PubMed Abstract | Publisher Full Text\n\nWild P: Electron microscopy of viruses and virus-cell interactions. Methods Cell Biol. in: Allan, D. (Ed.), Introduction to electron microscopy for biologists. Elsevier, San Diego, USA. 2008; 88. : 497–524. PubMed Abstract | Publisher Full Text\n\nWild P, de Oliveira AP, Sonda S, et al.: The herpes simplex virus 1 US3 regulates phospholipid synthesis. Virology. 2012a; 432(2): 353–360. PubMed Abstract | Publisher Full Text\n\nWild P, Engels M, Senn C, et al.: Impairment of nuclear pores in bovine herpesvirus 1-infected MDBK cells. J Virol. 2005; 79(2): 1071–1083. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWild P, Gabrieli A, Schraner EM, et al.: Reevaluation of the effect of lysoyzme on Escherichia coli employing ultrarapid freezing followed by cryoelectronmicroscopy or freeze substitution. Microsc Res Tech. 1997; 39(3): 297–304. PubMed Abstract | Publisher Full Text\n\nWild P, Käch A, Lucas MS: High resolution scanning electron microscopy of the nuclear surface in herpes simplex virus 1 infected cells. In: Schatten, H. (Ed.), Scanning electron microscopy for the life sciences. Cambridge University Press, New York USA. 2012b; 115–136. Publisher Full Text\n\nWild P, Kaech A, Schraner EM, et al.: Endoplasmic reticulum-to-Golgi transitions upon herpes virus infection [version 2; referees: 1 approved, 2 approved with reservations]. F1000Res. 2018; 6; 1804. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWild P, Leisinger S, de Oliveira AP, et al.: Nuclear envelope impairment is facilitated by the herpes simplex virus 1 Us3 kinase. Figshare. Fileset. 2019. http://www.doi.org/10.6084/m9.figshare.7586153.v2\n\nWild P, Leisinger S, de Oliveira AP, et al.: Herpes simplex virus 1 Us3 deletion mutant is infective despite impaired capsid translocation to the cytoplasm. Viruses. 2015; 7(1): 52–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWild P, Schraner EM, Adler H, et al.: Enhanced resolution of membranes in cultured cells by cryoimmobilization and freeze-substitution. Microsc Res Tech. 2001; 53(4): 313–321. PubMed Abstract | Publisher Full Text\n\nWild P, Schraner EM, Cantieni D, et al.: The significance of the Golgi complex in envelopment of bovine herpesvirus 1 (BHV-1) as revealed by cryobased electron microscopy. Micron. 2002; 33(4): 327–337. PubMed Abstract | Publisher Full Text\n\nWild P, Senn C, Manera CL, et al.: Exploring the nuclear envelope of herpes simplex virus 1-infected cells by high-resolution microscopy. J Virol. 2009; 83(1): 408–419. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWilson KL, Berk JM: The nuclear envelope at a glance. J Cell Sci. 2010; 123(Pt 12): 1973–1978. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWisner TW, Wright CC, Kato A, et al.: Herpesvirus gB-induced fusion between the virion envelope and outer nuclear membrane during virus egress is regulated by the viral US3 kinase. J Virol. 2009; 83(7): 3115–3126. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou ZH, Chiu W, Haskell K, et al.: Refinement of herpesvirus B-capsid structure on parallel supercomputers. Biophys J. 1998; 74(1): 576–588. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "46764",
"date": "18 Apr 2019",
"name": "Haidong Gu",
"expertise": [
"Reviewer Expertise HSV-1 virus-host interaction"
],
"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 manuscript “Nuclear envelope impairment is facilitated by herpes simplex virus 1 Us3 kinase”, the authors used transmission electron microscopy, Cryo-field emission scanning electron microscopy and super resolution confocal microscopy to examine the sizes and numbers of nuclear pores and nuclear membrane surface area in wild type HSV-1 or DUs3 virus infected cells. They conducted quantitative analysis and confirmed several previous observations such as nuclear membrane expands in HSV-1 infected cells, Us3 deletion leads to folds and invaginations of nuclear membrane, and Us3 kinase activity is necessary for capsid release into the cytoplasm, etc. They also found that nuclear pore number decreased HSV-1 infected cells, with more interpore distance measured, compared to that of mock infected cells. The authors generated beautiful EM images to carefully describe nuclear morphological changes that help us understand the HSV-1 egress process. Several issues need to be addressed before the manuscript can be indexed.\nIn Figure 3, the authors demonstrated nuclear envelope impairment in wt HSV-1 infected cells. In Figure 4, they described folds and invaginations of nuclear membrane in DUs3 infected cells. It would be more scientifically accurate if the authors compare both wt and DUs3 in the same TEM experiment, define the descriptive criteria for pore dilation and pore impairment, and then quantitate the pore impairment side-by-side. Via Figures 1 and 2, the authors showed pore size and number in mock and HSV-1 infected cells. They went on to measure and count pore size and number for statistical analysis. However, the scale bars in Figures 1 and 2 are different. Visually it is difficult to compare the pore sizes or to understand what they meant for “large clearly confined holes”. On page 7, three places in parenthesis it says “(Figure 5C)”, but it actually refers to Figure 5D. Proofreading should be done before submission. It causes a great deal of confusion.\n\nFigure 6: what does the red-orange color represent in 6B and 6C. Describe them in the figure legend. Also on page 7, the last sentence of conclusion says “we conclude that nuclear membranes enlarge upon HSV-1 infection, but pore formation is delayed or inhibited, and insertion of host cell membrane proteins ceased”, the last part of “insertion of host cell membrane proteins ceased” is unfounded from the data in this manuscript. On page 12, the authors showed NE impairment (Figure 3) in wt HSV-1 infected cells, and refer to previous publication that capsids are not released from PNS into cytoplasm in the absence of Us3 but capsids are found in the cytoplasm albeit in a number lower than wt infection, and then concluded DUs3 capsids enter cytoplasm via nuclear pore impairment. The part “via nuclear pore impairment” will need more evidence. Also on page 12, higher number of RC and capsids were observed in DUs3 infected cells, compared to wild type infection. How about the extracellular infectious virion production?\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? No source data required\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "49427",
"date": "17 Jun 2019",
"name": "Maria Kalamvoki",
"expertise": [
"Reviewer Expertise HSV-1 host interactions",
"vesicles trafficking",
"innate immunity."
],
"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 work with many novel concepts relevant to HSV-1 primary envelopment.\nA few minor points for clarification in the discussion are listed below:\nThe authors discuss about lipid biosynthesis controlled by Us3 that results in INM production. Is this an event restricted to the nuclear membrane? Can the authors discuss whether there is also increased anterograde vesicular trafficking and whether other organelles including the Golgi network display a similar surface increase? The authors stated that fully enveloped virions are present in the PNS. In case of ΔUs3 virus these virions are fully infectious. Can the authors discuss whether the envelope virions acquire from the INM has difference versus the envelope from the TGN? According to Figure 14 there are enveloped virions that move from Golgi to the plasma membrane without losing the primary envelope. Do the authors know whether there are differences between these virions and the virions that acquire envelope from TGN or vesicular structures? Is it nucleus disassembly and not rupture? Can the authors underscore that point? Do the authors think that ESCRT-III recruitment to the nuclear membrane might have some membrane repair function?\nMinor points to fix in the text: Abstract\nUnder the results \"decrement in pore numbers\" somewhere the authors need to state nuclear pores. Page 7, Figure 5C may be 5D in some occasions. Figures could have a legend to guide people about results they observe since some time is difficult to conclude from the figure.\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": []
}
] | 1
|
https://f1000research.com/articles/8-198
|
https://f1000research.com/articles/7-1489/v1
|
18 Sep 18
|
{
"type": "Research Article",
"title": "E-cadherin expression pattern during zebrafish embryonic epidermis development",
"authors": [
"María Florencia Sampedro",
"María Fernanda Izaguirre",
"Valeria Sigot",
"María Florencia Sampedro"
],
"abstract": "Background: E-cadherin is the major adhesion receptor in epithelial adherens junctions (AJs). On established epidermis, E-cadherin performs fine-tuned cell-cell contact remodeling to maintain tissue integrity, which is characterized by modulation of cell shape, size and packing density. In zebrafish, the organization and distribution of E-cadherin in AJs during embryonic epidermis development remain scarcely described. Methods: Combining classical immunofluorescence, deconvolution microscopy and 3D-segmentation of AJs in epithelial cells, a quantitative approach was implemented to assess the spatial and temporal distribution of E-cadherin across zebrafish epidermis between 24 and 72 hpf. Results: increasing levels of E-cadh protein parallel higher cell density and the appearance of hexagonal cells in the enveloping layer (EVL) as well as the establishments of new cell-cell contacts in the epidermal basal layer (EBL), being significantly between 31 and 48 hpf. Conclusions: Increasing levels of E-cadherin in AJs correlates with extensive changes in cell morphology towards hexagonal packing during the epidermis morphogenesis.",
"keywords": [
"Key words: E-cadherin",
"adherens junctions",
"zebrafish epidermis",
"enveloping layer (EVL)",
"epidermis basal layer (EBL)",
"deconvolution",
"3D segmentation"
],
"content": "Introduction\n\nThe skin is the largest organ of the body in direct contact with the environment. It has a complex structure, being constituted by many different tissues. Skin performs functions that are vital in maintaining body homeostasis, such as the control of body temperature and protection from physical damage and bacterial invasion. Sensory axons innervate the skin at early developmental stages enabling the embryo to sense mechanical, thermal, and chemical stimuli1.\n\nIn particular, the topology –connectivity, continuity and neighborhood–, and individual cell phenotypes of epidermis define the strength and permeability of the protective barrier between the organism and its environment. Although epidermis structure varies between aquatic and terrestrial organisms, its stratification is a common mechanism during development2.\n\nIn zebrafish, the prospective epidermis establishes soon after the end of gastrulation, constituted by the surface layer, enveloping layer (EVL), and the inner epidermal basal layer (EBL)3,4. EVL arises at mid-blastula stage (2.5 hours post fertilization; hpf) covering the whole embryo5. During epiboly EVL maintains tight joins to the yolk syncytial layer (YSL), and becomes the migration substrate for the underlying deep cells spreading during gastrulation6. At the tail bud stage (8–10 hpf), the epiblast forms the outer germ layer, the ectoderm, characterized as a pseudo-epithelial germ layer. From the non-neural ectoderm arises the EBL, which covers the whole embryonic surface underneath the EVL to form a two-layered epithelium by 10 hpf. At 24 hpf, the epidermis is a distinctive bilayer in which the basal layer actively produces collagen to form the basal membrane and the primary dermal stroma7,8. It is after three weeks that the epidermis becomes further stratified and develops into the adult teleost four-layered epidermal structure: cuticle, surface, intermediate and basal stratums9. In adult zebrafish, the EVL cells are replaced by those derived from basal keratinocytes8. Thus, the epidermis of adult zebrafish, as in mice, derives from basal stem cells, further expanding the similarities of epidermal ontogeny across vertebrates10.\n\nE-cadh is member of a superfamily of cadherins, calcium-dependent cell-cell adhesion molecules forming junctions along the apicolateral membranes of adjacent cells11–13. E-cadh plays a key role in determining cell polarity and differentiation, and thereby in the establishment and maintenance of metazoan tissue homeostasis2,14. As epithelia are constituted by cell phenotypes with the maximum polarity and whose identity is primarily specified by E-cadh, it is key to know its expression in the epidermis establishment and maintenance15. Furthermore, due to the mechano-transduction activity coupled to the acto-myosin cytoskeleton remodeling, E-cadh is involved in processes such as cell division orientation in planar polarized epithelia15 and collective cell migration16. Relevant as well is that E-cadh has been characterized as a potent suppressor of invasion and metastasis in epithelia17, which are usually located in direct contact with mutagenic and/or carcinogenic agents responsible for 85–90 % of human cancers18.\n\nIn zebrafish, E-cadh transcripts and proteins are maternally deposited. Reduced levels of the maternal and zygotic protein has been proved to delay epiboly progression with lethal phenotypes5. E-cadh is required for blastomeres adhesion during the cleavage stage and later during gastrulation and epiboly19–21. Indeed, as epiboly proceeds, EVL directs cell migration and the spreading of cells of the deep cells layer (DCL) in a process that requires dynamic cell contacts remodeling mediated by E-cadh6. Once the bi-layered epidermis is established, E-cadh role becomes more refined by keeping its integrity while actively remodeling cell-cell contacts within each layer. At tissue scale, this leads to cell rearrangements which establish regular geometric patterns, and loss of E-cadh results in altered epidermis topology5,14,22. Strikingly, scarce knowledge exists regarding epidermal spatiotemporal expression of E-cadh after epiboly stages. By combining 3D-deconvolution and segmentation of AJs in epidermal cells we were able to obtain a quantitative profile of E-cadh expression during normal epidermis morphogenesis from embryonic-to-larval life of zebrafish.\n\n\nMethods\n\nZebrafish strain of T/AB genetic background was used as wild-type. Male and female adults of 8-months-old were obtained from the Institute of Molecular and Cellular Biology of Rosario (IBR-CONICET-UNR), Argentina, and maintained at 28°C on a 14-h light/10-h dark cycle. Adult fishes were kept in rectangular glass tanks of 12 liters at a density of (1-2 fishes/liter). In each tank, chlorine free water was constantly aerated and filtered (ATMAN hang On filter HF 0100), and renovated by 1/3 twice a week, water temperature was maintained with a heater (Atman 200W). Water pH was kept between 7.8-8.2, salinity was maintained between 350–600 TDS and nitrates were controlled using biological films included in the filtering system. Fishes were fed twice a day with dried flakes (TetraMin) and twice a week with freshly hatched artemia cysts.\n\nAfter breeding, laid eggs were collected and maintained at 28°C. Then, embryos and larvae were staged according to Kimmel et al.23. Around 20 to 25 embryos were collected at 2.5, 18, 24, 31, 48 and 72 hpf, then dechorionated and sedated with buffered tricaine methylsulfonate (MS-222, Sigma) prior to fixation. Approximately, 10–15 fixed embryos per stage were processed for immunofluorescence detection of E-cadh.\n\nAdults and embryos were handled according to the ARRIVE guidelines and to the national guidelines from the Advisory Committee on Ethics of the Facultad de Bioquímica y Ciencias Biológicas de la Universidad Nacional del Litoral, Santa Fe, Argentina (Res. 229 and 388/2006).\n\nAll embryos were fixed in toto in Carnoy solution at room temperature (RT) for at least 2 h and processed according to Izaguirre et al.24. Briefly, they were washed in PBS and permeated in 1% Triton X-100/PBS pH 7.4 for 1 h. Then, washed in PBS pH 7.4 and incubated in normal goat serum (catalogue number: S-1000 Vector Laboratories, Burlingame, CA) for 45 min, followed by overnight incubation with primary antibody anti E-cadh at 4°C, three washes in PBS, and incubation with secondary goat anti-mouse IgG-FITC antibody at RT in darkness for 2 h. Finally, they were rinsed in PBS and mounted in 50% Glycerol-PBS for microscopy imaging. Embryos directly incubated with secondary antibody and normal goat serum, were used as negative controls.\n\nAntibodies. The 36/E-cadh monoclonal antibody recognizes the cytoplasmic domain of human E-cadh, regardless of phosphorylation status (clone 36 mouse IgG2a, catalogue number: 610181 Transduction Laboratories). It was diluted 1:150 and revealed with secondary goat anti-mouse IgG-FITC antibody (Sigma, catalogue number: F8771, St. Louis, MO) used at 1:100 dilution.\n\nThe spatial distribution of E-cadh in zebrafish epidermis was analyzed by fluorescence microscopy followed by image deconvolution and cell segmentation in 3D. The trunk was selected for the ease of orientation and image acquisition within the studied periods. Images were acquired with an inverted wide field sectioning microscope Olympus IX83 coupled to a digital camera CMOS-ORCA-Flash 2.8 (Hamamatsu), and commanded by Olympus Cell Sens software v. 1.13. Raw images were processed using FIJI v. 3.0. Sampling in xy was 0.182 µm with z-step every 0.33 µm. The epidermis was completely scanned along the trunk region. Exposure time was experimentally determined and fixed in order to avoid pixel intensity saturation and to minimize photobleaching.\n\nDeconvolution was applied to restore fluorescence, which improved contrast and z-resolution, enabling better definition of E-cadh in AJs for subsequent application of the 3D-segmentation tool. Quantification of E-cadh fluorescence intensity was carried throughout the epidermis bilayer (~ 6 μm) in calibrated 3D-ROIs set at 2500 µm2 × 0.33 µm × 20 slices (16500 µm3). First, deconvolution was performed on individual 3D-ROI by applying Richardson-Lucy algorithm25 running under the open source Deconvolution Lab 2 v 2.0.0, with a theoretical point spread function26. The Trainable Weka Segmentation Plugin v. 3.1.0, a classification tool based on machine learning in FIJI27 was applied on each deconvolved 3D-ROI so as to create a template that would automatically find the cell boundaries by providing trainable examples of membranes and cytosol (set as background). Each segmented 3D stack was further converted into 8-bit binary 3D-mask and multiplied by the corresponding deconvolved 3D-ROI to obtain the final “Result of Classification”. On each classified image E-cadh fluorescence was quantified as the sum of pixel intensities per 3D-ROI and expressed as raw integrated density (RawIntDen). This measurement was performed on at least six 3D-ROIs per embryo to cover the trunk region, in five embryos per developmental stage. The pipeline for the image processing, theoretical psf and classifier model files are available as Supplementary File 1 as well as an example output (Supplementary File 1, Supplementary File 1).\n\nOn each classified image, 3D-ROIs of fixed volume (10 μm2 × 3 μm deep) were selected along cell-cell contacts in EVL cells and fluorescence intensity was expressed as RawIntDen/cell-cell contact. To assess the fluorescence intensity in individual cells of the EVL, 3D-ROIs were manually outlined along cell perimeters to include the full membrane width and thickness and expressed as RawIntDen/cell area.\n\nCell morphology and cell area were assessed in EVL and EBL cells from previously selected 3D-ROIs. Round, 4-, 5-, 6-, 7- and 8-sided cells were counted using the \"polygon selection tool\" in the individual layers. Mean area was expressed in the image calibrated units. Cell packing index was scored for the EVL and expressed as number of cells/ ROI area (2500 μm2). Area of penta- and hexagonal cells from EBL and EVL were compared for all stages (Supplementary Supplementary File 2).\n\nFive animals in the specified stages were obtained from three to five independent experiments. Differences in E-cadh levels between developmental stages were analyzed using a Linear Mixed Model (LMM). The assumptions of the model were checked graphically (linearity, homoscedasticity, normality of residuals and independence). The non-normality of the data was tested using the Shapiro-Wilk test. The variable \"stage\" was considered as fixed effects (24, 31, 48 and 72 hpf). The random effects of the model are the number of embryos per stage (5) and number of 3D-ROIs per embryo (at least 6). This number of ROIs per embryo was estimated in order to cover >90% of the embryo trunk for the selected stages.\n\nDifferences in mean cellular area for the observed polygon types were analyzed using LMM containing the same fixed and random effects but adding the variable \"morphology\". The same statistical analysis was performed on the data set for the analysis of cell density and expressed as packing index (number of cell / ROI).\n\nThe Tukey's test was used for post-hoc pair-wise comparison when an effect or an interaction was found significant. Significant differences are denoted with *p < 0.05, **p < 0.01. Data were analyzed with RStudio software’s version 1.1.453 and plotted with the BoxPlotR application or InfoStat software version 2018.\n\n\nResults\n\nE-cadh expression pattern was determined in wild type zebrafish during epidermis development from 2.5 (blastula period) to 72 hpf. E-cadh protein was clearly detected in embryos at the blastula stage on epiblast cells (EVL) (Figure 1a). By 18 hpf, during primary organogenesis both epidermis layers are already established. At this stage, E-cadh was observed in AJs in EVL cells and weakly detected in EBL cells (Figure 1b-c). At later stages E-cadh labeling was observed as well as cytoplasmic dots, presumably in endocytic vesicles (Figure 1d).\n\na) at 2.5 hours post fertilization (hpf) b) at 18 hpf; c) Zoom of selection in b); d) at 48 hpf, showing dot labeling in cytoplasm of enveloping layer (EVL) cells ; e–h) 24 to 72 hpf embryos, displaying E-cadh membrane distribution in trunk, clearly visible in underlying EBL cells from 31 hpf. Images are contrast enhanced maximum intensity projections of 50 slices z-step: 0.33 µm. Objectives: UPLFLN 20X 0.75 NA and UPLFLN 40X 1.3 NA oil. Scale bar: 50 µm.\n\nIn embryos at 24 hpf, E-cadh labeling was observed in vertices (puncta adherens) as well as in micro-clusters along lateral cell-cell contacts in the EVL (Figure 1e). At 31 hpf the underlying EBL was barely visible, detected as a faint E-cadh immunolabeling. From 48 hpf onwards, fluorescence in the underlying EBL cells was clearly detected (Figure 1 f-h).\n\nDuring the transition from embryo to larval stages the growing detection of E-cadh along cell-cell contacts parallels noticeable changes in cell size and morphology in the epidermis bilayer (Figure 1, e–h), which were further analyzed.\n\nE-cadh levels were compared in the epidermis layers between stages 24, 31, 48 and 72 hpf, a period during which intense morphogenetic events lead to hatching and significant physiological changes occur for the resulting larvae epidermis to adapt to the aquatic environment. By implementing a 3D-Segmentation algorithm based on machine learning we were able to generate a mask to extract fluorescence intensity values along cell-cell contacts on 3D-ROIs covering the trunk epidermis bilayer and at each developmental stage (Figure 2a–d). With this approach we measured a significant increase of E-cadh levels between 31 and 48 hpf, consistent with the visible detection of E-cadh in the EBL and the visible increase in cell density in the EVL, which was subsequently quantitated.\n\na) 24 hpf and b) 31 hpf, with incipient E-cadh labeling of the underlying epidermis basal layer (EBL) cells; c) 48 hpf; d) 72 hpf, with stronger detection of E-cadh in the EBL (arrow heads). Panels are representative 3D-ROIs classified images shown as sum of intensities in projections of 20-optical section stacks covering the epidermis thickness (6.6 μm); e) Box-plot of means of RawIntDen/ROI per embryo. Objective, 40X, NA 1.3 oil. Statistical significances, ** p<0.01, *p<0.05.\n\nChanges in cell shape, area and density were quantitated in an attempt to correlate the observed increments in E-cadh expression with cell morphology changes characteristics of developing epithelia28. Similar to the other epithelia, the morphogenetic processes leading to epidermis topology development involve cell morphology changes with a predominance of hexagonal geometry in the outermost layer16,24,28. Therefore, the distribution of the cell polygons classes was analyzed in the EVL layer. While round-cells, 4- and 8-side cell polygons were only detected in 24, 31 and 48 hpf stages and represented no more than 7, 15 and 2 % respectively, pentagonal and hexagonal shaped cells predominated in all stages (Figure 3a). Hexagonal cells constituted approximately 34 % of EVL total cells in stage 24 hpf and 45–50% at 31, 48 and 72 hpf with a two-fold increase above pentagonal cells from 31 hpf. Hexagonal cells represented a ~ 50 % of the total cells analyzed consistent with previous reports for other species14.\n\na) Pie chart displaying percentages of x-sided polygons of embryos at 24 hpf (n= 81 cells), 31 hpf (n= 147 cells), 48 hpf (n= 133 cells) and 72 hpf (n= 102 cells). Hexagonal and pentagonal cells are the main cell morphologies observed in the EVL. Data obtained from 30 3D-ROIs in five animals for each stage. b) Mean area of cell types on EVL at 24 hpf (EVL n= 73), 31 hpf (EVL n= 147), 48 hpf (EVL n= 133) and 72 hpf (EVL n= 102).\n\nFor all polygon types assessed in the EVL the mean cell areas decreased within the same polygon type groups from 24 to 72 hpf (Figure 3b). In the EBL, and despite fewer cells were accessible for area measurements, this tendency was not evident (Supplementary Figure 2). Therefore, the visibly increment in cell density in the EVL was characterized by an establishment of hexagonal cell morphology.\n\nWe hypothesized then, that the significant increment of E-cadh levels in the epidermis bilayer between 31 and 48 hpf could reflect a contribution of the appearance of more cell-cell contacts per area, the addition of more protein to cell-cell contacts and the emergent detection of the protein in the underlying EBL. The separation of the EVL as a sub-stack in individual 3DROIs is not completely free from the fluorescence contribution of the underlying EBL layer. Therefore, to evaluate the contribution of individual EVL cells to the observed differences in global fluorescence intensity, this was quantified in individual cell-cell contacts of fixed volume (3D-ROIs, 10 μm2 × 3 μm deep) in the EVL or in individual cells (Figure 4). With this approach overlapping contacts with the underlying EBL cells were excluded from the measurements. Mean RawIntDen values along cell-cell contacts revealed that more E-cadh may populate individual contacts during embryonic epidermis morphogenesis and contribute to the total fluorescence increase, although differences were not significant between developmental stages (Figure 4a). Then, to elucidate whether the increase of the protein between 31 and 48 hpf could be due to the appearance of more cell-cell contacts per EVL area, cell density was estimated and expressed as a cell packing index (Figure 4b). However, this showed a two-fold increase in cell density by 31 hpf without significant change at 48 hpf. Then, fluorescence intensity was quantified in individual cells of the EVL manually outlined from the original 3D-ROIs, and expressed as RawIntDen/cell area (Figure 4c). As expected, the expression of E-cadh in the EVL followed a similar pattern as the one obtained for the global bilayer analysis. However, difference between 31 and 48 hpf was not significant. Together this data suggested that the main contribution to the significant increase in E-cadh in the bilayer is due to the growing detection of the protein in the underlying EBL observed at 48 hpf.\n\na) Box-plot of RawIntDen/ cell-cell contact in the EVL; b) Plot of mean of cell packing index calculated as average number of cells per individual ROIs (2500 µm2). Statistical significances, ** p<0.01; c) Box-plot of means of RawIntDen/average cell area in the EVL.\n\n\nDiscussion\n\nEpithelial architecture retains essential features such as apical/basal cell polarization, formation of cell–cell junctions, and the constitution of a paracellular diffusion barrier, all of which enable epithelia to serve a great diversity of biological functions29. The organization of mature epithelia into packaging of hexagonal shaped cells is a feature evolutionarily conserved14,30. This appears to be the optimum arrangement concerning the transduction of forces with minimum energy costs. The role of E-cadh in molecular architecture of epithelia has been studied extensively in animal tissues, partially owing to its decisive play in human cancers31. In frog and fish embryos, as in others, cadherins are the main adhesion factors responsible for regulating the shape of the embryo and its role has been thoroughly described during epiboly and gastrulation32–34. On established animal epidermis, E-cadh performs fine-tuned cell-cell contact remodeling to maintain tissue integrity while the body axis elongates, this is characterized by modulation of cell shape, size and density to achieve the stable hexagonal arrangement28,35,36.\n\nIn zebrafish, the prospective epidermis is established at 24 hpf as a bilayer formed by the enveloping layer (EVL) and the epidermal basal layer (EBL). Despite numerous descriptions about E-cadh role in epiboly and gastrulation, there is scarce information about E-cadh distribution in epidermis beyond this stage and during the embryo to larval transition. During this period body axis elongates from 1 mm at 24 hpf to 3.5 mm at 72 hpf and embryos undergo hatching asynchronously between 48 and 60 hpf23. Once in direct contact with water, the embryonic epidermis is the main protective barrier against pathogens. Therefore, we find it relevant to study the spatiotemporal distribution of E-cadh in zebrafish and elucidate a relationship between E-cadh levels, cell morphology and cell density in the epidermis bilayer from “embryos” to “larvae”.\n\nWe implemented a trainable 3D-segmentation tool in FIJI37 to extract fluorescence intensity values from epithelial cells in in toto immunolabeled epidermis. Global expression was estimated from these 3D-segmentation volumes excluding cytoplasm (E-cadh in endosomal or in reticulo-endoplasmic compartments) in the bi-layered epidermis from embryo to larval stages. At present, only one pipeline method was reported for segmentation and tracking of epithelial cells based on the detection of the AJs in voxels in the Drosophila notum and leg epithelium38.\n\nAt the membrane level, E-cadh protein concentrates in clusters detected as puncta adherens in cell vertices or as lateral micro-clusters at 24 hpf, that turns into a continuous belt structure from stage 31 hpf onwards, when intracellular E-cadh is also frequently observed in the cytoplasm of EVL cells.\n\nIntensity based analysis showed that growing levels of E-cadh along cell-cell contacts during zebrafish epidermis development correlate with cell morphology changes towards hexagonal geometry. Specifically, within a short period between 24 and 31 hpf, a ~two-fold increase in cell density parallels the appearance of penta- and hexagonal cells together representing ~75 % of the polygons classes similarly to other animal models14.\n\nGlobal bilayer analysis of E-cadh fluorescence intensity revealed a significant increase in protein expression between 31 and 48 hpf. In an attempt to establish the contribution of individual layers to the observed difference, the outermost EVL layer was analyzed by selecting individual cells or cell-cell contacts without the interference of the fluorescence coming from the basal layer.\n\nMean E-cadh levels measured in fixed cell-cell contact volumes of EVL cells showed a steady increase from 24 to 72 hpf, but without significant differences between stages. When E-cadh levels were estimated per average cell area in the EVL, a visible increase was observed between 31 and 48 hpf, although non-significant, indicating that the emergent detection of E cadh in the EBL from 31 hpf onwards may indeed contribute to the significant increase in the protein levels measured in the epidermis bilayer.\n\nWe cannot overlook that during this embryonic period characterized body elongation, either cell proliferation or adherens junctions remodeling, through active E-cadh trafficking could account for the observed increase in cell density in the epidermis39,40. These mechanisms has been well described for the hexagonal cell packing in the developing D. melanogaster pupal wing under polarized trafficking of E-cadh28 and should be further analyzed in vivo within this period in zebrafish.\n\nAs a novel outcome, a recent report proposed that tension generated by the E-cadh/AmotL2/actin filaments complexes plays a crucial role in developmental processes such as epithelial geometrical packing as well as generation of forces required for blastocyst hatching both in mouse and human35. In zebrafish, the hatching process is the result of combined enzymatic digestion of the chorion41 together with mechanical forces that drive the embryo out of the yolk sac in a similar way as observed for mouse and human blastocyst hatching. In the present work, a peak in E-cadh membrane level was detected in the epidermis around the time of spontaneous hatching. Therefore, it is conceivable, that cell morphology remodeling leading to hexagonally packed geometry followed by a significant increase in E cadherin may achieve high cell surface compactness and stiffness of the epidermis required for efficient mechanical disruption of the chorion.\n\n\nConclusions\n\nThe presented results show that during the establishment of embryonic epidermis in zebrafish growing levels of E-cadh protein correlates with increased cell density in the EVL and the establishments of new cell-cell contacts in the EBL more significantly between 31 and 48 hpf. This differentiated and compact epidermal tissue is most likely to support mechanical stress prior to hatching which starts around 48 hpf when the embryo contacts for the first time the aquatic environment directly.\n\nThe combination of classical immunofluorescence, image deconvolution with intensity based segmentation in 3D offers a powerful tool to study the spatial arrangement of cell-cell adhesion proteins and cell morphology in bi-layered epithelia that can be applied to cadherin morphants, to other species or processes such as wound healing and re-epithelialization of the skin.\n\n\nData availability\n\nDataset 1: Raw images for Figure 1 for 2.5 and 18 hours post fertilization (hpf). These can be viewed using FIJI or ImageJ 10.5256/f1000research.15932.d21781942\n\nDataset 2: Raw and processed images of 3D-ROIs for assessing RawIntDen, cell areas and cell morphology data for 24, 31,48 and 72 hours post fertilization (hpf) for Figure 2a and Dataset 3. Raw can be viewed using FIJI or ImageJ 10.5256/f1000research.15932.d21782043\n\nDataset 3: RawIntDen, cell areas and cell morphology counts for Figure 2–Figure 4 10.5256/f1000research.15932.d21782144",
"appendix": "Grant information\n\nThis work was supported by the National University of Entre Ríos, Argentina [PID-UNER 6145].\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\nWe thank Juan Ignacio Etchart for technical microscopy assistance; Sebastián Graziatti, from IBR-CONICET-UNR, for valuable advice in zebrafish husbandry; and Luciana Erbes and Angel Zeitoune for valuable advice in image processing. We thank Dr. José Biurrum Manresa for critical discussion and assistance with statistical methods, and Prof. Dr. Victor Hugo Casco, for his great support to this research by providing all the necessary equipment and access to the microscopy facilities in LAMAE.\n\n\nSupplementary material\n\nSupplementary File 1: Image processing pipeline for E cadherin fluorescence intensity segmentation and quantitation in 3D-ROIs selected in the trunk epidermis\n\nClick here to access the data.\n\nSupplementary File 2: theoretical psf stack for image processing pipeline\n\nClick here to access the data.\n\nSupplementary Figure 2: Mean cell area of the main cellular types of the enveloping layer (EVL) and epidermal basal layer (EBL) of Danio rerio from 24 to 72 hours post fertilization (hpf) (EVL n=48; EBL n=21), 31 hpf (EVL n=108; EBL n=30), 48 hpf (EVL n=101; EBL n=47) and 72 hpf (EVL n=79; EBL n=45)\n\nClick here to access the data.\n\n\nReferences\n\nCauna N: The free penicillate nerve endings of the human hairy skin. J Anat. 1973; 115(Pt 2): 277–88. PubMed Abstract | Free Full Text\n\nMuroyama A, Lechler T: Polarity and stratification of the epidermis. Semin Cell Dev Biol. 2012; 23(8): 890–896. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWarga RM, Nüsslein-Volhard C: Origin and development of the zebrafish endoderm. Development. 1999; 126(4): 827–838. PubMed Abstract\n\nChang WJ, Hwang PP: Development of zebrafish epidermis. Birth Defects Res C Embryo Today. 2011; 93(3): 205–214. PubMed Abstract | Publisher Full Text\n\nSlanchev K, Carney TJ, Stemmler MP, et al.: The epithelial cell adhesion molecule EpCAM is required for epithelial morphogenesis and integrity during zebrafish epiboly and skin development. PLoS Genet. 2009; 5(7): e1000563. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReig G, Cerda M, Sepúlveda N, et al.: Extra-embryonic tissue spreading directs early embryo morphogenesis in killifish. Nat Commun. 2017; 8: 15431. PubMed Abstract | Publisher Full Text\n\nLittle SC, Mullins MC: Extracellular modulation of BMP activity in patterning the dorsoventral axis. Birth Defects Res Part C Embryo Today Rev. 2006; 78(3): 224–242. PubMed Abstract | Publisher Full Text\n\nLe Guellec D, Morvan-Dubois G, Sire JY: Skin development in bony fish with particular emphasis on collagen deposition in the dermis of the zebrafish (Danio rerio). Int J Dev Biol. 2003; 48(2–3): 217–231. PubMed Abstract\n\nCline A, Feldman SR: Zebrafish for modeling skin disorders. Dermatol Online J. 2016; 22(8): pii: 13030/qt4ws351w8. PubMed Abstract\n\nLee RT, Asharani PV, Carney TJ: Basal keratinocytes contribute to all strata of the adult zebrafish epidermis. Hogan B, editor. PLoS One. 2014; 9(1): e84858. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHulpiau P, van Roy F: Molecular evolution of the cadherin superfamily. Int J Biochem Cell Biol. 2009; 41(2): 349–369. PubMed Abstract | Publisher Full Text\n\nYoshida C, Takeichi M: Teratocarcinoma cell adhesion: identification of a cell-surface protein involved in calcium-dependent cell aggregation. Cell. 1982; 28(2): 217–224. PubMed Abstract | Publisher Full Text\n\nHarris TJ, Tepass U: Adherens junctions: from molecules to morphogenesis. Nat Rev Mol Cell Biol. 2010; 11(7): 502–14. PubMed Abstract | Publisher Full Text\n\nGibson WT, Gibson MC: Cell topology, geometry, and morphogenesis in proliferating epithelia. Curr Top Dev Biol. Elsevier; 2009; 89: 87–114. PubMed Abstract | Publisher Full Text\n\nRagkousi K, Gibson MC: Cell division and the maintenance of epithelial order. J Cell Biol. 2014; 207(2): 181–188. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTheveneau E, Mayor R: Collective cell migration of epithelial and mesenchymal cells. Cell Mol Life Sci. 2013; 70(19): 3481–3492. PubMed Abstract | Publisher Full Text\n\nPetrova YI, Schecterson L, Gumbiner BM: Roles for E-cadherin cell surface regulation in cancer. Mol Biol Cell. 2016; 27(21): 3233–3244. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIzaguirre MF, Casco VH: E-cadherin roles in animal biology: A perspective on thyroid hormone-influence. Cell Commun Signal. 2016; 14(1): 27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBabb SG, Marrs JA: E-cadherin regulates cell movements and tissue formation in early zebrafish embryos. Dev Dyn. 2004; 230(2): 263–77. PubMed Abstract | Publisher Full Text\n\nKane DA, Hammerschmidt M, Mullins MC, et al.: The zebrafish epiboly mutants. Development. 1996; 123: 47–55. PubMed Abstract\n\nKane DA, McFarland KN, Warga RM: Mutations in half baked/E-cadherin block cell behaviors that are necessary for teleost epiboly. Development. 2005; 132(5): 1105–16. PubMed Abstract | Publisher Full Text\n\nSonawane M, Martin-Maischein H, Schwarz H, et al.: Lgl2 and E-cadherin act antagonistically to regulate hemidesmosome formation during epidermal development in zebrafish. Development. 2009; 136(8): 1231–1240. PubMed Abstract | Publisher Full Text\n\nKimmel CB, Ballard WW, Kimmel SR, et al.: Stages of embryonic development of the zebrafish. Dev Dyn. 1995; 203(3): 253–310. PubMed Abstract | Publisher Full Text\n\nIzaguirre MF, Larrea D, Adur JF, et al.: Role of E-cadherin in epithelial architecture maintenance. Cell Commun Adhes. 2010; 17(1): 1–12. PubMed Abstract | Publisher Full Text\n\nFish DA, Brinicombe AM, Pike ER, et al.: Blind deconvolution by means of the Richardson-Lucy algorithm. Journal of the Optical Society of America A. 1995; 12(1): 58–65. Publisher Full Text\n\nSage D, Donati L, Soulez F, et al.: DeconvolutionLab2: An open-source software for deconvolution microscopy. Methods. 2017; 115: 28–41. PubMed Abstract | Publisher Full Text\n\nArganda-Carreras I, Kaynig V, Rueden C, et al.: Trainable Weka Segmentation: a machine learning tool for microscopy pixel classification. Bioinformatics. 2017; 33(15): 2424–2426. PubMed Abstract | Publisher Full Text\n\nClassen AK, Anderson KI, Marois E, et al.: Hexagonal Packing of Drosophila Wing Epithelial Cells by the Planar Cell Polarity Pathway. Dev Cell. 2005; 9(6): 805–817. PubMed Abstract | Publisher Full Text\n\nTunggal JA, Helfrich I, Schmitz A, et al.: E-cadherin is essential for in vivo epidermal barrier function by regulating tight junctions. EMBO J. 2005; 24(6): 1146–1156. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi Y, Naveed H, Kachalo S, et al.: Mechanisms of Regulating Cell Topology in Proliferating Epithelia: Impact of Division Plane, Mechanical Forces, and Cell Memory. Connon CJ, editor. PLoS One. 2012; 7(8): e43108 . PubMed Abstract | Publisher Full Text | Free Full Text\n\nPetrova YI, Schecterson L, Gumbiner BM: Roles for E-cadherin cell surface regulation in cancer. Mol Biol Cell. 2016; 27(21): 3233–3244. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZalik SE, Lewandowski E, Kam Z, et al.: Cell adhesion and the actin cytoskeleton of the enveloping layer in the zebrafish embryo during epiboly. Biochem Cell Biol. 1999; 77(6): 527–542. PubMed Abstract | Publisher Full Text\n\nBruce AE: Zebrafish epiboly: Spreading thin over the yolk. Dev Dyn. 2016; 245(3): 244–258. PubMed Abstract | Publisher Full Text\n\nSchepis A, Nelson WJ: Adherens junction function and regulation during zebrafish gastrulation. Cell Adh Migr. 2012; 6(3): 173–8 . PubMed Abstract | Publisher Full Text | Free Full Text\n\nHildebrand S, Hultin S, Subramani A, et al.: The E-cadherin/AmotL2 complex organizes actin filaments required for epithelial hexagonal packing and blastocyst hatching. Sci Rep. 2017; 7(1): 9540. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLecuit T, Lenne PF: Cell surface mechanics and the control of cell shape, tissue patterns and morphogenesis. Nat Rev Mol Cell Biol. 2007; 8(8): 633–44. PubMed Abstract | Publisher Full Text\n\nArganda-Carreras I, Kaynig V, Rueden C, et al.: Trainable_ Segmentation: Release v3.1.2. zenodo. 2016; Publisher Full Text\n\nCilla R, Mechery V, Hernandez de Madrid B, et al.: Segmentation and tracking of adherens junctions in 3D for the analysis of epithelial tissue morphogenesis. PLoS Comput Biol. 2015; 11(4): e1004124. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLe TL, Yap AS, Stow JL: Recycling of E-cadherin: a potential mechanism for regulating cadherin dynamics. J Cell Biol. 1999; 146(1): 219–232. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWest JJ, Harris TJ: Cadherin Trafficking for Tissue Morphogenesis: Control and Consequences. Traffic. 2016; 17(12): 1233–1243. PubMed Abstract | Publisher Full Text\n\nYamagami K: Mechanisms of hatching in fish: secretion of hatching enzyme and enzymatic choriolysis. Am Zool. 1981; 21(2): 459–471. Publisher Full Text\n\nSampedro MF, Izaguirre MF, Sigot V: Dataset 1 in: E-cadherin expression pattern during zebrafish embryonic epidermis development. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15932.d217819\n\nSampedro MF, Izaguirre MF, Sigot V: Dataset 2 in: E-cadherin expression pattern during zebrafish embryonic epidermis development. F1000Research 2018. http://www.doi.org/10.5256/f1000research.15932.d217820\n\nSampedro MF, Izaguirre MF, Sigot V: Dataset 3 in: E-cadherin expression pattern during zebrafish embryonic epidermis development. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15932.d217821"
}
|
[
{
"id": "39415",
"date": "24 Oct 2018",
"name": "Tony J. C. Harris",
"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 paper describes the distribution of E-cadherin and the packing of cells within the developing epidermis of the zebrafish embryo. From 24-72 hours post fertilization the total amount of E-cadherin detected at all cell-cell contacts of the epidermis was found to increase. Over this same developmental period, hexagonal cells seemed to become more prevalent while tetragonal, octagonal and round cells become less frequent. Other parameters were quantified but were found not to be statistically significant between stages of development.\nA major concern with the analysis is the variability of the E-cadherin detection across the epidermis in single images (e.g. Figure 1e-h). The source of this variability was unclear. It was also unclear how the authors dealt with this variability for the quantifications.\n\nThe authors also noted “dot labeling in the cytoplasm” in panel 1d, but these puncta were not apparent in panel 1g which displayed the same developmental stage.\nStatistical tests were not applied to the analyses of polygon classes in Figure 3. Also it was not stated whether SD or SE bars are shown in Figure 3b.\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": "4084",
"date": "13 Dec 2018",
"name": "Valeria Sigot",
"role": "Author Response",
"response": "This paper describes the distribution of E-cadherin and the packing of cells within the developing epidermis of the zebrafish embryo. From 24-72 hours post fertilization the total amount of E-cadherin detected at all cell-cell contacts of the epidermis was found to increase. Over this same developmental period, hexagonal cells seemed to become more prevalent while tetragonal, octagonal and round cells become less frequent. Other parameters were quantified but were found not to be statistically significant between stages of development.A major concern with the analysis is the variability of the E-cadherin detection across the epidermis in single images (e.g. Figure 1e-h). The source of this variability was unclear.VS: If I understand well, your concern refers to the variability of E-cadh detection between stages and as well within single images (in fig 1 panel e-h). In this figure, single images (stages) are displayed as maximum intensity projections of 50 optical sections, in addition, they are larger than the ROIs used for quantitation aiming to show the curvature of the trunk and changes in cell morphology and density. As stated in the legend, they are contrast enhanced for visualization purposes. For quantitation, these and other images acquired were processed following the protocol (Suppl. File 1) and displayed in Figure 2, each panel representing a single ROI 3D of 50x50x6.6µm but here as the “sum of intensities” for 20 slices from which fluorescence intensity values were quantitated.It was also unclear how the authors dealt with this variability for the quantifications. VS: This was the main challenge of the approach because of the trunk morphology which changes during the studied period and the endogenous fluorescence of the yolk. Although the regions were selected at random along the trunk our criteria was to select the maximum ROI volume that cover a relatively flat region of the epidermis bilayer, with a minimum of 50 stacks including sections below and above the bilayer, to allow for proper restoration of fluorescence upon de-convolution and selecting regions from the yolk extension to the tail to minimize the contribution of the yolk fluorescence.VS: The LMM (Linear Mixed Model) was employed to analyse the statistical differences of the quantitated intensity values between stages for the non-normal distribution of the data. This was selected as a suitable model to consider the variability of E-cadh detection between embryos within each stage. In this approach, the number of embryos per stage (5) and the number of ROIs/embryo (6) were set as a random effect and the stages (4) set as the fixed effect.The authors also noted “dot labeling in the cytoplasm” in panel 1d, but these puncta were not apparent in panel 1g which displayed the same developmental stage.VS: I have already replaced the image for a more evident one, to show the E-cadh cytoplasmic dot pattern along the trunk at 48 hpf.Statistical tests were not applied to the analyses of polygon classes in Figure 3. Also it was not stated whether SD or SE bars are shown in Figure 3b.MFS: Figure 3 shows the distribution of all the polygon classes found in the EVL, in the studied stages. Although the pie chart displays the percentage of x-sided polygons, this represents a qualitative observation of the results and as you suggest, it would be more convenient, to support our results, to make a bar graph showing the mean values per n with its corresponding dispersion measure (SD; Standard Deviation). It is intended to show that there are significant differences between the main classes of polygons found, applying the corresponding statistical tests. MFS: It will be necessary to modify the legend of this figure, specifying that Figure 3a corresponds to a bar graph displaying mean percentages of x-sided polygons of embryos at 24 hpf (n = 81 cells), 31 hpf (n = 147 cells), 48 hpf (n = 133 cells) and 72 hpf (n = 102 cells). The bars will represent mean values with their standard deviations. The type of statistical analysis performed on the data and the p-value obtained will be specified. MFS: Regarding Figure 3b this shows the mean area of cell types at 24 hpf (EVL n = 73), 31 hpf (EVL n = 147), 48 hpf (EVL n = 133) and 72 hpf (EVL n = 102). In the same way, it will be necessary to complete this legend with statistical test applied, the obtained p-value and indicate that the bars represent cell mean area and its standard deviations (SD). All the data to build these figures were obtained from 30 3D-ROIs in five animals for each stage. The excel sheet that shows all these values (mean ± standard deviation) and the new Figure 3 will be completed and uploaded in \"data availability\" section."
}
]
},
{
"id": "39783",
"date": "12 Nov 2018",
"name": "Marc Muller",
"expertise": [
"Reviewer Expertise Zebrafish development",
"organogenesis and toxicology"
],
"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 represents a thorough microscopic analysis of E-Cadh expression, as detected by immunostaining, during zebrafish development in the two layers of epidermal cells.\nFig. 1: what does the white arrow in Fig. 1c point to? Also, weak detection in EBL cells is not so obvious at 18hpf, maybe another arrow (head) would be helpful. Fig. 2: Comparison of E-cadh levels between stages: how far can it be excluded that immunodetection by the antibodies may be systematically different at different stages, considering the important changes in the epidermal cells, especially between 31 and 48hpf (hatching)? This comment also applies to Fig. 4, where in addition the differences between stages are either non-significant or dismissed by the authors.\n\nOverall, this comment sheds some doubt onto the comparisons of intensities between stages, while the changes in topology (polygon classes) is more secure. I was wondering whether counting of the puncta adherens, independent of intensity, could be performed to further support the conclusions. Also, the weaker labelling of the deeper EBL layer would be further supported by performing immunodetection on sections.\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?\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? Partly",
"responses": [
{
"c_id": "4263",
"date": "13 Dec 2018",
"name": "Valeria Sigot",
"role": "Author Response",
"response": "Fig. 1: what does the white arrow in Fig. 1c point to? Also, weak detection in EBL cells is not so obvious at 18hpf, maybe another arrow (head) would be helpful. VS: The white arrow indicates a single puncta adherens in contacting EVL cells. Following your suggestion, an arrowhead will point the weak detection in EBL cells at 18 hpf. Fig. 2: Comparison of E-cadh levels between stages: how far can it be excluded that immunodetection by the antibodies may be systematically different at different stages, considering the important changes in the epidermal cells, especially between 31 and 48hpf (hatching)? This comment also applies to Fig. 4, where in addition the differences between stages are either non-significant or dismissed by the authors. VS: If we understood your question, your concern refers to how we can explain differences in E-cadh expression whilst cells are changing in size and shape so distinctly between stages and in particular between 31 and 48 hpf. Actually, between 31 and 48 hpf there is no significant increase in cell density in the EVL (Fig 4b), Thus, we exclude that the appearance of more cell-cell contacts per area (decrease in cell size and increase in the percentage of hexagonal cells) is the main contribution to the observed difference in E-cadh expression in the EVL. This is explained in the text, but probably the asterisks are misleading in figure 4b, (we will correct that in the new version). The graphs in Figure 4 a and c clearly show this \"trend\" in the increase in the expression level from 31 to 48 hpf, either by quantifying cadherin along fixed volumes of cell-cell contacts, or in individual cells of the EVL. However, this increment was not enough to account for the significant difference observed globally (EVL+EBL) thus we conclude that is the growing detection of E-cadh at the EBL which contributes to this difference between 31 and 48 hpf. We did not dismiss the non-significant results shown in Figure 4 which are only for EVL. The Linear Mixed Model approach proved robust enough for our analysis based on intensity information. Overall, this comment sheds some doubt onto the comparisons of intensities between stages, while the changes in topology (polygon classes) is more secure. I was wondering whether counting of the puncta adherens, independent of intensity, could be performed to further support the conclusions. We understand your concern and we are aware of the limitations that quantitative microscopy has, so far our intention was to be more accurate in measuring the change in protein expression than the amount itself of the protein at a particular stage. Even though, as you propose it is more secure to extract morphological data or counting puncta adherens, our approach supported by the LMM statistical method, allows analysing larger areas of epidermis being more representative and descriptive of the tissue. Also, the weaker labelling of the deeper EBL layer would be further supported by performing immunodetection on sections. It is true, but we prefer to take advantage of the in-toto labeling and imaging of the whole mounted embryos maintaining as much as possible an intact epidermis. In addition, as we do not have the resources to label specifically the basal cells for instance with an anti p63 antibody, we used this differential analysis of the global (EVL+EBL) and the EVL to account for the contribution of the E-cadh expression from the EBL. With our approach we aim to implement an image processing routine for extracting fluorescence intensity and cell morphology features in a bilayer with a single immunofluorescence and acquisition procedure, which can be applied as well for live embryos using a transgenic reporter protein in laboratories with limited equipment and resources."
}
]
},
{
"id": "40062",
"date": "13 Nov 2018",
"name": "James A. Marrs",
"expertise": [
"Reviewer Expertise Zebrafish cadherin adhesion"
],
"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 by Sampedro et al. provides an interesting and useful contribution to epidermis development in the zebrafish. Cadherin adhesion is among the mechanisms that control epidermis development, and the authors contribute an initial description of cadherin adhesion molecule expression changes during epidermis maturation. The image analysis and cell shape changes observed will provide a starting point for additional studies. However, I have some concerns about the manuscript.\nThe monoclonal antibody that the authors used to examine E-cadherin was raised to human E-cadherin cytoplasmic domain sequences. This antibody cross reacts with cadherins in various species and with cadherin types other than E-cadherin. E-cadherin is probably the most prevalent cadherin in the epidermis but this was not shown. If other cadherins are expressed, then this antibody is likely to detect those cadherins too. Flurorescence microscopy can be a useful way to measure protein expression, but several controls would be necessary to ensure that expression is being detected within the linear range of flurorescence detection. These controls were not evident, and expression was not validated with other approaches. The authors state that the prospective epidermis is established at 24 hpf, but this prospective structure is established at neurulation, which occurs significantly earlier.\nOverall, this is a useful contribution to our understanding of cadherin adhesion activity during epidermis development.\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": [
{
"c_id": "4281",
"date": "13 Dec 2018",
"name": "Valeria Sigot",
"role": "Author Response",
"response": "1-The monoclonal antibody that the authors used to examine E-cadherin was raised to human E-cadherin cytoplasmic domain sequences. This antibody cross reacts with cadherins in various species and with cadherin types other than E-cadherin. E-cadherin is probably the most prevalent cadherin in the epidermis but this was not shown. If other cadherins are expressed, then this antibody is likely to detect those cadherins too. VS: We are aware of this, indeed the supplier states in the datasheet that this antibody has some degree of cross-reactivity to P-Cadherin. This cadherin is expected to be expressed in the skin, its presence has been reported in the basal layer of the epidermis in mice most prominently in hair follicles (Tinkle et al. PNAS 2008 105 (40) 15405-15410; https://doi.org/10.1073/pnas.0807374105). In adult zebrafish, the presence of P-cadherin was screened through immunohistochemistry in regenerating blastema but without positive detection (Mateus et al.Development (2015) 142, 2752-2763 doi:10.1242/dev.119701. We did test an anti N cadherin, Clone 32/N-Cadherin (cat. 610920 BD Transduction Lab) IgG1 and we observed only a dim membrane labeling in the eye (figure below). 2-Fluorescence microscopy can be a useful way to measure protein expression, but several controls would be necessary to ensure that expression is being detected within the linear range of fluorescence detection. These controls were not evident, and expression was not validated with other approaches. VS: We understand your concern and we are aware of the limitations that quantitative microscopy has, so far our intention was to be accurate in measuring the change in protein expression than the amount itself of the protein at a particular stage.In the “methods” section we stated that all the acquisition settings were held equal for each 3D stack, but we will specify in the new version of the manuscript that lamp power was set at 6% and the exposure time set at 370 msec. With these settings we collected z-stacks with lower than desired contrast and counted on the improvement in SNR after applying the deconvolution algorithm. Under these conditions photo-bleaching of FITC could be neglected and the distribution of the available intensity levels for all embryonic development stages corresponded well to the linear dynamic range of detection of the CMOS-ORCA-Flash 2.8 (4500:1, calculated from the ratio of the full well capacity and readout noise).With our approach we aim to implement an image processing routine for extracting fluorescence intensity and cell morphology features in a bilayer with a single immunofluorescence and acquisition procedure, which can be applied as well to living embryos using a transgenic reporter protein in laboratories with limited equipment and resources. 3-The authors state that the prospective epidermis is established at 24 hpf, but this prospective structure is established at neurulation, which occurs significantly earlier. VS: Certainly, we misused the term “prospective” to refer to the immature bilayered epidermis this will be corrected in the manuscript"
}
]
}
] | 1
|
https://f1000research.com/articles/7-1489
|
https://f1000research.com/articles/8-190/v1
|
15 Feb 19
|
{
"type": "Research Article",
"title": "Does the mouse tail vein injection method provide a good model of lung cancer?",
"authors": [
"Nensi Shrestha",
"Zabeen Lateef",
"Orleans Martey",
"Abigail R. Bland",
"Mhairi Nimick",
"Rhonda Rosengren",
"John C. Ashton",
"Nensi Shrestha",
"Zabeen Lateef",
"Orleans Martey",
"Abigail R. Bland",
"Mhairi Nimick",
"Rhonda Rosengren"
],
"abstract": "Lung cancer drug development requires screening in animal models. We aimed to develop orthotopic models of human non-small lung cancer using A549 and H3122 cells delivered by tail vein injection. This procedure has been used previously for a mouse lung cancer (Lewis lung carcinoma) and as a model of human breast cancer metastasis to lung. We report that the procedure led to poor animal condition 7-8 weeks after injection, and produced lesions in the lungs visible at necropsy but we were unable identify individual cancer cells using immunohistochemistry. We conclude that if this method is to produce a model that can be used in drug experiments, improvements are required for cancer cell detection post mortem, such as by using of a fluorescently tagged human lung cancer cell line.",
"keywords": [
"human cells",
"lung cancer",
"mouse model",
"tail vein"
],
"content": "Introduction\n\nLung cancer causes more deaths worldwide than any other cancer in both males and females1. Good animal models of lung cancer are essential if treatments are to improve, but there are disadvantages to existing models of lung cancer. Non-small cell lung cancer (NSCLC) comprises approximately 85% of all lung cancer2, and therefore represents the lung cancer subtype where good models are most needed.\n\nMouse models of lung cancer fall into several categories. The first division that can be made is whether the cancer cells are of mouse or human origin. Cancer cells of mouse origin may be grafted onto a host mouse, or induced in tissues by genetic modification, chemical means, or spontaneously. Neither fully recapitulates human cancer. To study the effects of a drug on human cancer in mouse models requires cancer cell xenografts in immunocompromised mice, such as nude or severely immune comprised (SCID) mice. This may be as a solid tumour in the mouse flank or for increased pharmacokinetic validity, an orthotopic model where cells are directly grafted into the lungs. Grafting cancer cells into mouse lungs may take place by several different methods. The cells may be directly injected into the lung, (i.e., by intrathoracic implantation via puncture3), or the cells may be introduced into the airways of the mouse, causing a bronchial tumour4.\n\nAn advantage of intrathoracic implantation, by direct puncturing through the intercostal space to lung parenchyma, is that it avoids thoracotomy or intubation, but the method is disadvantaged by the risk of pneumothorax, intrathoracic haemorrhage and haemoptysis3. Methods of bronchial implantation without surgical thoracotomy have been developed4, but these are disadvantaged by a risk of death during cancer cell inoculation. We therefore experimented with another method of cancer cell inoculation in the mouse lung; engraftment via vascular delivery and pulmonary entrapment. This method has been successfully used to create lung tumour nodules in the lungs of immunocompetent mice using Lewis lung carcinoma cells5, but to our knowledge has not been used to study lung cancer using human lung cancer cells in immunocompromised mice. We therefore carried out a study using SCID mice, inoculating them via tail vein injection either with A549 or H3122 human adenocarcinoma cells. A549 cells are sensitive to Kirsten sarcoma virus protein (Kras) inhibition6 and the H3122 cells are sensitive to anaplastic lymphoma kinase (ALK) inhibition7.\n\n\nMethods\n\nHuman lung adenocarcinoma cells (A549) were maintained in complete growth media (RPMI1640, ThermoFisher, USA) with 2% heat-inactivated foetal bovine serum (FBS, Sigma-Aldrich, NZ). Human NSCLC adenocarcinoma cells harbouring the EML4-ALK variant one (H3122) were maintained in 5% FBS RPMI 1640. All cell lines were also maintained in 2 mM L-glutamine and 1% streptomycin/penicillin (100 μg/mL, Sigma- Aldrich, AU), and grown in a humidified incubator at 5% CO2, 95% O2 and 37°C. At 80–90% confluence, cells were passaged with 1X TrypLE (A459 and H3122 cell lines, ThermoFisher, NZ).\n\nMale immunocompromised SCID mice were purchased from Animal Resources Centre, Australia. All animal experiments were performed after approval by the University of Otago (AEC #9/17). SCID mice were housed in pathogen-free condition with sterile woodchip bedding supplied with sterile food (Reliance rodent diet, Dunedin, NZ) and water. Mice were kept in a room maintained at temperature of 21–24°C on a scheduled 12 h light/dark cycle.\n\nA total of 48 male 12-week-old SCID mice, 20–30 g in weight, were divided into two identical experiments comprised of 24 mice each. As this is a method development study, no data existed with which to carry out a power analysis. Instead, we designed our experiment based on the maximum number of SCID mice we could import in one batch, dividing the mice into two groups in order to trial two different human adenocarcinoma cell lines. Mice were then sub-divided equally between 3-, 5-, and 8-week duration experiments. Two mice were not injected with cells. As this was not a hypothesis test study, we did not allocate mice to groups according to a randomization protocol, but assigned mice to groups sequentially by individual cage that they were housed in (allocated independently by animal technicians). In each experiment, mice were restrained and injected into the medial tail vein (Figure 2) with either 1x105 A549 cells or 1x105 H3122 cells suspended in 100 µl of phosphate buffer saline (PBS). Mice were weighed daily and monitored for mobility, respiratory distress, and signs of pain daily for up to 8 weeks. A weight loss of more than 20% was deemed to be unacceptable and would lead to early sacrifice of the mouse. At the end of 3 weeks in each experiment 4 mice from each group (i.e., 8 mice across the two repeats of the experiment) were euthanized by CO2 exposure and perfused with isotonic PBS followed by 10% formalin. Organs were rapidly removed, weighed and kept in 10% formalin for 48 h followed by 30% starch solution for 24 h at 4°C. The organs were embedded in Optimal cutting temperature compound (OCT) and snap-frozen in liquid nitrogen. Organs removed in this way were brain, kidney, spleen, liver, lungs, heart, and testes. Lungs were quickly photographed prior to freezing and fixation. This was repeated for another 4 mice in each group at 5 weeks and 7–8 weeks following injection, and compared with lungs taken from SCID mice that were not injected with lung cancer cells, from another experiment.\n\nData points are means and error bars are SEM. Up to week three there were fur mice for each cell type. Four mice were then sacrificed, and another four mice at week 5.\n\nBar heights are means and error bars are SEM (n=4).\n\nFrozen, fixed lungs were embedded in OCT and 6-µm sections cut and mounted on poly-L-lysine-treated frosted slides. Mounted sections were either stained with haematoxylin QS (H-3404, Vector laboratories, USA) and eosin (Sigma Aldrich, USA) or probed with antibodies for ALK (Cat# 3633, RRID:AB_11127207, Cell Signaling Technology, USA phospho-ALK (Cat# 6962, RRID:AB_10828357, Cell Signaling Technology, USA) (H3122 cell injected mice only) or human Ki67 Cat# M7249, RRID:AB_2250503, (Dako, Denmark and Abcam, UK).\n\nLung sections were fixed with acetone and methanol (1:2) solution for 10 min and were incubated with haematoxylin followed by eosin. Excess of eosin was removed by rinsing in 95% ethanol and slides were dehydrated in a series of 95%–100% ethanol. The sections were then soaked in xylene and mountant in DPX mounting solution.\n\nFor immunohistochemistry, lung sections were fixed with acetone for 10 min at room temperature and then treated with 0.3% of hydrogen peroxide in methanol for 20 min. Prior to incubation with antibodies, antigen retrieval was performed by boiling at 95°C in either water bath or in a decloaker chamber (Biocare Medical, USA) in citrate buffer (10 mM citric acid, 0.05% Tween-20) pH 6 for 20–30 min (Ki67) and EDTA buffer (1 mM EDTA and 0.05% Tween 20) pH 9 for 30–40 min (ALK/p-ALK). A range of blocking techniques were also trialled, including avidin-biotin blocking for biotinylated secondary antibodies. After incubation with primary antibodies overnight, washed slides were then incubated for up to 2 h with either goat anti-Rabbit IgG, HRP conjugate (Cat# 401353-2ML, RRID:AB_10690659, Millipore, US) or for up to 45 min with biotinylated goat anti-rabbit IgG (Cat# 550338, RRID:AB_393618BD, Biosciences, USA) for subsequent labelling with HRP-conjugated streptavidin( Cat# PA5-54066, RRID:AB_2639134, Thermo Fisher Scientific, USA). Following washing, all slides with stained with 3,3'-diaminobenzidine (DAB,BD Pharmingen, USA). Coverslip mounted sections were then examined by two examiners blinded to the treatment groups using a Nikon RM229 microscope.\n\nWeight gain over time was analysed using linear regression. Organ weights at necropsy were analysed using one-way ANOVA with Bonferroni post hoc tests. All statistical analyses were carried out using GraphPad Prism v7.\n\n\nResults\n\nMice that had been inoculated with A549 human lung cancer cells were slightly heavier than mice inoculated with H3122 cells, as shown by a 1.8 g difference in weight at baseline (Figure 1; F = 884.5. DFn = 1, DFd = 111, P<0.0001, linear regression). However, this gap was maintained during the experiment, such that mice inoculated with A549 cells showed no difference in rate of weight gain from those inoculated with H3122 cells (0.068 g/day and 0.075g/day, respectively, a non-significant difference, F = 2.549. DFn = 1, DFd = 110, P=0.1133, linear regression). However, at over 50 days after inoculation, mice began to show signs of distress, manifested by hunched posture, immobility, rough coats, and laboured breathing. One mouse had proptosis (protruding eyes); this mouse was sacrificed, and all other mice were then sacrificed within a day (i.e., mice in the 8-week group were sacrificed in the 8th week after injection). At this point, mice inoculated with A549 cells appeared to lose weight by 0.08 g/day, although this was not significant (F = 1.52. DFn = 1, DFd = 6, P=0.2638, linear regression). Raw values for body weight, alongside all other raw data, are available on figshare12.\n\nAt necropsy, organs were weighed; there were no significant (P > 0.05, one-way ANOVA with Bonferroni post hoc tests) differences between organ (including lung) weights taken at weeks 3, 5, or 8. (Figure 2). However, examination of the lungs showed that numerous superficial white opacities began to appear by week 5, minimally apparent in week 3 mice lungs and absent in a control mouse (no cancer cells injected) (Figure 3A-C and Figure 4A-C).\n\n(A-C) lungs from mice injected with cells at 3 (A), 5 (B), and 8 (C) weeks after cell injection. Note the white patches on the lungs, which appeared by week 5; D-F. Haematoxylin and Eosin stain of lungs from mice at 3 (D), 5 (E), and 8 (F) weeks after cell injection using a 10x objective. Squares and arrows indicate areas of high cell density; (G-I) Area shown in insets from panels (D-F) respectively, using a 20x objective. (J-L) Area shown in insets from images (D-F), respectively, using a 40x objective. Scale bars: D-I 100 µm; J-L 50 µm.\n\n(A) lungs from a mouse that did not receive a cancer cell injection; B-C. lungs from mice injected with cells at 5 (B), and 8 (C). Note the white patches on the lungs, which appeared by week 5 (as for H3122 cells above). (D-F) Haematoxylin and Eosin stain of lungs from mice at 3 (D), 5 (E), and 8 (F) weeks after cell injection using a 10x objective. Squares and arrows indicate areas of high cell density. (G-I) Area shown in insets from images D-F respectively, using a 20x objective. (J-L) Area shown in insets from images (D-F) using a 40x objective. Scale bars: D-I 100 µm; J-L 50 µm.\n\nHaematoxylin and eosin staining of the lungs did not show tumour cell nodules with distinct edges (Figure 3 and Figure 4) consistent with the dispersed cancer cell pattern observed by Rashid et al.8 using breast cancer cells. Notably, in earlier weeks for both lung cancer cell types, lung sections consisted of a sparse network of bronchioles, and alveolar ducts and sacs with infrequent areas where cells were aggregated in the parenchyma (Figure 3–Figure 4). These areas were more extensive by week 5, and increased through to week 8.\n\nHowever, we were unable to confirm that areas of cellular density corresponded to cancer cells, as we were unable to obtain positive staining for tumour cell markers using immunohistochemistry. In mice inoculated with ALK-positive H3122 cells, we did not find specific staining with either ALK or p-ALK antibodies (with secondary antibody only sections showing high amounts of non-specific staining). Similarly, when we looked in lung sections from mice inoculated with either cell, we could not distinguish primary antibody specific from non-specific labelling for any human cell marker, including Ki67. Although we used a range of antigen-retrieval techniques and different secondary antibodies (both directly conjugated to HRP, and biotinylated) we could not detect any primary antibody specific labelling in inoculated mouse lungs compare to control mouse lungs. There are at least two possible reasons for this. First, cancer cells may have been present but not detected by immunolabelling methods sufficiently due to low concentrations of secondary antibody. This may be because of over-fixation with formalin. To test this hypothesis will require a repeat of the experiment, testing a range of fixation methods or other visualisation methods (such as fluorescently labelling cells). As SCID mice are only available by importation in New Zealand, the country in which these experiments were carried out, this was not possible in the current study. But it is also possible that cancer cells have failed to engraft in lung parenchyma, such that cellular aggregations in haematoxylin and eosin-stained sections were either artefacts or else pathological features secondary to embolism. This second interpretation, however, is difficult to reconcile with the time-dependent appearance of superficial lesions on the lungs, loss of body condition, and increasing density of cells in lung histology, most apparent 8 weeks after tumour cell injection.\n\n\nDiscussion\n\nLung cancer drug development is a difficult process, and using mouse models to screen novel drugs is a critical part of it. Development of new models is therefore part of the process of lung cancer drug development. We carried out these experiments in order to test whether tail vein injection of two commonly used human lung adenocarcinoma cell lines could recapitulate aspects of human lung cancer, to facilitate efficient drug testing. This procedure has successfully been used previously only with one mouse lung cancer cell line (Lewis lung carcinoma)5, where female C57 immunocompetent mice, age 6 weeks, were tail vein injected with 2 × 106 cells in 100 µl PBS). However, although our procedure did lead to poor animal condition in SCID mice 7–8 weeks after injection, as well as lesions in the lungs apparent at necropsy and histological differences, we were unable to identify individual cancer cells using immunohistochemistry. However, we do not yet conclude because of this that a useful model of lung cancer may not be produced by this method.\n\nThe pathological consequences of the tumour cell injection are consistent with thromboembolism, with areas of apparent hypoxia on the lungs at necropsy. In other types of cancer, tail vein injection has tended to lead to sudden death due to a thromboembolism. This and the lack of a more gradual progression of tumour burden has led to criticism of tail vein injection as a model of breast cancer metastasis8. Cancer cell injection into any blood vessel is potentially embolic9, and thromboembolism is a significant cause of death in human lung cancer10, subsequent to the development of tumour nodules. Thus, due to the lack of visible lung cancer nodules using the tail vein injection method, the sudden decline in the mice (at 7–8 weeks after injection), the signs of ischaemia in mouse lungs beginning at week 5, but the difficulty in detecting individual cancers in our experiments, we conclude that the model requires further development if it is to be of value in drug development. We outline a strategy for this in our concluding paragraph below.\n\nFirst, confirmation of hypoxic lesions in the lungs could be ascertained by tetrazolium chloride staining in fresh lung slices11. Second, the structure of the lung is such that relatively strong fixation methods were required to ensure good morphology—more so than for most other tissues in our experience—which can compromise antigen exposure. We aim to overcome this problem in a future experiment through use of a fluorescently-tagged human lung cancer cell line. This approach will facilitate detection of individual cancer cells (by fluorescence microscopy or flow cytometry). Notably, in the study referenced above, Lewis lung carcinoma cells were labelled with GFP, and were thereby visualised by fluorescence microscopy in dissected lung tissue5.\n\n\nData availability\n\nRaw data on body and organ weights for each mouse, alongside uncropped images of the lungs of mice injected with each cell type, are available on figshare. DOI: https://doi.org/10.6084/m9.figshare.7633508.v112.",
"appendix": "Grant information\n\nThis work was funded by a grant from the Otago Medical Research Fund, Grant No. AG 327.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nIslami F, Torre LA, Jemal A: Global trends of lung cancer mortality and smoking prevalence. Transl Lung Cancer Res. 2015; 4(4): 327–38. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGridelli C, Rossi A, Carbone DP, et al.: Non-small-cell lung cancer. Nat Rev Dis Primers. 2015; 1: 15009. PubMed Abstract | Publisher Full Text\n\nLiu X, Liu J, Guan Y, et al.: Establishment of an orthotopic lung cancer model in nude mice and its evaluation by spiral CT. J Thorac Dis. 2012; 4(2): 141–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNakajima T, Anayama T, Matsuda Y, et al.: Orthotopic lung cancer murine model by nonoperative transbronchial approach. Ann Thorac Surg. 2014; 97(5): 1771–5. PubMed Abstract | Publisher Full Text\n\nZhao M, Suetsugu A, Ma H, et al.: Efficacy against lung metastasis with a tumor-targeting mutant of Salmonella typhimurium in immunocompetent mice. Cell Cycle. 2012; 11(1): 187–93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMou H, Moore J, Malonia SK, et al.: Genetic disruption of oncogenic Kras sensitizes lung cancer cells to Fas receptor-mediated apoptosis. Proc Natl Acad Sci U S A. 2017; 114(14): 3648–3653. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShaw AT, Engelman JA: ALK in lung cancer: past, present, and future. J Clin Oncol. 2013; 31(8): 1105–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRashid OM, Nagahashi M, Ramachandran S, et al.: Is tail vein injection a relevant breast cancer lung metastasis model? J thorac dis. 2013; 5(4): 385–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchuh JC: Trials, tribulations, and trends in tumor modeling in mice. Toxicol Pathol. 2004; 32(suppl 1): 53–66. PubMed Abstract | Publisher Full Text\n\nChew HK, Davies AM, Wun T, et al.: The incidence of venous thromboembolism among patients with primary lung cancer. J Thromb Haemost. 2008; 6(4): 601–8. PubMed Abstract | Publisher Full Text\n\nLinsell O, Ashton JC: Cerebral hypoxia-ischemia causes cardiac damage in a rat model. Neuroreport. 2014; 25(10): 796–800. PubMed Abstract | Publisher Full Text\n\nAshton J, Shrestha N: Does the mouse tail vein injection method provide a good model of lung cancer? For F1000Research. figshare. Fileset. 2019. http://www.doi.org/10.6084/m9.figshare.7633508.v1"
}
|
[
{
"id": "45833",
"date": "26 Mar 2019",
"name": "Malcolm Tingle",
"expertise": [
"Reviewer Expertise Adverse drug reactions",
"toxicology",
"species differences in xenobiotic disposition",
"kinetics and toxicity"
],
"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 study has attempted to use a relatively straightforward approach to determine whether injection of adenocarcinoma cells via a tail vein in mice can produce a viable model for lung cancer, presumably with a view to screening potential therapies for human disease. This appears as a reasonable aim, given that the current models are far from perfect, with the potential for a high failure rate of animals through technical difficulties with intrathoracic implantation. Although the current study is limited in its size and scope, it has produced results that are worthy of publication, in part because the model does not appear to have been as successful as perhaps the authors had initially hoped and that along the way there are clear welfare issues that should feature in any ethical harm: benefit consideration by an animal ethics committee.\nOne of the key aims of this type of study is to mimic as closely as possible human disease. Unfortunately, the authors report that despite clear pathological changes occurring in the lungs of the animals, it was not possible to identify the presence of the cells of interest. Without further work-up, this would render the current strategy as potential for the screening of selective therapeutic approaches. The possibility for further workup may be limited by the fact that the tail injection of these adenocarcinoma cells appeared to produce lung damage that adversely impacted on the health of the animals, resulting in a decision to euthanise the animals. In this respect, the authors are to be congratulated on not merely adhering to some obvious metric, such as body weight, but also taking into account more subjective measures. One can only assume that ‘laboured breathing’, presumably as a consequence of the thromboembolism, was in fact quite marked, although the fact that animals did not suffer dramatic weight loss (Figure 1) indicates that they were still eating, and more importantly, not becoming dehydrated. To a large extent, any further attempts to develop this model will be limited by ethical consideration and oversight as science.\nThe manuscript is succinct and given the findings of this preliminary study, the conclusions seem appropriate. However, I would recommend that the authors change the term ‘sacrifice”, since it really has no place in scientific literature, since it does not implicitly infer that the animal was killed humanely. The authors use the terms as if synonyms, “A weight loss of more than 20% was deemed to be unacceptable and would lead to early sacrifice of the mouse. At the end of 3 weeks in each experiment 4 mice from each group (i.e., 8 mice across the two repeats of the experiment) were euthanized by CO2 exposure” However, there may be marked differences between the religious sacrifice of an animal and accepted species-specific practice of euthanasia for animals in scientific experiments covered by legislation.\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": "50732",
"date": "29 Jul 2019",
"name": "Giulio Francia",
"expertise": [
"Reviewer Expertise metastasis"
],
"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 neat paper describing tail vein injection to attempt to generate an in vivo model of lung cancer in mice. It would be helpful if the authors explain why they chose to use male SCID mice, and not female mice. They should also mention that PCR analysis of the lungs could be one way to look for potential metastases (by amplification of sequences that are human-specific). I think it would be helpful to include a side by side comparison (maybe a table?) of the tail vein results and the models mentioned in the intro, or results from other experiments that have established models using the cell lines mentioned in this study. They mention non-small cell lung cancer is the most prevalent subtype of lung cancer, but don’t discuss how the tail vein injection method could potentially model this specific type of lung cancer. The paper mentions intrathoracic implantation. Have the authors tried this method or are they merely stating it is an alternative option? Are the results comparable to the tail vein injection? They don’t state was size needle they use to inject in the tail vein. The authors could mention that a more laborious attempt (not without drawbacks) would be spontaneous metastasis assays, by first injection a primary tumor – followed by surgical resection – and then allowing several weeks for the development of spontaneous metastases.\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/8-190
|
https://f1000research.com/articles/8-187/v1
|
15 Feb 19
|
{
"type": "Case Report",
"title": "Case Report: Levetiracetam causing acute liver failure complicating post-operative management in a neurosurgical patient",
"authors": [
"Sharanya Jayashankar",
"Sunil Munakomi",
"Vignesh Sayeerajan",
"Prakash Kafle",
"Pramod Chaudhary",
"Jagdishchandra Thingujam",
"Deepak Poudel",
"Iype Cherian",
"Sunil Munakomi",
"Vignesh Sayeerajan",
"Prakash Kafle",
"Pramod Chaudhary",
"Jagdishchandra Thingujam",
"Deepak Poudel",
"Iype Cherian"
],
"abstract": "Background: Herein we report a rare case of acute liver failure due to levetiracetam, which has been considered to have an excellent safety profile with minimal hepatic side effects. Case presentation: A 55-year-old male patient presenting with sudden onset dizziness, slurring of speech and headache was operated for posterior fossa cerebellar hematoma. His post-surgical period was complicated by development of icterus with elevation of liver enzymes. After ruling out common inciting factors, it was decided to stop levetiracetam which was given prophylactically for preventing seizures owing to presence of external ventricular drain. From the next day patient had dramatic improvement in liver functions and sensorium. Conclusions: We would like to highlight this side effect that is potentially life threatening, though rare, of levetiracetam, which is very commonly used in today’s practice and fast superseding all other time-tested antiepileptics.",
"keywords": [
"Levetiracetam",
"anti-epileptics",
"hepatotoxicity"
],
"content": "Introduction\n\nSeizure complicates major subsets of patients with stroke, and newer anti-epileptics are being favored in many clinical settings for seizure prophylaxis due to their good safety profile1. Levetiracetam has become one of the most commonly used antiepileptic in current practice for treatment as well as prophylaxis against seizures.\n\nDrug-induced liver injury owing to antiepileptic drugs (AED) is well recognized2. It has been reported to occur more commonly with phenytoin and carbamazepine, and very rarely with valproate2. The consequences of such an injury can be alarming, resulting in harbingering death or the need for liver transplantation. Therefore, newer AED with minimal or no hepatic metabolism are being favored as first line drugs3.\n\nHerein, we report a case of acute liver injury following levetiracetam usage in a post-operative patient of intracerebellar hemorrhage at our Neurosurgery Intensive Care Unit. We implicate a rare but life-threatening effect of a very commonly used anti-epileptic drug. There have been only few case reports of acute liver failure following use of levetiracetam and none were in post-operative neurosurgery cases4–6.\n\n\nCase report\n\nA 55-year-old male patient from a remote village in Biratnagar presented to our emergency department with complaints of sudden onset dizziness, slurring of speech and headache. He was a known hypertensive but not on regular medication or regular follow-up. Neurological examination revealed Glasgow Coma Scale (GCS) of eye opening 4; Verbal 5; and Motor 5, on admission with his bilateral pupils equal and reactive to light. He had no focal neurological deficits or features of meningeal irritation. An emergent Computerized Tomography scan of the head showed features suggestive of cerebellar bleed with fourth ventricle compression with herniation and ventricular extension. While arranging for cerebral angiography, there was a sudden fall in GCS to E1V1M3, and thereby the patient underwent emergency suboccipital craniectomy with evacuation of cerebellar bleed with placement of external ventricular drain.\n\nThe patient’s post-operative medications included ceftriaxone (1gram intravenous every 8th hourly), prophylactic levetiracetam (500 milligram intravenous every 12th hourly), Pantoprazole (40 milligram intravenous every 12th hourly), amlodipine (5 milligram via nasogastric tube 12 hourly), Losartan (50 milligram 12 hourly via Nasogastric tube), and Metoprolol (50 milligram via nasogastric tube 12 hourly). His immediate post-operative GCS improved to E3VtM6.\n\nHowever, on 3rd post-operative day, the GCS fell to E1VTM4. Repeat CT head was uneventful. The patient was noted to have gross icterus and his liver function test revealed total bilirubin of 9.4 mg/dl (normal, 0.1mg/dl), direct 2.0 mg/dl (normal, 0-0.35mg/dl); aspartate aminotransaminase/serum glutamic-oxaloacetic transaminase (AST/SGOT) of 911 IU/L; (normal, 10–40 IU/L); alanine aminotransferase/serum glutamic-pyruvic transaminase (ALT/SGPT)of 926 IU/L (normal, 10–40 IU/L); alkaline phosphatase (ALP) of 298; (normal, 40–112 U/L); International Normalized ratio (INR) of 1.09 (normal, <1.1). Complete blood counts were done to rule out sepsis and were normal. Ultrasound of the abdomen and peripheral smear (for identifying features of obstructive jaundice as well as portal hypertension and ruling out hemolysis for raised bilirubin respectively) were normal. So, a possibility of drug induced liver injury causing acute hepatic failure was considered. Since none of the drugs prescribed were commonly implicated to have hepatotoxic effects, we considered the possibility of levetiracetam following a thorough literature search and hence stopped the drug. We also prescribed prophylactic hepatic encephalopathy regimen with strict monitoring of urine output, GCS, watching for seizures and features of upper gastrointestinal bleed. From the second day of stoppage of the drug, repeat laboratory tests showed gradual improvement in liver functions (Figure 1 and Figure 2) paralleling clinical improvement.\n\nAlthough restarting the patient with the same drug and aided with liver biopsy would be more diagnostic, in our case, the patient’s hepatic function rapidly normalized following stoppage of only levetiracetam from our prescribed drug lists. Therefore, we sufficiently concluded that levetiracetam caused the hepatotoxicity. Though rare, we would like to stress upon the importance of keeping this rare but life-threatening complications of levetiracetam in mind, as it can have profound effect on the timely and corrective management of the patient. There was no episode of recurrence of jaundice seen in the patient within the ensuing 3 weeks.\n\n\nDiscussion\n\nSeizures complicate up to 20% of cases with spontaneous intracerebral hemorrhage7. There is no high level of evidence favoring the use of a specific AED7. Levetiracetam is one of the most commonly used AED in current clinical practice due to its relatively good drug safety profile, and most adverse effects mentioned are usually mild to moderate in intensity8. Levetiracetam does not inhibit or induce hepatic enzymes and most of it is eliminated unchanged by the kidneys. Thus, because it is minimally protein bound and lacks metabolism by the liver, the risk of hepatotoxicity is low. Thus, levetiracetam has a wide safety margin8.\n\nHowever, while reviewing literature, we found a few case reports citing hepatotoxicity with levetiracetam usage4–6. As per the current recommendations by the Council for International Organization of Medical Sciences (CIOMS) for diagnosing drug induced liver injury, our case was of hepatocellular variant9. However, there is no consensus or diagnostic modality in correctly determining the implicated drugs, and thereby this has to be relied solely on the basis of empirical decision as to discontinue or modify drugs during such a scenario9.\n\nTan et al. reported incidence of fulminant liver failure owing to levetiracetam4. Syed and Adams also reported a case of liver failure following prophylactic levetiracetam usage in a patient with head injury5. Sethi et al. reported a post-traumatic head injury patient who developed asymptomatic elevation of hepatic enzymes following levetiracetam usage6.\n\n\nConclusion\n\nThough safe and free of major side effects when comparing to older AED, it is however prudent to note that there are reports of liver injury following levetiracetam, ranging from asymptomatic elevation of transaminases to fulminant hepatic failure. Though routine liver or renal function monitoring may not be needed, it is advisable to keep the patient informed of such possible side effects with the use of newer AED like levetiracetam.\n\n\nConsent\n\nSince the patient was not fully conscious and alert enough to understand the concept of signing consent (since he was in a rehabilitation phase following intracranial hemorrhage), written informed consent for the publication of the relevant clinical and radiological data was obtained from the patient’s wife.\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\nMehta A, Zusman BE, Choxi R, et al.: Seizures After Intracerebral Hemorrhage: Incidence, Risk Factors, and Impact on Mortality and Morbidity. World Neurosurg. 2018; 112: e385–e392. PubMed Abstract | Publisher Full Text\n\nVidaurre J, Gedela S, Yarosz S: Antiepileptic Drugs and Liver Disease. Pediatr Neurol. 2017; 77: 23–36. PubMed Abstract | Publisher Full Text\n\nVidaurre J, Gedela S, et al.: Topical Review-Antiepileptic Drugs and Liver Disease; Division of Pediatric Neurology, Department of Pediatrics, Nationwide Children’s Hospital, Columbus, Ohio.\n\nTan TC, de Boer BW, Mitchell A, et al.: Levetiracetam as a possible cause of fulminant liver failure. Neurology. 2008; 71(9): 685–6. PubMed Abstract | Publisher Full Text\n\nSyed AA, Adams CD: Acute liver failure following levetiracetam therapy for seizure prophylaxis in traumatic brain injury. Case Rep Clin Med. 2012; 1(2): 41–4. Publisher Full Text\n\nSethi NK, Sethi PK, Torgovnick J, et al.: Asymptomatic elevation of liver enzymes due to levetiracetam: a case report. Drug Metabol Drug Interact. 2013; 28(2): 123–4. PubMed Abstract | Publisher Full Text\n\nArboix A, García-Eroles L, Massons JB, et al.: Predictive factors of early seizures after acute cerebrovascular disease. Stroke. 1997; 28(8): 1590–1594. PubMed Abstract | Publisher Full Text\n\nHarden C: Safety profile of levetiracetam. Epilepsia. 2001; 42 Suppl 4(s4): 36–9. PubMed Abstract | Publisher Full Text\n\nDanan G, Benichou C: Causality assessment of adverse reactions to drugs--I. A novel method based on the conclusions of international consensus meetings: application to drug-induced liver injuries. J Clin Epidemiol. 1993; 46(11): 1323–1330. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "46050",
"date": "25 Mar 2019",
"name": "Subish Palaian",
"expertise": [
"Reviewer Expertise Pharmacovigilance"
],
"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: It is a timely report representing an important area of clinical practice and hence deserves indexing.\n\nSpecific comments:\n\nAuthors must mention these points in the case report section:\nWhether this patient had a history of liver problems in the recent past? Any previous history of allergy to any other drugs? Was there use of potentially hepatotoxic general anaesthetic agent? Is he taking any other medications at the time of admission? Is the patient an alcoholic? Any other predisposing factors that could have potentially caused/augmented this adverse drug reaction?\n\nCausality assessment: It will be nice for the authors to perform a ‘Causality assessment’ and grade the Causal relationship between the drug and the event. I recommend using Naranjo scale. Severity assessment: The severity assessment gives an opportunity for the authors to categorize the adverse drug reaction based on its severity. Authors could sue the ‘Hartwig Scale’ for this purpose. In ‘Case report’ Section, …remote village in Biratnagar, Nepal. (add ‘Nepal’) ‘Discussion section’ should emphasize on the management pattern patient (in relation to the adverse drug reactions) highlighting any supportive treatment provided in this patient. This could be even mentioned partly under the ‘Case report’ section.\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": "46049",
"date": "01 Apr 2019",
"name": "Pathiyil Ravi Shankar",
"expertise": [
"Reviewer Expertise Pharmacovigilance",
"pharmacoepidemiology"
],
"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\nMy specific comments to further strengthen the manuscript are shown below.\n\nAbstract Section: Case presentation subsection: Line 5, the word ‘inciting’ can be replaced by a more suitable word. Conclusions subsection: Line 1, ‘adverse effect’ may be a more appropriate term that ‘side effect’. Key words: More key words can be added. Key words can be arranged alphabetically. Page 2, column 1, Case report section, paragraph 1, Line 9: Do the authors mean emergency instead of emergent?\nThe quality of written English is generally good but corrections may be required in a few places. I agree with the other reviewer that the authors can carry out a causality, severity and preventability assessment of the adverse drug reaction.\n\nAlso, information about the medicines used preoperatively and during anesthesia will be helpful as suggested by the other reviewer. Does the institution have a functioning pharmacovigilance center? Was the adverse drug reaction reported to the center?\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/8-187
|
https://f1000research.com/articles/7-1211/v1
|
07 Aug 18
|
{
"type": "Research Article",
"title": "Increasing Epstein-Barr virus infection in Chinese children: A single institutional based retrospective study.",
"authors": [
"Kiran Devkota",
"Maio He",
"Meng Yi Liu",
"Yan Li",
"You Wei Zhang",
"Kiran Devkota",
"Maio He",
"Meng Yi Liu",
"Yan Li"
],
"abstract": "The Epstein-Barr virus (EBV) is a common virus in humans and the most common causative agent of Infectious Mononucleosis. EBV primary infection has recently risen in some countries and children below 2 years of age are highly susceptible. The clinical manifestations in children with EB virus infection involve multiple systems, causing severe illness, meaning attention should be paid during diagnosis and treatment. Objective: This single institution based retrospective study was carried out with the aim of estimating the overall prevalence of EBV infection and identifying high-risk age group among children. Methods: This study include total 253 patients under 15 years of age found to be positive for EBV DNA by PCR who were admitted to the Pediatrics Department of Renmin Hospital,(Shiyan, China) during a 4-year period from 2014 to 2017. Patients were divided into three groups; 0-<4years, 4-<6years and 6-<15years. We then calculated the percentage and prevalence of EBV DNA-positive cases. Results: The yearly EBV prevalence rate was 4.99 per 1000 admissions in 2014, 6.97 per 1000 admissions in 2015, 10.42 per 1000 admissions in 2016, and 12.16 per 1000 admissions in 2017. Out of 253 EBV-positive cases, those under 4 years had the highest rate of EBV infection (74.7%). The rate drops to 11.06% in the 4-6 years group, and was 14.22% in the 6-15 years group. Those between 6 months and 1 year are those at the highest risk. Conclusion: The rate of hospital admission of children due to EBV infection is increasing day by day. Children under 4 years of age are highly susceptible to infection and children of age between 6 months and 1 year are the high-risk group for EBV infection.",
"keywords": [
"Infectious mononucleosis",
"Prevalence",
"EBV DNA",
"Epstein–Barr virus"
],
"content": "Introduction\n\nThe Epstein-Barr virus (EBV) is the most common herpesvirus in humans and the most common causative agent of infectious mononucleosis1. It is also known as the “kissing disease”2. EBV is an acute infection with a characteristic symptomatic triad of fever, sore throat and lymphadenopathy. Sprunt and Evans in 1920 coined the term infectious mononucleosis to describe an acute infectious disease accompanied by atypical large peripheral blood lymphocytes2. EBV primary infection has recently risen in some countries3 and children below 2 years of age are highly susceptible4,5. EBV is transmitted primarily via oral secretions and may be transmitted via penetrative sexual intercourse6. Transmission may occur by the exchange of saliva among children. EBV is not spread by non-intimate contact, environmental sources, or fomites6. During late adolescence 50–70 percent of teenagers get infected with infectious mononucleosis2. Though it has a self-limiting course, it may sometimes lead to numerous rare, atypical and threatening manifestations. The clinical manifestations in children with EBV infection involve multiple systems and can cause severe illness, meaning that attention should be paid during diagnosis and treatment. The diagnosis of EBV infection is based on clinical features such as- fever, pharyngitis, lymphadenopathy, hepatomegaly, and splenomegaly along with leukocytosis with a predominance of lymphocytes, >10% atypical lymphocytosis, heterophile antibodies (assessed via monospot test), PCR for EBV DNA and serological testing including antibodies for viral capsid antigens, early antigens, and Epstein-Barr nuclear antigen.\n\nEBV DNA PCR has high specificity and sensitivity for identifying patients with infectious mononucleosis7.\n\n\nMethods\n\nWe retrospectively collected the EBV infection status in children <15 years old at Renmin Hospital, 3rd Affiliated Hospital of Hubei University of Medicine, Shiyan, (Hubei, China) during the 4-year period from January 1, 2014, to December 31, 2017. In total, 253 patients had with EBV infection and were positive for EBV DNA. At birth, neutrophils make up around 61% of total leukocytes and lymphocytes make up around 31%. After birth, the number of neutrophils goes down and the lymphocyte number goes up, with both reaching about 45% around the 1st week of life. This process continues and by the age of 4 years, lymphocytes reaches around 50% and neutrophils reach around 42%. On growing older, the proportion of lymphocytes starts to fall and that of neutrophils start to increases. By the age of 6 years, the proportion of neutrophils reaches up to 51% and that of lymphocytes falls to 42%8. Owing to this age-specific leukocytes differential, we divided patients into three age groups: <4 years, 4–<6 years and 6–<15 Years. We also made further age-specific groupings, as follows: <30 days, 1–<6 months, 6–<12 months, 1 year, 2 years, 3 years, 4 years, 5 years, 6 years, 7 years, 8 years, 9 years, 10 years, 11 years, 12 years, 13 years, and 14 years to find out the risk group for EBV infection. A diagnosis of EBV infection was achieved using real time PCR at the Pathology Department at Renmin Hospital.\n\nReal-time PCR ABI iiA7 was used for quantitation of EBV DNA. The primers used, targeting the EBNA-1 fragment of EBV, were as follows: 5’-GTAGAAGGCCATTTTTCCAC-3’ (forward) and 5’-TTTCTACGTGACTCCTAGCC-3’ (reverse). PCR was conducted using the following thermocycling conditions: 93°C for 2 min, followed by 10 cycles of 93°C for 45 sec and 55°C for 60 sec, and then 30 cycles of 93°C for 30 sec and 55°C for 45 sec.\n\nAll data were analyzed using Microsoft Excel 2010. Age-specific prevalence was calculated. Prevalence was calculated as follows:\n\nPrevalence = number of EBV-positive children under 15 years admitted to hospital / number of total hospital admissions for children under 15 years\n\n\nResults\n\nOut of the total of 253 patients, 151 (60%) were male and 102 (40%) were female. The male to female ratio was 3:2 (Figure 1).\n\nThe number of EBV DNA-positive cases observed increased each year. There were 36 EBV DNA positive cases in 2014 (total admissions, 7202) with a prevalence of 5.00 per 1000 admissions, 43 on 2015 (total admissions, 6163) with a prevalence of 6.98 per 1000, 77 on 2016 (total admissions, 7384) with prevalence of 10.43 per 1000 and 97 on 2017 (total admissions, 7972) with prevalence of 12.17 per 1000 admissions (Figure 2, Figure 3).\n\nOver the 4 years studied here, the numbers of hospitalized children were highest in the 0 to < 4 years group. Of 253 EBV-positive patients, 189 (74.70%) were in group 0 to less than 4 years, 28 (11.06%) in the group of children aged 4 to <6 years, and 36 (14.23%) in those aged 6 to <15 years. Each year, in the group of children under 4 years the percentage of EBV positive cases were more and rate were in increasing trend (Figure 4, Figure 5).\n\nWe calculated the age-specific prevalence of EBV infection to identify the high-risk group. The number of positive cases was highest in the age group 6 months- <1 year, which decreased as age increased. Prevalence is also high in this age group (Table 1 and Figure 6, Figure 7).\n\nP, Prevalence of Epstein-Barr virus (EBV)-positive cases per 1000; C, number of EBV-positive cases; N, total number of hospital admissions.\n\n\nDiscussion\n\nThe incidence of EBV infection is higher in male children in Northern China10 and Turkey11. In India, the male to female ratio of EBV infection in hospitalized children is 2:112,13. A Korean study found the overall male-to-female ratio of EBV infection to be 1.53:114. Our study had a male to female ratio of 3:2. During adolescence, women acquire before men the first infection by EBV15. In the US EBV antibody titers were significantly higher for females16.\n\nWe have found that in children under 4 years, the percentage of EBV-positive cases increased each year. However, in children aged 4–<6 years this decreased, but increased in those aged 6 to <15 years. Out of the 253 EBV positive patients, those aged under 4 years made up the highest proportion (74.7%). This drops to 11.06% in those 4–<6 years, and 14.22% in those 6–<15 years. In the study done on the Northern and Southern part of China, the seroprevalence of EBV infection is more than 50% before age 32. Serological evidence of EBV infection is found in around 84% of Chinese children aged >9 years, with peak incidence observed at age 2–3 years17. However, in a study done by Gao et al., the incidence of EBV-IM peaked in children at age of 4–<6 years in Northern China10. In Taiwan, the seropositive rate of EBV is high in children aged 2 years4. A Danish study found that EBV infection is common in young children, and children under 3 years of age constitute the largest group of hospitalizations for acute EBV infection5. In a study conducted in Poland, age of infection occurred in two peaks, i) in children aged 1 to 5 years (62%), and ii) in teenagers (24.6%)18. In most developing countries nearly 70% of patients are seropositive for EBV by the age of 2 years19. However in USA, the seroprevalence increased with age, ranging from 54.1% for 6–8 year-olds to 82.9% for 18–19 year-olds16.\n\nWe found hospitalization for mononucleosis in all age groups. The number of positive cases was higher in the age group >6 months but <1 year, which decreases as age increases. The prevalence is also high on age group 6 month to 1 year. This indicates that the age group 6 months to less than 1 year is a high-risk group. The most common age group for hospitalization with acute EBV infection in Denmark was 1–2 years5. In Asia and other developing countries most of the children are infected with EBV in early life, mostly before the age of 1 year.20. According to Cocuz et al., admissions for infectious mononucleosis were prevalent in young children, with most occurring in the 1–3 years age group (32.31% of the total IM Cases), followed by those 4–<6 years old (27.69% of the total IM Cases), then those 11–16 years old (26.15% of the total IM Cases) and finally those 7–10 years old (13.84% of the total IM Cases)21.\n\nSeveral prior studies have reported in the last decade which shows the changes in the epidemiology of EBV infection. A Japanese study showed that the seroprevalence of EBV in 5–7 years old children was higher than 80% before the early 1990s which decreased to 59% in the years 199520. Similarly in the USA, the study showed that the seroprevalence in 6–19 year olds declined from 72% in 2003–2004 to 65% in 2009–201022. But, the EBV primary infection is increasing in England and Wales23. Therefore, we aimed to determine the epidemiological condition of EBV infection over the last years in the Pediatrics Department of Renmin Hospital, Shiyan, China. The EBV positivity rate in hospitalized children is increasing every year. Prevalence is also increased each year. In the years 2000 to 2016, the EBV infection rate in France has increased, whereas its seroprevalence has decreased3.\n\nAlthough most EBV infections are self-limiting, sometimes they may lead to rare, atypical and threatening manifestations. Although serious complications during the acute phase of primary EBV infection are rare1, neurological complications, like meningoencephalitis, acute encephalitis, acute cerebellitis, transverse myelitis, and myeloradiculitis, occur more frequently in children under 2 years of age18,24,25. Furthermore, in immunocompromised individuals, there was an association observed between EBV with several tumors following reactivation of the virus from latency26.\n\nSince this study was conducted in children admitted to hospital, the results might lack generalization to the entire population, but may indicate trends and bring up questions deserving further prospective study.\n\nIncreasing primary infection of EBV in children may be due to many reasons, including that the virus is active among the population around Shiyan, airborne transmission27 of the virus is higher in this area, multiple caregivers for each infant, bottle feeding, unnecessary kissing, feeding with chewed food to babies, or through hospital acquired EBV infection e.g. from health care personals, doctors or nurses. There are several reports on the intrauterine transmission of EBV, but none has been substantiated by appropriate viral studies28,29. Besides, doctors may be more familiar and experienced with the clinical presentation, symptoms, and signs of infectious mononucleosis.\n\nThe next steps should be a focus on awareness to parents and caregivers of children, and development of a vaccine against EBV to reduce the burden of EBV infection in future.\n\n\nConclusion\n\nThe rate of hospital admission of children due to EBV infection is increasing. Children under 4 years of age are highly susceptible to infection and children of age between 6 months and 1 year are the high-risk group for EBV infection. Vaccination against EBV must be considered to reduce the burden of EBV infection in future.\n\n\nData availability\n\nDataset 1. The number of total admissions and admissions of Epstein-Barr virus (EBV)-positive children under 15 years of age for each of the years 2014–2017. This dataset also contains stratifications of EBV-positive individuals by age and sex. DOI: https://doi.org/10.5256/f1000research.15544.d2121419.",
"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\nAcknowledgments\n\nWe wish to thank Dr Li Lian Director of Pathology Department, Renmin Hospital Dr. Liu Zheng Mei, Dr. Ke Wei, Dr. Tian Cai Xia and entire staff of doctors and nurses at the Department of Pediatrics, Renmin Hospital.\n\n\nReferences\n\nDunmire SK, Hogquist KA, Balfour HH: Infectious Mononucleosis. Curr Top Microbiol Immunol. 2015; 390(Pt 1): 211–240. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXiong G, Zhang B, Huang MY, et al.: Epstein-Barr virus (EBV) infection in Chinese children: a retrospective study of age-specific prevalence. PLoS One. 2014; 9(6): e99857. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFourcade G, Germi R, Guerber F, et al.: Evolution of EBV seroprevalence and primary infection age in a French hospital and a city laboratory network, 2000-2016. PLoS One. 2017; 12(4): e0175574. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen CY, Huang KY, Shen JH, et al.: A large-scale seroprevalence of Epstein-Barr virus in Taiwan. PLoS One. 2015; 10(1): e0115836. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTopp SK, Rosenfeldt V, Vestergaard H, et al.: Clinical characteristics and laboratory findings in Danish children hospitalized with primary Epstein-Barr virus infection. Infect Dis (Lond). 2015; 47(12): 908–14. PubMed Abstract | Publisher Full Text\n\nJenson HB: Chapter 154: Epstein - Barr virus. In Nelson Textbook of Pediatrics. 20th Edition. Elsevier, Philadelphia: Elsevier Inc. 2016; 1586–1590.\n\nJiang SY, Yang JW, Shao JB, et al.: Real-time polymerase chain reaction for diagnosing infectious mononucleosis in pediatric patients: A systematic review and meta-analysis. J Med Virol. 2016; 88(5): 871–876. PubMed Abstract | Publisher Full Text\n\nRadha G, Elizabeth J: “Chapter 14: Hematology.” In The Harriet Lane Handbook, 20th Edition. Elsevier, Philadelphia: Elsevier Inc. 2015; 313–314.\n\nDevkota K, He M, Liu MY, et al.: Dataset 1 in: Increasing Epstein-Barr virus infection in Chinese children: A single institutional based retrospective study. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15544.d212141\n\nGao LW, Xie ZD, Liu YY, et al.: Epidemiologic and clinical characteristics of infectious mononucleosis associated with Epstein-Barr virus infection in children in Beijing, China. World J Pediatr. 2011; 7(1): 45–9. PubMed Abstract | Publisher Full Text\n\nCengiz AB, Cultu-Kantaroğlu O, Seçmeer G, et al.: Infectious mononucleosis in Turkish children. Turk J Pediatr. 2010; 52(3): 245–54. PubMed Abstract\n\nRamagopalan SV, Giovannoni G, Yeates DG, et al.: Sex ratio of infectious mononucleosis and possible relevance to multiple sclerosis. Mult Scler. 2013; 19(3): 359–61. PubMed Abstract | Publisher Full Text\n\nBalasubramanian S, Ganesh R, Kumar JR: Profile of EBV associated infectious mononucleosis. Indian Pediatr. 2012; 49(10): 837–8. PubMed Abstract | Publisher Full Text\n\nSon KH, Shin MY: Clinical features of Epstein-Barr virus-associated infectious mononucleosis in hospitalized Korean children. Korean J Pediatr. 2011; 54(10): 409–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTrastoy Pena R, Costa Alcalde JJ, Rodríguez Calviño J, et al.: [Infection by the Epstein-Barr virus between the years 2006-2015 in the health area of Santiago de Compostela. Relationship with age and sex]. Rev Esp Quimioter. 2017; 30(6): 468–471. PubMed Abstract\n\nDowd JB, Palermo T, Brite J, et al.: Seroprevalence of Epstein-Barr virus infection in U.S. children ages 6-19, 2003-2010. PLoS One. 2013; 8(5): e64921. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuang Y, Wei C, Zheng K, et al.: The impact of serological features in Chinese children with primary or past Epstein-Barr virus infections. Virol J. 2013; 10: 55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMillichap JG: Epstein-Barr Virus Neurologic Complications. Pediatric Neurology Briefs. 2015; 29(11): 88. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChabay PA, Preciado MV: EBV primary infection in childhood and its relation to B-cell lymphoma development: a mini-review from a developing region. Int J Cancer. 2013; 133(6): 1286–92. PubMed Abstract | Publisher Full Text\n\nTakeuchi K, Tanaka-Taya K, Kazuyama Y, et al.: Prevalence of Epstein-Barr virus in Japan: trends and future prediction. Pathol Int. 2006; 56(3): 112–6. PubMed Abstract | Publisher Full Text\n\nCocuz ME, Cocuz IG: Infectious Mononucleosis in Children- Current Clinical and Epidemiological Aspects. Bulletin of the Transilvania. University of Braşov. 2016; 9(58): 55–60. Reference Source\n\nBalfour HH Jr, Sifakis F, Sliman JA, et al.: Age-specific prevalence of Epstein-Barr virus infection among individuals aged 6-19 years in the United States and factors affecting its acquisition. J Infect Dis. 2013; 208(8): 1286–93. PubMed Abstract | Publisher Full Text\n\nMorris MC, Edmunds WJ, Hesketh LM, et al.: Sero-epidemiological patterns of Epstein-Barr and herpes simplex (HSV-1 and HSV-2) viruses in England and Wales. J Med Virol. 2002; 67(4): 522–7. PubMed Abstract | Publisher Full Text\n\nÇelik T, Çelik Ü, Tolunay O, et al.: Epstein-Barr virus encephalitis with substantia nigra involvement. J Pediatr Neurosci. 2015; 10(4): 401–403. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMazur-Melewska K, Breńska I, Jończyk-Potoczna K, et al.: Neurologic Complications Caused by Epstein-Barr Virus in Pediatric Patients. J Child Neurol. 2016; 31(6): 700–8. PubMed Abstract | Publisher Full Text\n\nAli AS, Al-Shraim M, Al-Hakami AM, et al.: Epstein- Barr virus: Clinical and Epidemiological Revisits and Genetic Basis of Oncogenesis. Open Virol J. 2015; 9: 7–28. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPayne DA, Mehta SK, Tyring SK, et al.: Incidence of Epstein-Barr virus in astronaut saliva during spaceflight. Aviat Space Environ Med. 1999; 70(12): 1211–3. PubMed Abstract\n\nHorwitz CA, McClain K, Henle W, et al.: Fatal illness in a 2-week-old infant: diagnosis by detection of Epstein-Barr virus genomes from a lymph node biopsy. J Pediatr. 1983; 103(5): 752–755. PubMed Abstract | Publisher Full Text\n\nGoldberg GN, Fulginiti VA, Ray CG, et al.: In utero Epstein-Barr virus (infectious mononucleosis) infection. JAMA. 1981; 246(14): 1579–1581. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "37372",
"date": "12 Sep 2018",
"name": "Anna Mania",
"expertise": [
"Reviewer Expertise infectious diseases",
"pediatrics"
],
"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 presents results of retrospective analysis conducted to estimate the overall prevalence of EBV infection and identify high risk groups among children. This study included 253 patients under 15 years of age positive for EBV DNA by PCR, were admitted to the Pediatrics Department in China during a 4-year period from 2014 to 2017. Patients were divided into three groups; 0-<4years, 4-<6years and 6-<15years. The percentage and prevalence of EBV DNA-positive cases was calculated on that basis. The highest rate of EBV infection (74.7%) was observed in the group under 4 years of age, 11.06% and 14.22% in the 4-6 years and in the 6-15 years group, respectively. The authors mention increasing number of EBV-infected individuals in recent years. Noticing the highest number of cases in children between 6 month to 1 year.\nCertain limitations of the study were visible:\nThe authors confirm EBV infection by PCR, however I could not find the information concerning the type of the specimen – blood, saliva, urine, anything else? Significant proportion of EBV infected patient may be asymptomatic; the term infection is not equivalent to the disease, therefore it is a pity that data concerning clinical symptoms are not mentioned.\nMinor comments:\nCiting articles are not equivalent to given numbers e.g –study conducted in Poland is given at the 25 positions in the list, not 18;\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?\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": "4397",
"date": "15 Feb 2019",
"name": "Kiran Devkota",
"role": "Author Response",
"response": "Respected Professor Thank you so much for you time and consideration. I have tried to make clear on the queries that you have noted."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1211
|
https://f1000research.com/articles/8-181/v1
|
14 Feb 19
|
{
"type": "Case Report",
"title": "Case Report: Rare comorbidity of celiac disease and Evans syndrome",
"authors": [
"Syed Mohammad Mazhar Uddin",
"Aatera Haq",
"Zara Haq",
"Uzair Yaqoob",
"Syed Mohammad Mazhar Uddin",
"Aatera Haq",
"Zara Haq"
],
"abstract": "Background: Celiac disease is an immune-mediated enteropathy due to permanent sensitivity to gluten in genetically predisposed individuals. Evans syndrome is an autoimmune disorder designated with simultaneous or successive development of autoimmune hemolytic anemia and immune thrombocytopenia and/or immune neutropenia in the absence of any cause. Case Report: We report a rare case of Celiac disease and Evans syndrome in a 20-year-old female who presented to us with generalized weakness and shortness of breath. Her examination finding included anemia, jaundice, and raised jugular venous pulse. Her abdominal exam revealed hepatosplenomegaly. Her laboratory values showed microcytic anemia, leukocytosis and thrombocytopenia. To rule out secondary causes of idiopathic thrombocytopenia purpura, we tested viral markers for Human immunodeficiency virus, Epstein bar virus, Cytomegalovirus and performed a Helicobacter pylori test, all of which were negative. We also ruled out idiopathic thrombocytopenia purpura associated with any thyroid disorder. For celiac disease, we took anti-tissue transgulataminase titers of IgA and IgG which confirmed the diagnosis of celiac disease. For the diagnosis of Evans syndrome, despite a negative serum coombs test initially, her bone marrow sample showed a positive Coombs test along with immune mediated hemolytic anemia and immune mediated thrombocytopenia. The patient was treated with prednisone which was tapered off and counseling was provided regarding a gluten free diet. Conclusion: Although rare, tests for Evans syndrome (and other coexisting autoimmune problems) should be performed in patients with celiac disease.",
"keywords": [
"Celiac disease",
"Evans syndrome",
"autoimmune disease"
],
"content": "Introduction\n\nCeliac disease (CD) is defined as an immune-mediated enteropathy due to permanent sensitivity to gluten in genetically predisposed individuals. It is known to affect roughly 1% of the population worldwide. This occurs when genetically predisposed individuals consume gluten, which is a storage protein in wheat, and other wheat related grain species (e.g. barley and rye)1. Several conditions such as dermatitis herpetiformis, autoimmune thyroiditis and type 1 diabetes mellitus has been reported with CD2.\n\nEvans syndrome (ES) is an autoimmune disorder designated with simultaneous or successive development of autoimmune hemolytic anemia (AIHA) and immune thrombocytopenia (ITP) and/or immune neutropenia in the absence of any cause. However, it may be associated with conditions such as systemic lupus erythematosus (SLE), lymphoproliferative disorders or primary immunodeficiencies3. The pathway of the disease is chronic and relapsing, with unknown pathophysiology. Studies have reported that autoantibodies target antigens on red blood cells and platelet, leading to hemolytic anemia and isolated thrombocytopenia2. We report a unique case of 20-year-old female with celiac disease and Evans syndrome (ES).\n\n\nCase presentation\n\nA 20-year-old Sindhi female student with no known comorbid presented to the emergency department with a complaint of generalized weakness and shortness of breath over the previous 15 days. According to the patient herself, the generalized weakness was progressive and with increasing intensity to such an extent that it hampered her daily activities. On top of that she was also experiencing shortness of breath which was also progressive. However, she denied any orthopnea, paroxysmal nocturnal dyspnea (PND), fever, rash, altered bowel habits, cough, joint pain and any acute history of blood loss. According to the patient’s past medical history, she had on and off loose stools from 9 years of age, which resolved by age 16. Furthermore, 2 years back she was admitted to a nearby hospital with generalized weakness and jaundice. There is no official documentation but reportedly she was also transfused with 2 blood bags. Workup and diagnosis were not completed during her stay as she was non-compliant and left against medical advice at that time. All other tests were normal and her menstrual history was also normal. On examination, her vitals were blood pressure 110/60 mmHg, (reference, 120/80mm/hg); pulse 90 beats/minute (reference range, 70–100 beats/minute); temperature 98°F (reference range, 97–99°F) and respiratory rate 22 breaths/minute (reference range, 12–20 breaths/minute). Her general physical examination showed anemia, jaundice and clubbing, along with a raised jugular venous pulse. Her respiratory, cardiovascular system and central nervous examination were normal. However, her abdominal examination showed hepatomegaly (with liver palpable up to one finger) and splenomegaly (with spleen palpable up to 3 fingers below the costal margin) with the rest of the examination being normal.\n\nBased on the history and examination, we ordered pertinent laboratory work up along with other tests. Her base line laboratory values were hemoglobin 2.3 g/dL (reference range, 11.1 – 14.5 g/dL); mean corpuscular volume (MCV) of 73.2 (reference range, 76 – 96); total lung capacity (TLC) 4.2×109/L; platelets 92000/mm3 (reference range, 150 – 400000/mm3); sodium 140 meq/L (reference range, 135–145 meq/L); potassium 3.6 meq/L (reference range, 3.5–5 meq/L); chloride 108 meq/L (reference range, 97–107 meq/L); calcium 8.5 meq/L (8.5–10.2 meq/L); magnesium 2.2 (reference range, 1.5–2.5 meq/L); total bilirubin 1.97 (reference range, 0.1–1.2 meq/L); serum glutamic pyruvic transaminase (SGPT) 59 units/L (reference range, 7–56 units/L); alkaline phosphatase (ALP) 123 IU/L (reference range, 44–147 IU/L); serum glutamic-oxaloacetic transaminase (SGOT) 40 units/L (reference range, 5–40 units/L); total protein 7.7 g/dL (reference range, 6–8.3 g/dL); albumin 3.5 g/dL (reference range, 3.5–5 g/dL) and prothrombin time to international normalized ratio (PT/INR) 11 seconds (reference range, 11–13.5 seconds).\n\nAs she presented with shortness of breath, we also performed an echocardiography and chest x-ray, both were within normal limits. At this point in time, we assumed that the patient’s symptoms and signs could be due to some autoimmune disorder (such as SLE, rheumatoid arthritis (RA)), immune mediated disorders (such as celiac disease), thyroid disorders (example immune thrombocytopenia purpura), Evan syndrome (due to thrombocytopenia and hemolytic anemia) and even chronic liver disease. Hence, in order to rule out the differentials and get to a possible diagnosis we conducted more tests. The patient iron profile showed ferritin 115.5 ng/mL (reference range, 12–150 ng/mL), serum iron 82 mcg/dL (reference range, 50–170 mcg/dL), total iron binding capacity (TIBC) 258 mcg/dL (reference range, 250–370 mcg/dL), transferrin saturation 37% (normal 25–35%) and her B12 levels were 2000 pg/mL (reference range, >200 pg/mL), both were in normal range. To rule out thyroid disorders, we checked her thyroid profile [thyroid stimulating hormone (TSH) 2.75 mU/L, (reference range, 0.5–4.0 mU/L); tri-iodothyronine (fT3) 2.45 pg/mL, (reference range, 2.3–4.2 pg/mL); thyroxine (fT4) 1.71 ng/dL, (reference range, 0.8–1.8 ng/dL)], which was within normal limits. She was also negative for anti-nuclear antibody, anti-Smith antibodies, anti-double stranded DNA antibody, anti-smooth muscle antibody and anti-mitochondrial antibody, thereby ruling out the possibility of autoimmune disorders. Furthermore, her blood Coombs test was also negative. To rule out secondary causes of idiopathic thrombocytopenia purpura (ITP), we tested for viral markers of human immunodeficiency virus (HIV), Epstein bar virus (EBV), cytomegalovirus (CMV) and performed a Helicobacter pylori test, all of which were negative. For celiac disease, we took anti-tissue transglutaminase (anti-TTG) titers of IgA and IgG, which came out to be 353 U/mL, (reference, <20 U/mL); and 419 U/mL (reference, <20 U/mL) respectively. The elevated titers confirmed the diagnosis of celiac disease. However, at this point in time, in order to improve symptoms of the patient, she was transfused with 2 bags of blood. Her bone marrow biopsy showed a positive coombs test with immune mediated hemolytic anemia and immune mediated thrombocytopenia. Hence, we established a diagnosis of celiac disease with Evans syndrome. However, in order to confirm our diagnosis of celiac disease, endoscopy was scheduled, but the patient did not consent to the procedure. Moreover, we started oral prednisone therapy 40 mg once daily which was tapered off and stopped in four weeks and the patient was counseled about a gluten free diet. Within these four weeks, the patient felt better and was discharged.\n\n\nDiscussion\n\nCeliac disease can result from gluten, several environmental triggers and immune factors. Gluten is the absolute protein component of wheat, whereas Gliadin is the alcohol-soluble fraction of gluten. The immune response to Gliadin leads to an inflammatory reaction in the small intestine, with infiltration of the lamina propria and epithelium with inflammatory cells along with villous atrophy. Furthermore, the development of celiac disease is highly linked to alleles that codes for HLA-DQ2 or HLA-DQ8 proteins, which are the products of the two HLA genes in adults. Women, due to unknown reasons, are more prone to the disease. Moreover, the frequency of autoimmune diseases in general is greater in women as compared to men and diseases like osteoporosis and iron deficiency anemia, which evoke work-up of celiac disease, are also common in women. The classic presentation is diarrhea, which may be associated with abdominal pain. However, studies have reported in the past decade that diarrhea is the chief complaint in less than 50% of cases. Moreover, with insufficient data on classic symptoms, the notion of silent celiac disease has arisen with the advent of serologic screening. Silent presentation includes iron deficiency anemia, incidental findings on endoscopy for other symptoms such as gastro esophageal reflux and osteoporosis. Furthermore, other less common findings can include constipation, neurologic symptoms, elevated liver enzymes, hypoproteinemia and hypocalcaemia. In addition, another study states that over the past 5 years, one third of the celiac disease patients were diagnosed by serological screening without gastrointestinal symptoms, indicating it can also be asymptomatic4,5.\n\nFurthermore, hematologic manifestations of celiac disease consist of anemia (due to iron, folate or vitamin B12 deficiency), coagulopathy (due to vitamin K deficiency) and, very rarely, leukopenia and thrombocytopenia. Unlike in classic celiac disease, serum folate, vitamin B12 and ferritin can be normal in silent2. Our patient had no folate, vitamin B12, iron or vitamin K deficiency, thereby further indicating the presence of silent celiac disease. In addition, the anemia present in our case was considered to be due to the autoimmune destruction of red blood cells. Our patient was female, and therefore at higher risk of celiac disease, and the only symptoms at presentation were generalized fatigue, shortness of breath and mild jaundice. The gold standard test for diagnosis of celiac disease is a duodenal biopsy, which shows the characteristic finding of intraepithelial lymphocytosis, crypt hyperplasia and villous atrophy. As far as the serological tests are concerned, sensitivity of both endomysial antibodies and anti-tissue transglutaminase antibodies (TTGA) is greater than 90%, with IgA tissue transglutaminase antibodies (TTGA) being more highly ranked4,6. In our case, we were unable to perform a confirmatory endoscopic biopsy as the patient did not consent to the procedure; therefore celiac disease was solely diagnosed from IgA tissue transglutaminase (TTGA) and IgG tissue transglutaminase (TTGA) levels.\n\nEvans syndrome (ES), a rare condition affecting only 0.8% to 3.7% with either immune thrombocytopenia purpura (ITP) or autoimmune hemolytic anemia (AIHI), is diagnosed only after eliminating all other possibilities. Hence, other possible differentials for immune cytopenias such as systemic lupus erythematosus (SLE), acquired immunodeficiency syndrome and autoimmune lymphoproliferative disorders should be ruled out prior to diagnosis2,3. In our case, the patient had negative SLE workup and her HIV status was non-reactive. However, despite a negative coombs test from the blood sample, our patient later had a positive coombs test from a bone marrow sample indicating autoimmune hemolytic anemia (AIHA).\n\n\nConclusion\n\nOverall, the association of celiac disease with Evans syndrome (ES) is very rare. To our knowledge, only one study has reported 2 adult cases of patient having celiac disease and autoimmune hemolytic anemia (AIHA), with only one of the cases having both autoimmune hemolytic anemia (AIHA) and thrombocytopenia in the entire literature7. Despite a rare coexistence of celiac disease and Evans disease (ED), we advise Coombs test (on both the blood and bone marrow sample) in all patients of celiac disease presenting with anemia with normal hematinics.\n\n\nConsent\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": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nMubarak A, Wolters VM, Gmelig-Meyling FH, et al.: Tissue transglutaminase levels above 100 U/mL and celiac disease: a prospective study. World J Gastroenterol. 2012; 18(32): 4399–4403. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYarali N, Demirceken F, Kondolat M, et al.: A rare condition associated with celiac disease: Evans syndrome. J Pediatr Hematol Oncol. 2007; 29(9): 633–635. PubMed Abstract | Publisher Full Text\n\nMichel M, Chanet V, Dechartres A, et al.: The spectrum of Evans syndrome in adults: new insight into the disease based on the analysis of 68 cases. Blood. 2009; 114(15): 3167–3172. PubMed Abstract | Publisher Full Text\n\nGreen PH, Cellier C: Celiac disease. N Engl J Med. 2007; 357(17): 1731–1743. PubMed Abstract | Publisher Full Text\n\nCollin P, Kaukinen K, Mäki M: Clinical features of celiac disease today. Dig Dis. 1999; 17(2): 100–106. PubMed Abstract | Publisher Full Text\n\nVolta U, Villanacci V: Celiac disease: diagnostic criteria in progress. Cell Mol Immunol. 2011; 8(2): 96–102. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMiller DG: Coeliac disease with autoimmune haemolytic anaemia. Postgrad Med J. 1984; 60(707): 629–630. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "49207",
"date": "17 Jun 2019",
"name": "Lucia Terzuoli",
"expertise": [
"Reviewer Expertise Autoimmunity diseases laboratory."
],
"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 report is an interesting work, which adds another case to these rare comorbidities. The only point, which in my opinion, should be improved, is the bibliographical note no. 2 in the Introduction. The authors speak of \"Several conditions such as determinatis herpetiformis, autoimmune thyroiditis and type 1 diabetes mellitus reported with CD, but in literature the bibliography refers to Evans syndrome. Perhaps it would be more correct to add other types of citations, such as, for example:\nKotze LMDS et al. Immune mediated diseases in patients with celiac disease and their relatives: a comparative study of age and sex. Arq Gastroenterol 2018; 55:346-3511. Ferrari SM et al. The association of other autoimmune diseases in patients with Graves’ disease (with or without ophtalmopathy): review of the literature and report of a large series. Autoimmun Rev 2019; 18:287-2922. Nederstigt C et al. Associated auto-immune disese in type 1 diabetes patients: a systematic review and meta-analysis. Eur J Endocrinol 2019; 180:135-1443.\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": "4700",
"date": "18 Jun 2019",
"name": "Uzair Yaqoob",
"role": "Author Response",
"response": "Thank you for your valuable comments"
}
]
},
{
"id": "50760",
"date": "05 Jul 2019",
"name": "Kofi Clarke",
"expertise": [
"Reviewer Expertise 1.IBD 2.Celiac Disease 3.Graduate Medical 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\nCase is written in great detail.\nEvans syndrome is uncommon, and the association with celiac disease is extremely uncommon. Since the pretest probability is very low in patients with celiac disease, I would not recommend routine screening for Evans syndrome in all patients with celiac disease. It will be prudent to screen in only the appropriate clinical setting.\n\nRecommend amend the first line in the discussion section celiac disease results from gluten ingestion in genetically predisposed individuals. There is no convincing evidence of disease onset only from environmental triggers.\n\nIt will be helpful to include a line in the history on excluding any medications/supplements that can cause hemolytic anemia.\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": []
}
] | 1
|
https://f1000research.com/articles/8-181
|
https://f1000research.com/articles/8-180/v1
|
14 Feb 19
|
{
"type": "Research Note",
"title": "Massive open online courses on biomedical informatics",
"authors": [
"Cender Udai Quispe-Juli",
"Victor Hugo Moquillaza-Alcántara",
"Katherine Arapa-Apaza",
"Victor Hugo Moquillaza-Alcántara",
"Katherine Arapa-Apaza"
],
"abstract": "This study aimed to identify the characteristics of massive open online courses (MOOCs) related to biomedical informatics offered in several plataforms. We conducted an observational study on specialized MOOCs platforms to find courses related to biomedical informatics, in 2018. Our search identified 67 MOOCs on biomedical informatics. The majority of MOOCs were offered by Coursera (71.6%, 48/67), English was the most common language (95.5%, 64/67). The United States developed the majority of courses (73.1%, 49/67), with the vast majority of MOOCs being offered by universities (94%, 63/67). The majority of MOOCs were in bioinformatics (56.7%, 38/67) and data science (47.7%, 32/67). In conclusion, the MOOCs on biomedical informatics were focused in bioinformatics and data science, and were offered in English by institutions in the developing world.",
"keywords": [
"computer-assisted instruction",
"medical informatics",
"bioinformatics",
"education",
"continuing education"
],
"content": "Introduction\n\nBiomedical informatics (BI) is “the interdisciplinary field that studies and pursues the effective uses of biomedical data, information, and knowledge for scientific inquiry, problem solving and decision making, motivated by efforts to improve human health”1. BI has an important role in healthcare therefore health professionals well-trained in the use of information and communication technology are needed2.\n\nIn developing countries, a few of they have developed training programmes in biomedical or health informatics3. These programmes are short courses, Master’s programmes and even sub-specialty programmes3. However, these do not supply the need for health professionals with BI skills. The internet allows health professionals to have easy access to educative sources through platforms such massive open online courses (MOOCs)4.\n\nMOOCs are accessible through the web and open to registration for people all around the world that want to participate in higher education courses; they are recognized for their educational quality and flexibility schedules4,5. MOOCs materials are free of charge, however, in some courses one can pay to get a certificate of completion4,5.\n\nThey represent one strategy to reduce costs and enable continuous education in developing countries; especially learners with English language proficiency, computer literacy, and internet access5. This study aimed to identify the characteristics of MOOCs related to biomedical informatics.\n\n\nMethods\n\nA search of MOOCs was performed in several learning platforms, including Coursera®, EdX®, FutureLearn®, Udacity®, FunMOOC®, UniMOOC®, MiriadaX®, Alison®, Iversity®, Open2Study® and P2PU®, in order to find courses about biomedical informatics. The search was made from 31 October to 27 November, 2018. The following keywords were used: biomedical informatics, telemedicine, telehealth, remote consultation, mobile health, mHealth, eHealth, medicine technology, biomedical technology, IT Health and bioinformatics.\n\nInformation was obtained on the platform where the MOOC was hosted, data regarding the institution offering the course, and the original language. The disciplines were categorized into: Bioinformatics, Images, Clinical Informatics, Public Health Informatics and Data science (a course could approach more than one discipline). Likewise, the data of the duration of the course and its cost in dollars were also obtained.\n\nThe data obtained were analyzed in STATA version 14. The categorical results were reported by frequencies and percentages, while the numerical results were reported by measures of central tendency and dispersion, after analysis of normality using the Shapiro-Wilk test.\n\n\nResults\n\nThe analyses of the data identified 67 MOOCs offered on biomedical informatics in the world. The majority (71.64%) were offered by Coursera, followed by EdX (13.43%) and FutureLearn (13.43%). The majority of MOOCs were offered from institutes from the United State of America (73.13%). Out of these, the majority were offered by universities. The large majority of these MOOCs, were offered by the University of California San Diego, followed by Johns Hopkins University. Finally the language breakdown of MOOCs related to biomedical informatics shows that the vast majority of MOOCs ( 95.52%) were offered in English (Table 1). The details of each course are shown in the Table 2.\n\nFrom the MOOCs, disciplines related to biomedical informatics courses were analyzed. Some courses taught more than one subject at a time. The majority of these MOOCs, 56.72% (n = 38), were in bioinformatics and 47.76% (n = 32) in data science (Figure 1).\n\nThe average cost of the courses was $49, which ranged from zero cost (free) to $672. Likewise, the average length of the MOOCs considered for the review was 5 weeks (Min: 2, Max: 36), with an average activity of 3 hours per week (Min: 1, Max: 30).\n\n\nDiscussion\n\nWithin the educational platforms, Coursera® offered the majority of courses focused on BI, as showed in a previous study of health and medicine5; authors prefer to upload their content more often in Coursera® because this is the most used platform6.\n\nRegarding the countries, The United States, China, and the United Kingdom develop the majority of courses. There appears to be a correlation between countries that generate more MOOCs and those with higher scientific output7. It’s important to consider that Russia appears among the leading developers of MOOCs; this could be explained by the student exchanges that the Russian educational institutions have been promoting8.\n\nThe majority of courses were offered in English, with a few having subtitles in other languages such as in previous studies9. A possible explanation for this might be the vast majority of them were made in an English-speaking country. Another explanation could be the development level of BI in these countries.\n\nMost courses approached both bioinformatics and data science, maybe because both are tools to personalized medicine and this was been a growing field in the last few years10. Therefore it is necessary to develop more courses focused on health informatics.\n\nThis study has some limitations, such as such as only English language courses being included, and the incomplete coverage of all MOOC platforms. However, the platforms studied are those that have the most health or medicine courses5. This is the first study that has assessed MOOCs in the area of BI. In addition, the data shows a list with all the names, languages and prices of the courses.\n\nThe recommendation of this study is to diversify the BI courses into other disciplines. We suggest further studies in this area that focus on evaluating the quality of MOOCs.\n\n\nConclusion\n\nThe majority of MOOCs on Biomedical informatics were focused in bioinformatics and data science, they were offered in English by institutions in the developing world.\n\n\nData availability\n\nUnderlying data is available from figshare\n\nFigshare: Dataset 1. Data base of Massive Open Online Courses on Biomedical Informatics https://doi.org/10.6084/m9.figshare.7582016.v211\n\nLicence: CC0 1.0 Universal",
"appendix": "Grant information\n\nThis research was funded with a contribution of Peruvian National Science and Technology Fund (FONDECYT).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nKulikowski CA, Shortliffe EH, Currie LM, et al.: AMIA Board white paper: definition of biomedical informatics and specification of core competencies for graduate education in the discipline. J Am Med Inform Assoc. 2012; 19(6): 931–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMantas J, Ammenwerth E, Demiris G, et al.: Recommendations of the International Medical Informatics Association (IMIA) on Education in Biomedical and Health Informatics. Methods Inf Med. 2010; 49(2): 105–20. PubMed Abstract | Publisher Full Text\n\nBlas MM, Curioso WH, Garcia PJ, et al.: Training the biomedical informatics workforce in Latin America: results of a needs assessment. BMJ Open. 2011; 1(2): e000233. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPickering JD, Swinnerton BJ: An Anatomy Massive Open Online Course as a Continuing Professional Development Tool for Healthcare Professionals. Med Sci Educ. 2017; 27(2): 243–52. Publisher Full Text\n\nLiyanagunawardena TR, Williams SA: Massive open online courses on health and medicine: review. J Med Internet Res. 2014; 16(8): e191. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAnggraini A, tanuwijaya C, oktavia T, et al.: Analyzing MOOC Features for Enhancing Students Learning Satisfaction. Journal of Telecommunication, Electronic and Computer Engineering. 2018; 10(4). Reference Source\n\nScimago Lab: Ranking countries. Scimago Journal Rank. [Cited on: 12-12-2018]. Reference Source\n\nZiganshin BA, Sadigh M, Yausheva LM, et al.: Developing medical education capacity in Russia: twenty years of experience. BMC Med Educ. 2017; 17(1): 24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCulquichicón C, Helguero-Santin LM, Labán-Seminario LM, et al.: Massive open online courses in health sciences from Latin American institutions: A need for improvement? [version 1; referees: 2 approved]. F1000Res. 2017; 6: 940. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlyass A, Turcotte M, Meyre D: From big data analysis to personalized medicine for all: challenges and opportunities. BMC Med Genomics. 2015; 8: 33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoquillaza Alcántara VH, Arapa Apaza KL, Quispe-Juli CU: Data base of Massive Open Online Courses on Biomedical Informatics. 2019. http://www.doi.org/10.6084/m9.figshare.7582016.v2"
}
|
[
{
"id": "44479",
"date": "15 Feb 2019",
"name": "Edward Meinert",
"expertise": [
"Reviewer Expertise Digital health",
"eLearning"
],
"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 have developed a manuscript examining the use of biomedical informatics in current MOOC implementations. This contribution could be very useful for those looking to examine MOOC implementations in this subject domain.\nThere are a few areas I believe the authors could further develop to improve the work:\n1. While the manuscript introduces the subject matter, there is not background for why the course of investigation is merited. There is most likely a clear rationale for this investigation, the authors could improve the manuscript to provide it.\n2. Methodologically the authors have not provided details on data extraction methods nor how their keywords were classified and harmonised among the various MOOC platforms. Additionally, to be medical literature, it would have been useful to use a systematic search method applied with this context as this would be more familiar to readers.\n3. The authors examine cost, number of implementations and further classifications. In their own right, these are big areas and the authors would do well to consider analysis and comparison of factors of economic impact, learning design, etc - the current discussion and analysis is only approaching these areas at a superficial interpretive level.\nI recommend the authors address the rationale, strengthen the methodological approach and pick particular areas of analysis to strengthen the richness of this manuscript.\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? 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? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "44480",
"date": "25 Feb 2019",
"name": "Tharindu Liyanagunawardena",
"expertise": [
"Reviewer Expertise Learning Technology",
"MOOCs",
"OERs",
"Accessibility"
],
"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 aimed to identify the characteristics of MOOCs related to biomedical informatics and it was an interesting read. However, there are some serious issues in the presentation of this work. I find some of the sentences difficult to parse. For example, Introduction - second sentence, Introduction - second paragraph, first sentence. Overall, I think a professional proof reading would improve the article's readability considerably.\n\nThis study has searched for MOOCs using many well-known MOOC platforms. It would have improved the search had the search included MOOC aggregation service for searching. In the Results section it is claimed that \"67 MOOCs offered on biomedical informatics in the world\" - I think this needs to account for the methodology and scale back to what the search was - selected platforms. It is a little confusing to hear about the \"average cost of courses\" given that the searches were on MOOCs and whether these were fees to access the course materials or for certification. The authors need to acknowledge that major non-English MOOC platforms such as Edraak and XuetangX were not consulted in this search when making statements about course languages. \"However, the platforms studied are those that have the most health or medicine courses5\" - can this statement be substantiated? \"This is the first study that has assessed MOOCs in the area of BI\" - can this be substantiated?\nI find the conclusion \"The majority of MOOCs on Biomedical informatics were focused in bioinformatics and data science, they were offered in English by institutions in the developing world\" not supported by the evidence and misleading.\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? I 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/8-180
|
https://f1000research.com/articles/8-179/v1
|
14 Feb 19
|
{
"type": "Research Article",
"title": "Oral mucositis in patients undergoing radiotherapy for head and neck cancer: An observational cross-sectional study",
"authors": [
"Marwa Ayad Nouri Al-Qalamji",
"Khudair Jassim Al-Rawaq",
"Dalya Saad Abbood Al-Nuaimi",
"Ali Ghalib Mahmood Noori",
"Khudair Jassim Al-Rawaq",
"Dalya Saad Abbood Al-Nuaimi",
"Ali Ghalib Mahmood Noori"
],
"abstract": "Background: Radiotherapy (RT) is indispensable in the treatment of head and neck cancer (HNC). Oral mucositis (OM) is a complication in HNC patients undergoing RT. This study aimed to identify the incidence, distribution of OM, and its effect on treatment breaks in a section of HNC in patients in Iraq. Methods: This is an observational, descriptive cross-sectional study. In total, 50 patients with primary HNC, treated with external beam RT, from 30th April to 10th September 2017 at Baghdad Radiation Oncology and Nuclear Medicine Center were included in the study. Cases of OM were graded according to the World Health Organization scale. Results: 80% of patients were below 65 years, and the male/female ratio was 2.6:1. Tumor sub-sites were nasopharynx (36%), larynx (22%), parotid (14%) and tongue (12%). 74% were smokers during or before starting RT. 86% were in stages III or IV. Incidence of OM was 72%; 16% grade I, 40% grade II, and 16% grade III. OM occurred in 93% females and 64% males, and 79% received concurrent regimens. Conclusions: OM occurred in HNC treated by RT, more in females, who received chemotherapy plus RT, and those with tumors of the oral cavity and nasopharynx. OM-related unplanned breaks may interrupt treatment schedule. HNC imposes a double burden in Iraq as it attacks a productive age group, and the vast majority of the patients included in this study were diagnosed in advanced stages.",
"keywords": [
"Head and neck cancer",
"Oral mucositis",
"Concurrent chemo-irradiation"
],
"content": "Introduction\n\nGlobally, head and neck cancer (HNC) accounts for 550,000 cases and 380,000 deaths, annually1. In the US, HNC accounts for 3% of cancers2, and in Europe, it was estimated to be 4% in 20123. Men are affected more than women with a ratio of 2:1 and 4:1 for US and Europe, respectively. In Iraq, the incidence of HNC is <2%4. The primary causes of HNC are tobacco and alcohol, and viral infections including Epstein-Barr virus and hepatitis virus5–7.\n\nRadiotherapy (RT) plays a central and evolving role in the treatment of HNC8 and is used with or without chemotherapy (CT) as a definitive or adjuvant treatment8. The oral cavity is susceptible to direct and indirect toxic effects of cancer CT and RT9. This risk reflects high rates of cellular turnover for the lining mucosa, a diverse and complex micro-flora, and trauma to oral tissues during normal oral function10–12. Oral mucositis (OM) is a debilitating side effect of RT13 and is exacerbated by concomitant CT14, which can begin 1–2 weeks after initiation of RT as asymptomatic erythema often progressing to erosion and ulceration. The ulcers are painful, covered by a white fibrinous pseudo-membrane, associated with dysphagia and decreased oral intake15,16.\n\nRadiation therapy is an important method used in the treatment of head and neck cancers and, like all other methods used in treatment, it is not without the side effects of treatment. The injuries that occur to mucous membranes of the oral cavity are only part of those effects, which we can avoid and reduce them before and during and after treatment by adhering to the recommendations of the treating physician and the work of therapeutic planning. This study aimed to identify the incidence, distribution of OM, and its effect on treatment breaks in a section of HNC in patients in Iraq.\n\n\nMethods\n\nThis is an observational, cross-sectional study for HNC treated by external beam radiation therapy (EBRT), which included patients 50 patients, who were patients from 30th April to 10th September 2017.\n\nThe study was conducted at Baghdad radiation Oncology and Nuclear Medicine center, Bagdad, Iraq.\n\nPatients who fit the eligibility criteria during the study dates who were scheduled for treatment were included in the study.\n\nDose of EBRT: 50-70 Gray, with a standard fractionation. Each fraction is 2Gy and 5 fractions per week. Radiation delivered with 3D conformal technique, using Elekta infinity, and Elekta synergy machines.\n\nEligibility criteria of patients was: primary HNC; T3 or T4 disease; positive nodes; residual disease; positive margins; perineural invasion; lymphovascular infiltration; extracapsular extension; treatment as described above.\n\nPatients were excluded who had comorbid conditions, were treated in a palliative way, and those with metastasis or with a bad performance status.\n\nWe conditionally collected data from patient files when they attended follow-up at the in-patient or/and out-patient clinic, or when these patients made visits to our center.\n\nVariables collected:\n\nVariables collected included: patient’s gender, age and smoking status; tumor histopathology, stage, grade, subsites, and primary or metastases; radiotherapy dose, fractions, interpretations and oral mucositis onset.\n\nAssessment by World Health Organization’s scale of OM was performed as follows: Grade 0, no OM; Grade I, soreness; Grade II, erythema, ulcers, able to eat solids; Grade III, Ulcers, liquid diet only; Grade IV, alimentation not possible.\n\nThe collected data was categorized and analyzed by T-test to identify the incidence of OM and its distribution. SPSS IBM version 22 was used.\n\nWritten informed consent was obtained from the patients for the publication of their data in this article, and the study was conducted according to the ethical standards established by the 1964 Declaration of Helsinki. The Medical Ethical Committee of Baghdad University approved this study (code:611) on 18/04/2017.\n\n\nResults\n\nIn the total patient population 72% were men, while 28% were women; male/female ratio was 2.5:1. The mean age was 53.3±11 years, and majority (38%) were between the ages of 55 and 64 years. Patients aged 65 years or more composed 20% of the total population (Table 1). In total, 76% received CT before or concurrent with RT. The vast majority (86%) had advanced stages III or IV of cancer. 74% were smokers, during or before starting RT (Table 1); 86% of these were men, while 43% were women. Seven sub-sites were observed. The highest was the nasopharynx, followed by the larynx, parotid, and then the tongue (Figure 1).\n\nIn total 72% of patients had an incidence of OM, with no patients with grade IV; 40% were grade II, and grades I and III appeared in 16% of patients (Figure 2). OM occurred in 100% of young patients, below 35 years, 50% among 45–54 years old, and 90% in patients ≥ 65 years. 93% of women developed OM compared to 64% of men (Table 2). OM appeared in 50% of patients treated with RT only, and 79% of those treated with concurrent protocols (Table 2). Grade III OM occurred in 18% of patients treated with RT and CT, while it occurred in 8% with RT only. OM appeared in 42% of patients with stage II cancer, and 77% and 75% of those with stages III and IV. OM occurred among 73% smokers and 69% nonsmokers. But severe OM (≥ grade III) occurred in 26% of smokers and 11% of non-smokers (Table 2). In this study, the majority of OM cases (47%) came from nasopharynx tumors (Table 2).\n\nRT, radiotherapy; CT, chemotherapy.\n\nIn total, 20% of patients had single breaks in their treatment schedule; the total break days was 20 days, giving an average of 2 days per break (Figure 3). Unplanned breaks were observed more in men, those who smoke, those with both RT and CT treatment, those with stage IV cancer, and those with grade III OM (Table 3).\n\nRT, radiotherapy; CT, chemotherapy.\n\n\nDiscussion\n\nIn most countries around the world, HNC is most common in men, and the male/female ratio ranges from 2:1 to 4:17,17; in the UK, the ratio of male/female was 2.7:1, and in Australia it was 2.6:118,19. In our study the ratio was 2.6:1, which is consistent with studies elsewhere. In a similar study, Vera et al. found the mean age of 450 HNC patients was 61.3 ±12.3 years14, and it was 53.3 ±11 years in our study. According to Cancer Research UK, 50% of HNC cases each year are diagnosed in people aged 65 and over18. Some of our patients were aged 65 years but only constituted to 20% of the study population. This difference in age incidence can be attributed to many causes, such as the rising age incidence of HNC in the West17, higher life expectancy in those countries20, differences in environment, and differences in life style21.\n\nIn the current study OM occurred in 72% of patients, while in other studies OM occurred among virtually all patients who are undergoing RT11,21. Eilers and Million consider being female, young age and elderly as risk factors for developing OM22. This agrees with our study; 93% of females had OM, and there were high rates observed in the age groups of <35 years and >65 years.\n\nIn our study, grade III OM occurred in 18% treated with both therapies, while its appearance in 8% only among those who received RT only. Lalla et al. found a high incidence of OM with primary tumors in the oral cavity, oropharynx or in the nasopharynx16. This is well observed in the current study in which all patients had primary tumors. All patients with oral cavity, tongue, maxillary sinus and post-cricoid tumor developed OM. 94% of patients with nasopharynx tumor had OM and 47% were seen in nasopharynx patients. The tongue, oral cavity, and maxillary tumors constituted 36% of OM, while parotid tumor contribution was 11%.\n\nThe main principles of treatment by radiation is to deliver the total fractionated dose without interruptions. However, in daily clinical practice unplanned treatment interruptions are inevitable. Bese et al. concluded that patients, because of moderate or severe ulcerative OM had 15.8% and 46.8% incidences of RT break, respectively23. In our study, among 36 patients with OM, 10 of them (27.8%) needed breaks. The percent of breaks was high in males (35%), smokers (30%), those who were treated with CT plus RT (30%), those with stage IV (44%), and those with grade III OM (50%). In this study, two patients had breaks before the end of the 3rd week, and four had unplanned breaks at the 6th and 7th weeks, so a total of six patients (12%) had breaks at a critical treatment time. However, the break durations were only 1–3 days, but it is important to mention that in most of the times, the break is decided by patients themselves (subjective). It is convenient to mention here the conclusion of Bonomi et al., that OM is not only painful but also decreases the patient’s willingness to continue treatment24.\n\n\nConclusions\n\nOM is an ongoing toxicity of RT, yet it still represents an important clinical challenge and causes burden to patients and caregivers. Most patients with HNC treated by radiation develop OM. The sub-site of the tumor is a main risk for development of OM. It was observed that young and old ages, combined RT plus CT, and advanced stage of tumor are associated with high incidence and severe OM. Patients with OM are at high risk of unplanned breaks in radiation. HNC in Iraq attack young and middle age people, which may lead to increases on its social and economic burden.\n\n\nRecommendations\n\n1. Using a multidisciplinary approach for oral management of HNC, before, during and after treatment.\n\n2. Provision of psychological care and support services for these types of patients.\n\n3. Education of patients and families regarding oral care.\n\n4. Encouragement and support of multi-center studies and researches.\n\n5. Raising competency of dentists, primary health care physicians and dermatologists to ensure early detection of HNC.\n\n\nData availability\n\nZenodo: Excel sheet file of 50 head and neck cancers whom suffer from oral mucositis due to radiotherapy, http://doi.org/10.5281/zenodo.254320425.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nGlobal Burden of Disease Cancer Collaboration, Fitzmaurice C, Allen C, et al.: Global, Regional, and National Cancer Incidence, Mortality, Years of Life Lost, Years Lived With Disability, and Disability-Adjusted Life-years for 32 Cancer Groups, 1990 to 2015: A Systematic Analysis for the Global Burden of Disease Study. JAMA Oncol. 2017; 3(4): 524–548. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSiegel RL, Miller KD, Jemal A: Cancer Statistics, 2017. CA Cancer J Clin. 2017; 67(1): 7–30. PubMed Abstract | Publisher Full Text\n\nGatta G, Botta L, Sánchez MJ, et al.: Prognoses and improvement for head and neck cancers diagnosed in Europe in early 2000s: The EUROCARE-5 population-based study. Eur J Cancer. 2015; 51(15): 2130–2143. PubMed Abstract | Publisher Full Text\n\nAlsarraj MA, et al.: Iraqi Cancer Board, Iraqi Cancer Registry Center. 2011.\n\nRettig EM, D'Souza G: Epidemiology of head and neck cancer. Surg Oncol Clin N Am. 2015; 24(3): 379–96. PubMed Abstract\n\nPalanianppan N, Owadally W, Evans M: Epidemiology and etiology of head and neck cancers. In Practical Clinical Oncology. 2nd Edition. Hanna L, Crosby T, Macbeth F. Cambridge University Press, UK. 2015. Reference Source\n\nKerstin M, Bruce E, Michael E: Epidemiology and risk factors for head and neck cancer. UpToDate. 2017. Reference Source\n\nYom SS: Radiation treatment of head and neck cancer. Surg Oncol Clin N Am. 2015; 24(3): 423–436. PubMed Abstract\n\nLalla RV, Brennan MT, Schubert MM: Oral complications of cancer therapy. In: Yagiela JA, Dowd FJ, Johnson BS, et al., eds.: Pharmacology and Therapeutics for Dentistry. 6th ed. St. Louis, Mo: Mosby Elsevier, 2011. Reference Source\n\nKeefe DM, Schubert MM, Elting LS, et al.: Updated clinical practice guidelines for the prevention and treatment of mucositis. Cancer. 2007; 109(5): 820–31. PubMed Abstract | Publisher Full Text\n\nElting LS, Cooksley CD, Chambers MS, et al.: Risk, outcomes, and costs of radiation-induced oral mucositis among patients with head-and-neck malignancies. Int J Radiat Oncol Biol Phys. 2007; 68(4): 1110–20. PubMed Abstract | Publisher Full Text\n\nLalla RV, Peterson DE: Oral mucositis. Dent Clin North Am. 2005; 49(1): 167–184, ix. PubMed Abstract | Publisher Full Text\n\nSonis ST, Elting LS, Keefe D, et al.: Perspectives on cancer therapy-induced mucosal injury: pathogenesis, measurement, epidemiology, and consequences for patients. Cancer. 2004; 100(9 Suppl): 1995–2025. PubMed Abstract | Publisher Full Text\n\nVera-Llonch M, Oster G, Hagiwara M, et al.: Oral mucositis in patients undergoing radiation treatment for head and neck carcinoma. Cancer. 2006; 106(2): 329–336. PubMed Abstract | Publisher Full Text\n\nMaria OM, Eliopoulos N, Muanza T: Radiation-Induced Oral Mucositis. Front Oncol. 2017; 7: 89. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLalla RV, Sonis ST, Peterson DE: Management of oral mucositis in patients who have cancer. Dent Clin North Am. 2008; 52(1): 61–77, viii. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWarnakulasuriya S: Global epidemiology of oral and oropharyngeal cancer. Oral Oncol. 2009; 45(4–5): 309–316. PubMed Abstract | Publisher Full Text\n\nCancer Research UK. 2014. Reference Source\n\nHead and neck cancer in Australia. Australian Government, Cancer Australia. 2017. Reference Source\n\nhttp://data.worldbank.org/indicator/SP.DYN.LE00.IN.\n\nElting LS, Keefe DM, Sonis ST, et al.: Patient-reported measurements of oral mucositis in head and neck cancer patients treated with radiotherapy with or without chemotherapy: demonstration of increased frequency, severity, resistance to palliation, and impact on quality of life. Cancer. 2008; 113(10): 2704–13. PubMed Abstract | Publisher Full Text\n\nEilers J, Million R: Prevention and management of oral mucositis in patients with cancer. Semin Oncol Nurs. 2007; 23(3): 201–12. PubMed Abstract | Publisher Full Text\n\nBese NS, Hendry J, Jeremic B: Effects of prolongation of overall treatment time due to unplanned interruptions during radiotherapy of different tumor sites and practical methods for compensation. Int J Radiat Oncol Biol Phys. 2007; 68(3): 654–661. PubMed Abstract | Publisher Full Text\n\nBonomi M, Camille N, Misiukiewicz K, et al.: Assessment and management of mucositis in head and neck cancer patients. Clin Invest. 2012; 2(12): 1231–1240. Reference Source\n\nAl-Qalamji MAN, Al-Rawaq KJ, Al-Nuaimi DSA, et al.: Excel sheet file of 50 head and neck cancers whom suffer from oral mucositis due to radiotherapy. F1000research. 2019. http://www.doi.org/10.5281/zenodo.2543204"
}
|
[
{
"id": "52881",
"date": "04 Sep 2019",
"name": "Pierfrancesco Franco",
"expertise": [
"Reviewer Expertise Radiation oncology",
"head and neck cancer",
"clinical oncology"
],
"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 on the bone marrow changes during RT-CT in anal cancer patients as assessed by FDG-PET performed at week 2 of treatment. The study is of interest. Bone marrow sparing approaches tend to rely on the fact that bone marrow is a static organ at risk to be avoided during planning and delivery, but this is not the case, since both CT and RT may change the relative proportion of active/whole bone marrow and the spatial distribution.\nI have some comments:\n\nIntroduction:\nI would cite HPV infection as a risk factor.\n\nMethods:\nI would not call eligibility criteria out; it is a retrospective study and hence I would rather define patients’ characteristics.\n\nNo data on chemotherapy are present (type, timing, regimen); please provide details.\n\nWith respect to RT, please provide details on the setting (definitive RT, adjuvant RT); if part of the cohort includes post-operative patients, please describe.\n\nWhich OM scoring scale did you use? Please specify.\n\nResults:\nPlease provide data on the timing of worst OM; at which week of treatment.\n\nDiscussion:\nPlease cite and discuss Franco et al. (20171).\n\nGeneral comment:\nI would suggest the authors to have their manuscript revised by a native speaker. Language needs to be improved.\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?\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": "54828",
"date": "29 Oct 2019",
"name": "Tatsuhiko Miyazaki",
"expertise": [
"Reviewer Expertise Pathology",
"Oncoloy and Genomic 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\nIn this manuscript entitled “Oral mucositis in patients undergoing radiotherapy for head and neck cancer: An observational cross-sectional study”, the authors reported the cross-sectional analyses of oral mucositis of 50 head and neck cancer cases which had undergone radiation therapy. They presented 36 out of 50 cases (72%) manifested oral mucositis. Location of the tumor and gender were revealed to correlate the incidence of OM significantly, but no parameter was detected with significant difference regarding unplanned break of the therapy. The data and curation result have several matters to address.\nMajor points:\nThis manuscript shows limited impact of scientific novelty, nevertheless has a potent importance for indexing when considering the mean of regional statistical studies in Iraq.\n\nThe number of studied cases looks inadequate for the strength of statistical analyses. Even after the first publication in F1000 Research, the authors should make an effort to increase the number of analysed cases and revise the data. It might be expected that type of therapy should reveal a statistically significant difference in the incidence of OM.\n\nRegarding the scoring system of OM, the authors employed WHO scale of OM only. This scale system should be simple and easy to value but there are more precise novel scoring systems such as Oral Mucositis Assessment Scale (Toro et al., 20071). The authors should at least refer to the OM assessment system in the discussion part.\n\nIn the method part, it looks inappropriate to test the significance using simple t-test in each parameter. At least, the authors should employ the multivariate analysis which might be available on SPSS software.\n\nThere was no presented data regarding the chemotherapy (type, timing, regimen as well as actual dose value). The authors should manifest the data and show the statistical analyses regarding this point.\n\nSettings of the radiation therapy were not described in detail and also there was no data about surgery. If the patients got RT and chemotherapy only, the authors should describe it.\n\nMinor points:\nIn the introduction part, the authors described Human Hepatitis Virus as a risk factor, but in the cited references, it looks like Human Papilloma Virus should be the risk factor.\n\nFigure 1: The presentation of the pie chart looks unsuitable. Improvement of the colour usage (it’s not necessarily in colour figure) is recommended.\n\nIn the results (Table 2 and 3), it looks like smoking should be a big risk factor of incidence of HNC. The authors might be better to mention it more strongly.\n\nConclusion: It might be better to conclude in medical scientific terms.\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? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "65428",
"date": "07 Jul 2020",
"name": "Daniela Pierannunzio",
"expertise": [
"Reviewer Expertise Epidemiology",
"Statistics",
"Cancer Registries Data"
],
"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 of “Oral mucositis patients undergoing radiotherapy for head and neck cancer: An observational cross-sectional study” reported the occurrence of oral mucositis in 50 patients undergoing radiotherapy for head and neck cancer in Iraq. They estimated the incidence of oral mucositis and the related grade (assessed by WHO scale) by age, gender, type of tumor, type of therapy and smoking habit. The study is interesting but I have some suggestions to improve the clarity of the paper and the impact of the results and their interpretation:\nRisk factors: why do the authors provide specific analysis for smoker patients and not for drinkers or for those with viral infections?\n\nExclusion criteria: how many patients were excluded from 30th April to 10th September 2017?\n\nData collection: it is not clear how they collect data (they asked patients to fill in a questionnaire, they extract data from medical records, …).\n\nTable 1: there are not only patients' demographic data (change title); add p value description in the text; 86% of men where smokers and 43% of women (change phrase in the text and modify in the Discussion).\n\nOccurrence of OM by age: the suggestion is described 15-44, 45-65 and ≥65 years old.\n\nTable 2: it is not clear if p value 0.355 refers to OM or in Grade III OM for smoker/non smoker; it could be more informative to show % of OM in every type of tumor (17/18=94,4% of OM in nasopharynx patients, 1/11=9,1% in larynx patients…).\n\nFigure 3: clarify the meaning of x-axis; the meaning is, for example: 3 patients broke treatment after 6 weeks from the first RT?\n\nTable 3: are the authors sure that the only reason to break treatment is OM? In the Discussion they reported: “that in most of the times the break is decided by patients themselves”. This aspect needs to be more detailed: how many times does it happen? In which way did they break the treatment?\n\nWas OM the only reason for breaking the treatment? All differences by sub groups are not significant (add a comment in the text).\n\nRecommendations: are not related with results reported in the article (especially for n. 2 and 3).\n\nIn general, the English needs to be revised.\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? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "65431",
"date": "29 Jul 2020",
"name": "Osama Muhammad Maria",
"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 of the article titled; “Oral mucositis patients undergoing radiotherapy for head and neck cancer: An observational cross-sectional study”, investigated the oral mucositis (OM) rate, grades and effect on treatment continuation in a sample of 50 head and neck cancer (HNC) patients undergoing treatment at Baghdad Radiation Oncology and Nuclear Medicine Center, Iraq, from 30th April to 10th September 2017. The article may present an important statistical tool for Iraqi HNC patients, however, many issues should be addressed in order to improve the scientific impact of the article.\nThe population sample size looks relatively small compared to similar studies. I think, if the sample size increased, that would show significant differences for the authors’ selected variables, especially when applying more complicated significance tests other than simple T-test.\n\nAt the abstract conclusion section:\n\nThe authors mentioned OM occurred more in female patients who received chemotherapy (CT) plus radiotherapy (RT). But, they did not illustrate the chemotherapy details (type, dose, duration, and whether it was definitive or adjuvant). Please, track this throughout the article sections and include it at the study statistics.\n\nThe authors stated “OM-related unplanned breaks may interrupt treatment schedule”. The authors did not show data with significant difference for that conclusion. Again, the sample size needs to be increased.\n\nAt the introduction section:\n\nThe authors mentioned different primary causes for HNC, yet, they investigated smoking only. Are other variables investigated for the study population?\n\nThe last paragraph of the introduction section should be reformatted and attached to references.\n\nThe methods section:\n\nThe authors did not mention surgery as a treatment modality for HNC patients. This should be addressed and included at the study statistics.\n\nThe authors selected variables:\n\nPlease, state “tumor stage” instead of “stage”\n\nPlease, state “OM grade” instead of “grade”.\n\nDid you show any data for OM onset?\n\nDid you investigate viral infections as well?\n\nTreatment unplanned breaks issue needs to be discussed in detail and data sorted as well. Who decided the break? Which treatment was interrupted in “CT plus RT” group? And, for how long was it? Were the breaks caused by OM only?\n\nPlease, discuss why there was no significant difference in all groups (Table 3).\n\nI would recommend reformatting the recommendation section to reflect the study results and conclusion.\n\nTable 1: please, adjust clearly the P-value (done for which groups) and mention in the text.\n\nFigure 2: please, add the axis title for the graph, and make the graph simple.\n\nTable 2: better to change the % of total OM for the “Type of tumor” to be the % of OM within the same tumor type. Example: Nasopharynx has 17 OM patients out of 18 patients, so the % of OM in Nasopharynx patients will be 17/18*100 = 94.4 %. Apply for all then readjust the statistics for that. Also, please, mention the P-value in text.\n\nFigure 3: please, write the axis titles for the figure. Make it simple.\n\nTable 3: The P-value for the tumor stages is not clear. Please, specify.\n\nIn total, the article English language should be improved. I suggest to it be reviewed by a native English speaker.\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? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-179
|
https://f1000research.com/articles/7-1847/v1
|
26 Nov 18
|
{
"type": "Research Article",
"title": "Borneo herbal plant extracts as a natural medication for prophylaxis and treatment of Aeromonas hydrophila and Pseudomonas fluorescens infection in tilapia (Oreochromis niloticus)",
"authors": [
"Esti Handayani Hardi",
"Rudy Agung Nugroho",
"Irawan Wijaya Kusuma",
"Wiwin Suwinarti",
"Agung Sudaryono",
"Rita Rostika",
"Esti Handayani Hardi",
"Irawan Wijaya Kusuma",
"Wiwin Suwinarti",
"Agung Sudaryono",
"Rita Rostika"
],
"abstract": "Background: This study aims to describe the antibacterial and immunostimulant abilities of Boesenbergia pandurata (BP), Solanum ferox (SF) and Zingiber Zerumbet (ZZ) plant extracts to treat and prevent Aeromonas hydrophila and Pseudomonas fluorescens infection on Tilapia (Oreochromis niloticus). Methods: Tilapia (initial weight 15±2 g) were injected intramuscularly (0.1 ml/fish) with a combination of A. hydrophila and P. fluorescens at a density of 1×105 CFU ml-1 of each bacteria. Treatment trials were performed at day 7 post-injection with each combined extract, while the prevention trial was performed by including the combined extract into the diet for six and seven days prior to injection. Various combinations of extract—60 ml SF extract/kg feed with 40 ml ZZ/kg feed (SF60/ZZ40), SF50/ZZ50, BP90/SF10, and BP50/SF50—were mixed with a commercial diet and used in both treatment and prevention trials. Haematological and immunological parameters were performed every week for four weeks. Results: In prevention trials, tilapia fed SF50/ZZ50 showed a significant increase of white and red blood cells from weeks 2 to 4. Similarly, significantly increased haematocrit was also found in tilapia fed SF50/ZZ50 in the treatment trial but not in the prevention trial. However, haemoglobin of tilapia in both trials was not affected by any of the various combinations of extract in the diet. Furthermore, phagocytic, respiratory burst, lysozyme activity indexes and survival rate of fish fed with combined extracts were found to be significantly higher than controls. Moreover, the amount of pathogenic bacteria in fish that were fed combined extracts was also lower than the control and was significantly different at week 4. Conclusions: This study indicates that the addition of combined extract into feed has a positive effect on the tilapia's immune system. The SF50/ZZ50 combination appears to improve the innate immune system of tilapia to treat and prevent bacterial infections.",
"keywords": [
"Imunomodulator",
"Concoction",
"Aeromonas hydrophila",
"Pseudomonas fluorescens",
"Prophylaxis"
],
"content": "Introduction\n\nTilapia (Oreochromis niloticus) is one of the most widely cultivated fish species in Indonesia. Tilapia is a freshwater fish that can be easily cultivated1. According to Pridgeon2 and Harikrishnan et al.3, freshwater fish culture is inseparable from bacterial infections which are caused by motile Aeromonas septicaemia, furunculosis, edwardsiellosis and Aeromonas hydrophila. Further, Aeromonas species have been identified as major causative bacteria and a serious pathogen in fish4,5. In Indonesia, particularly East Kalimantan, infection of A. hydrophila and Pseudomonas fluorescens in fish results in high mortality rates of up to 60–80%. In fish, both of these bacteria cause stresses, exophthalmia, ulcers, and watery-looking organs, particularly gallbladder rupture6–8. In addition, combined bacterial infection in fish is also common, such as infections found in tilapia caused by Salmonella agalactiae and A. hydrophila9,10.\n\nTo reduce high mortalities of cultured fish, aquaculturists and researchers use antibiotics to prevent and treat infection. Nevertheless, due to concerns for maintaining eco-friendly environments, the application of antibiotics should be avoided, because they may enhance antibiotic-resistant pathogens, increase the accumulation of drugs in fish tissue and trigger immunosuppression11. Methods of controlling these infections should be developed as soon as possible because the pathogen disease type has significantly increased12, while the type of pathogen that leads to edema in the cultivation area still cannot be overcome. One of the effective and safe methods for disease control in aquaculture is by improving the defence system of the fish through the provision of natural immunostimulants13, through the use of several plant extracts.\n\nVarious plant extracts, such as Indian almond leaves (Terminalia catappa), oats (Avena sativa), oyster mushroom (Pleurotus ostreatus), nettle (Urtica dioica), sea grass (Cymodocea serrulata) and beetroot (Beta vulgaris) have been used as alternatives to antibiotics5,14–16. Plant extracts also contain levamisole13 and saponin17 which can enhance the work of nonspecific immune systems and increase the activation of phagocytosis14. Further, the plant extracts of Boesenbergia pandurata (BP) and Zingiber zerumbet (ZZ) from East Kalimantan have in vitro and in vivo antibacterial activity against A. hydrophila bacteria, while Solanum ferox (SF) has been found to be an antibacterial agent for P. fluorescens bacteria. Similarly, for the prevention and treatment of bacterial infections in tilapia, BP and ZZ are also effective for treating A. hydrophila and P. fluorescens infection8,18.\n\nThe incorporation of some extracts for the prevention and treatment of bacterial infections is likely to increase the effectiveness because some materials can work synergistically, so that the infection of both bacteria in the fish body can be controlled optimally. However, research regarding the combination of plant extracts to treat and prevent bacterial infection is limited. This study therefore aims to determine the effectiveness of the combination of three extracts (BP, ZZ and SF) to prevent and treat bacterial infections of A. hydrophila and P. fluorescens in tilapia.\n\n\nMethods\n\nIn total, 450 Tilapias (Initial weight 15 ± 2 g, age ±2.5 months, random sex) were obtained from Teluk Dalam Village in Tenggarong Seberang, Kutai Kartanegara, Indonesia. The fish were randomly distributed and assigned into five aquariums in triplicate, representing four treatments and one control. The fish were kept in the laboratory for two weeks for acclimatization in the aquarium (60×40×30 cm). Each aquarium was filled with 60 l of freshwater and the water was changed by as much as 50% every 2 days to remove remaining faeces and inedible feed. The average temperature of the water was 27°C. The feed given in the acclimation phase was a commercial feed (PT Rama Jaya Mahakam, Kutai Kartanegara East Kalimantan-Supplier) at a rate of 5% of the body weight of the fish per day. The bacteria used for the challenge test were A. hydrophila (EA-01) and P. fluorescens (EP-01), which was provided from the Aquatic Microbiology Laboratory, Faculty of Fisheries and Marine Sciences, Mulawarman University, Indonesia. To bring about bacterial challenge, a combination of bacteria at density of 105 CFU ml-1 of each bacteria was used. Each fish was injected intramuscularly with 0.1 ml of the suspension of the bacteria.\n\nThe plant materials, BP, SF and ZZ, were collected from a traditional market in Samarinda City, East Kalimantan, Indonesia. The plants were cleaned, cut and dried at 40°C for 48 hours in the oven, finely powdered and stored at -4°C for the further extraction stage. Ethanol solution (95%) was used to extract the plant materials, following a method described by Limsuwan & Voravuthikunchai19.\n\nThis treatment and prevention trials were carried out for 28 days. The treatment experiments were conducted with five combination treatments with the following stages: tilapia (average initial weight 15 ± 2 g, n = 30 fish per group, random sex) were injected intramuscularly (0.1 ml) with a mixture of A. hydrophila and P. fluorescens bacteria, each bacteria at density105 CFU ml-1. At day 7 after injection, the fish were fed with feed combined with extract as follows (ml per kg feed): P1, 60 ml SF extract/kg feed with 40 ml ZZ extract/kg feed (SF60/ZZ40); P2, SF50/ZZ50; P3, BP90/SF10; P4, BP50/SF50; and P5, fed with no additional extract (control). All fish were fed twice a day ad satiation. The remaining feed was siphoned out before the next feeding.\n\nMeanwhile, the prevention trial was performed by providing the same feeding combination and procedure for 6 days prior to intramuscular injection of the fish with 0.1 ml of mixed bacteria at day 7. After injection, feeding combination was continued until the 4th week. Haematological and immunological parameters were measured every week after the injection with bacteria until week 4.\n\nAt days 14, 21 and 28 following bacterial challenge, haematological profiles of fish (n=3 per treatment group) were observed. Fish were anesthetized using 50 mg l-1 MS 222 (Sigma Aldrich, USA) per dm3 water. The fish blood was taken through the caudal vein, using a 1 ml syringe rinsed with 10% trisodium citrate anticoagulant (fish were kept alive after blood withdrawal). Total red blood cells (RBC) (106 per mm3) and white blood cells (WBC) (103 per mm3) were determined manually using an improved Neubauer counting chamber. The number of WBC was calculated using the method of Blaxhall and Daisley20. Haemoglobin (Hb) was measured spectrophotometrically at 540 nm using the cyanmethemoglobin method17. The haematocrit (Htc %) was counted using the microcentrifuge and heparinized was used as a standard solution. Meanwhile, phagocytic activity was determined using a modification of previous methods20,21.\n\nTo obtain serum, the fish blood was taken from the caudal veins and collected in an Eppendorf tube and centrifuged at 3,000 rpm for 3 minutes. Serum was then incubated at 44°C for 20 min to activate the complement22. Serum was stored in the refrigerator at 4°C for the next antibody titre observation. Measurements of antibody titres were performed using 25 μl PBS and inserted into microplate at holes 1 to 12, with the serum being inserted into hole 1 (25 μl) and then diluted into 11 holes. A total of 25 µl of bacteria (A. hydrophila and P. fluorescens) were inserted into holes 1 to 12, the mixture homogenized, and stored for 2 hours in an incubator at 37°C, followed by storing at 4°C overnight in a refrigerator. For analysis, observing the antibody titre was carried out, indicated by the agglutination reaction in the last hole.\n\nRespiratory burst activity test was performed using nitro blue tetrazolium (NBT) reagent, using the method outlined by Secombes and Olivier23. Meanwhile, lysozyme activity was performed using a microtiter plate ELISA reader at wavelength of 520 nm, following the method described by Soltani and Pourgholam24.\n\nBoth A. hydrophila and P. fluorescens (the pathogenic bacteria) were used for challenge testing (n = 10 fish per aquarium, in triplicates per group). The survival rate (SR) and relative percent survival (RPS) of the fish were recorded on a daily basis for 4 weeks25.\n\nResults are expressed as means ± standard error (SE) and the data were analysed using SPSS version 22 (SPSS, Inc., USA). The data of WBC, RBC, haematocrit, Hb, TPC, phagocytic index, respiratory burst and lysozyme activity were subjected to ANOVA, followed by Duncan’s post hoc test to evaluate significant differences among the groups of treatments. The percentage of fish survival were arcsine-transformed. All tests were significant at P < 0.05.\n\n\nResults\n\nThe present results revealed that the total WBC count of tilapia in the treatment and prevention trials were significantly increased (P<0.05) from weeks 2–4 post-administration with combined extracts. The highest increase of WBC was found in tilapia fed with SF50/ZZ50. Similarly, total RBC and haematocrit of tilapia fed SF50/ZZ50 in the treatment trial showed a significant increase after week 2, while tilapia fed SF60/ZZ40 in the prevention trial led to a positively enhanced result from weeks 2–4. Further, haemoglobin of fish both in treatment and prevention trials were not affected by any various combination of extracts (Table 1).\n\nData shown as mean±standard deviation. Different superscript letters (a,b,c) in the same column in each variable and each treatment or prevention trial showed significantly different at P<0.05. WBC, white blood cells; RBC, red blood cells; BP, Boesenbergia pandurate; SF, Solanum ferox; ZZ, Zingiber zerumbet; SF60, 60 ml SF extract/kg feed.\n\nAll combination extracts fed to fish in the treatment (Figure 1) and prevention (Figure 2) trials increased the phagocytic index. The phagocytic index of fish fed SF50/ZZ50 in the diet, in both in treatment and prevention trials, were significantly higher than control and increased from the 2nd to 4th week of the post-challenge test.\n\nBP, Boesenbergia pandurate; SF, Solanum ferox; ZZ, Zingiber zerumbet.\n\nBP, Boesenbergia pandurate; SF, Solanum ferox; ZZ, Zingiber zerumbet.\n\nThe respiratory burst activity of infected fish fed with combination extract increased from week 2 to week 4 in the treatment trial (Figure 3). In addition, SF50/ZZ50 (ml per kg feed) combination extract resulted in a significantly different respiratory burst to other combinations of extracts and the control. Meanwhile, in the prevention test, infected fish fed SF50: ZZ50 combination extract in week 4 were significantly higher than control and other combinations of extracts (P<0.05) (Figure 4).\n\nBP, Boesenbergia pandurate; SF, Solanum ferox; ZZ, Zingiber zerumbet.\n\nBP, Boesenbergia pandurate; SF, Solanum ferox; ZZ, Zingiber zerumbet.\n\nThis study revealed that lysozyme activity of infected tilapia fed SF60/ZZ40, BP90/SF10 and BP50/SF50 combinations of extract did not show a significant increase (P<0.05) at weeks 2 and 4 in the treatment test. However, starting from weeks 2–4, the addition of SF50/ZZ50 combination extract in the diet of fish resulted in significantly better lysozyme activity (Figure 5). Meanwhile, in the prevention test at weeks 2 and 4, the lysozyme activity of tilapia fed SF50/ZZ50 was significantly higher (P<0.05) (Figure 6) than in other combinations.\n\nBP, Boesenbergia pandurate; SF, Solanum ferox; ZZ, Zingiber zerumbet.\n\nBP, Boesenbergia pandurate; SF, Solanum ferox; ZZ, Zingiber zerumbet.\n\nThe overall combination of extracts administered to treat and prevent infection by A. hydrophila and P. fluorescens may decrease the number of bacteria in the fish body until the 4th week of observation (Table 2). The bacterial density, in both the treatment and prevention trials was lower than in the control. Total bacteria of A. hydrophila and P. fluorescens in tilapia fish fed combination extract in the treatment trial decreased from weeks 2–4. Further, the lowest bacterial density in tilapia was obtained from the fish fed SF 50/ZZ 50 combination extracts in their diet, which was also significantly different (P<0.05) compared to the control.\n\nData shown as mean±standard deviation. Different superscript letters (a,b,c,d) in the same column in each treatment or prevention trial showed significantly different at P<0.05. BP, Boesenbergia pandurate; SF, Solanum ferox; ZZ, Zingiber zerumbet; SF60, 60 ml SF extract/kg feed.\n\nThe administration of extract with different combinations on tilapia injected with A. hydrophila and P. fluorescens bacteria increased the SR and RPS when compared to those not given the extracts (Table 3 and Table 4). The percentage of survival of tilapia in treatment and prevention trials with combination extracts of SF 50: ZZ 50 had the highest SR compared to the other combinations of extract.\n\nData shown as mean±standard deviation. Different superscript letters (a,b,c) in the same column in each treatment or prevention trial indicate significant differences at P<0.05. BP, Boesenbergia pandurate; SF, Solanum ferox; ZZ, Zingiber zerumbet; SF60, 60 ml SF extract/kg feed.\n\nData shown as mean±standard deviation. Different superscript letters (a,b,c) in the same column in each treatment or prevention trial indicate significant differences at P<0.05. BP, Boesenbergia pandurate; SF, Solanum ferox; ZZ, Zingiber zerumbet; SF60, 60 ml SF extract/kg feed.\n\n\nDiscussion\n\nThe number of infectious diseases caused by pathogenic bacteria such as A. hydrophila have become a pivotal concern in fish culture, causing high economic losses owing to high mortality rates5. The use of plant-based extracts as immunodulators has been applied to increase survival and immune system of fish to prevent or cure bacterial pathogen. Several plant extracts that contain active phytochemicals have been found and used as supplements in the feed of fish25–28.\n\nThe current study found that the WBC of tilapia infected by both bacteria in the prevention and treatment trials increased significantly (P<0.05), while the RBC of tilapia infected by both bacteria in the prevention and treatment trials decreased significantly (P<0.05). This result is similar to those of a previous study, which stated that the WBC increased in order to tackle the infection, while the RBC was decreased in tilapia infected with Streptococcus agalactiae bacteria29, S. iniae10, A. hydrophila and Pseudomonas sp.7. In contrast, tilapia fed with a combination of extracts SF60/ZZ40 showed a similar RBC value both in treatment and prevention trials. In addition, tilapia fed SF50/ZZ50 in treatment trial resulted the highest RBC at the end of the trial. The Hb and Htc values were unchanged during the first week of all treatments including control; the decrease in Htc and Hb values occurred in controls without extract from weeks 2–4 post-infection in the prevention and treatment trials. This result indicated that the combined administration of the extracts was capable of improving the performance of the fish immune system by producing more WBC, thus making the fish more able to suppress the growth of bacteria in the body.\n\nRBC, WBC, Hb and Htc can be used as an indicator of the blood profile in fish with respect to the innate immune defence and regulation of immunological function30. WBC are particularly responsible for providing protection or resistance to disorders caused by infectious pathogens and non-infectious factors (nutrition, temperature and handling)31. Total value of WBC also describes the health status and immune system of the fish. In addition to haematological statues, the Hb content decreases due to RBC swelling and poor Hb mobilization of the spleen and other haematopoesis organs32.\n\nBesides blood profiles, the phagocytic index, respiratory burst and lysozyme activity are good indicators for immunological status of fish during infection periods. The present results revealed that infected fish treated with a compound extract of SF50/ZZ50 showed the highest IP and increased from weeks 2–4 post-injection. These results are in line with the results of a previous study, which found that fish treated with immunostimulants usually show enhanced phagocytic cell activities33. Fish have several types of phagocytic leukocytes, which are part of WBC, in the peritoneal cavity, and various tissues. The phagocytic activity is also associated with the production of oxygen free radicals by using respiratory bursts, which are important events in bactericidal pathways in fish34,35. In addition, Secombes and Olivier23 revealed that the release of superoxide anions, hydrogen peroxide and hypochlorous acid into the phagosome and extracellular space during the respiratory burst can be considered the pivotal mechanisms involved in the bactericidal activity of macrophages.\n\nTotal lysozyme level is a tool to measure the humoral component of the non-specific defence mechanism (innate immunity), which can be used to detect infections or injections of foreign material, including bacteria36–38. The present findings determined that tilapia fed SF 50: ZZ 50 had significantly higher (P<0.05) lysozyme activity. This finding is in line with past research, stating that the lysozyme activity of Jian carp (Cyprinus carpio var. Jian)39 and large yellow croaker, Pseudosciaena crocea40 were increased after being fed with traditional Chinese medicine formulated from Astragalus root (Radix astragalin seu Heydsari) and Chinese Angelica root (R. angelicae Sinenesis).\n\n\nConclusion\n\nA combination of plant extracts was found to affect the health status of tilapia when compared with control. A combination of extracts of SF and ZZ (50:50) provides the optimum protection against bacterial infections of A. hydrophila and P. fluorescens in both prevention and treatment assays.\n\n\nData availability\n\nRaw data for Tables and Figures can be accessed on OSF, DOI: https://doi.org/10.17605/OSF.IO/A42JB41.\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\nThis research is supported by the Ministry of Research and Technology of the Republic of Indonesia for the support of research funds provided through the National Strategic Research Institutions Fiscal Year 2018, contract No. 121/UN17.41/KL/2018.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe research team would like to thank the Department of Aquaculture, Faculty of Fisheries and Marine Sciences Mulawarman University, East Kalimantan, for the support of facility and equipment during the research.\n\n\nReferences\n\nFAO, Food and Agriculture Organization of the United Nations: Fisheries and Aquaculture Statistics. 2015; Rome, Italy. Reference Source\n\nPridgeon JW, Mu X, Klesius PH: Expression profiles of seven channel catfish antimicrobial peptides in response to Edwardsiella ictaluri infection. J Fish Dis. 2012; 35(3): 227–237. PubMed Abstract | Publisher Full Text\n\nHarikrishnan R, Balasundaram C, Kim MC, et al.: Innate immune response and disease resistance in Carassius auratus by triherbal solvent extracts. Fish Shellfish Immunol. 2009; 27(3): 508–515. PubMed Abstract | Publisher Full Text\n\nCarriero MM, Mendes Maia AA, Moro Sousa RL, et al.: Characterization of a new strain of Aeromonas dhakensis isolated from diseased pacu fish (Piaractus mesopotamicus) in Brazil. J Fish Dis. 2016; 39(11): 1285–1295. PubMed Abstract | Publisher Full Text\n\nNugroho RA, Manurung H, Nur FM, et al.: Terminalia catappa L. extract improves survival, hematological profile and resistance to Aeromonas hydrophila in Betta sp. Arch Pol Fisheries. 2017; 25(2): 103–115. Publisher Full Text\n\nHardi E, Pebrianto C: Isolasi dan uji postulat Koch Aeromonas sp. dan Pseudomonas sp. pada ikan nila (Oreocromis niloticus) di Sentra Budidaya Loa Kulu Kabupaten Kutai Kartanegara. J Ilmu Perikanan Tropis. 2012; 16(2): 35–39. Reference Source\n\nHardi EH, Kusuma IW, Suwinarti W, et al.: Antibacterial activities of some Borneo plant extracts against pathogenic bacteria of Aeromonas hydrophila and Pseudomonas sp. Aquaculture, Aquarium, Conservation & Legislation-International Journal of the Bioflux Society (AACL Bioflux). 2016; 9(3): 638–646. Reference Source\n\nHardi EH, Saptiani G, Kusuma IW, et al.: Immunomodulatory and antibacterial effects of Boesenbergia pandurata, Solanum ferox, and Zingiber zerumbet on tilapia, Oreochromis niloticus. Aquaculture, Aquarium, Conservation & Legislation. 2017; 10(2): 182–190. Reference Source\n\nRijkers G, Teunissen A, Van Oosterom R, et al.: The immune system of cyprinid fish. The immunosuppressive effect of the antibiotic oxytetracycline in carp (Cyprinus carpio L.). Aquaculture. 1980; 19(2): 177–189. Publisher Full Text\n\nSumiati T, Sukenda NS, Lusiastuti A: Development of ELISA method to detect specific immune response in Nile tilapia O. niloticus vaccinated against A. hydrophila and S. agalactiae. Jurnal Riset Akuakultur. 2015; 10: 243–250.\n\nRaa R, Rorstad G, Engstad R, et al.: The use of immunostimulants to increase resistance of aquatic organisms to microbial infections. Diseases in Asian aquaculture. 1992; 39–50. Reference Source\n\nSudheesh PS, Al-Ghabshi A, Al-Mazrooei N, et al.: Comparative pathogenomics of bacteria causing infectious diseases in fish. Int J Evol Biol. 2012; 2012: 457264. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFindlay V, Munday B: The immunomodulatory effects of levamisole on the nonspecific immune system of Atlantic salmon, Salmo salar L. J Fish Dis. 2000; 23(6): 369–378. Publisher Full Text\n\nBaba E, Acar Ü, Öntaş C, et al.: The use of Avena sativa extract against Aeromonas hydrophila and its effect on growth performance, hematological and immunological parameters in common carp (Cyprinus carpio). Italian Journal of Animal Science. 2016; 15(2): 325–333. Publisher Full Text\n\nBilen S, Ünal S, Güvensoy H: Effects of oyster mushroom (Pleurotus ostreatus) and nettle (Urtica dioica) methanolic extracts on immune responses and resistance to Aeromonas hydrophila in rainbow trout (Oncorhynchus mykiss). Aquaculture. 2016; 454: 90–94. Publisher Full Text\n\nDevi KN, Dhayanithi NB, Kumar TT, et al.: In vitro and in vivo efficacy of partially purified herbal extracts against bacterial fish pathogens. Aquaculture. 2016; 458: 121–133. Publisher Full Text\n\nYuniar I, Darmanto W, Soegianto A: Effect of saponin-pods extract Acacia (Acacia mangium) to hematocrit, hemoglobin at Tilapia (Oreochromis niloticus). UNEJ e-Proceeding. 2017; 67–69. Reference Source\n\nHardi EH, Kusuma IW, Suwinarti W, et al.: Short Communication: Antibacterial activity of Boesenbergia pandurata, Zingiber zerumbet and Solanum ferox extracts against Aeromonas hydrophila and Pseudomonas sp. Nusantara Bioscience. 2016; 8(1): 18–21. Publisher Full Text\n\nLimsuwan S, Voravuthikunchai SP: Boesenbergia pandurata (Roxb.) Schltr., Eleutherine americana Merr. and Rhodomyrtus tomentosa (Aiton) Hassk. as antibiofilm producing and antiquorum sensing in Streptococcus pyogenes FEMS Immunol Med Microbiol. 2008; 53(3): 429–36. PubMed Abstract | Publisher Full Text\n\nBlaxhall PC, Daisley KW: Routine haematological methods for use with fish blood. J Fish Biol. 1973; 5(6): 771–781. Publisher Full Text\n\nWatanuki H, Ota K, Tassakka AC, et al.: Immunostimulant effects of dietary Spirulina platensis on carp, Cyprinus carpio. Aquaculture. 2006; 258(1–4): 157–163. Publisher Full Text\n\nSecombes CJ: Isolation of salmonid macrophages and analysis of their killing activity. Techniques in fish immunology. 1990; 137–154. Reference Source\n\nSecombes CJ, Olivier G: Host—Pathogen Interactions in Salmonids. In Furunculosis. Elsevier. 1997; 269–296. Publisher Full Text\n\nSoltani M, Pourgholam R: Lysozyme activity of grass carp (Ctenopharingodon idella) following exposure to sublethal concentrations of organophosphate, diazinon. 2007. Reference Source\n\nCitarasu T: Herbal biomedicines: a new opportunity for aquaculture industry. Aquaculture International. 2010; 18(3): 403–414. Publisher Full Text\n\nMadhuri S, Mandloi AK, Govind P, et al.: Antimicrobial activity of some medicinal plants againts fish pathogen. International Research Journal of Pharmacy. 2013; 3(4): 28–30.\n\nChakraborty SB, Horn P, Hancz C: Application of phytochemicals as growth-promoters and endocrine modulators in fish culture. Rev Aquac. 2014; 6(1): 1–19. Publisher Full Text\n\nSivasankar P, Anix Vivek Santhiya A, Kanaga V: A review on plants and herbal extracts against viral diseases in aquaculture. Journal of Medicinal Plants Studies. 2015; 3(2): 75–79. Reference Source\n\nSeeley K, Gillespie P, Weeks B: A simple technique for the rapid spectrophotometric determination of phagocytosis by fish macrophages. Mar Environ Res. 1990; 30(1): 37–41. Publisher Full Text\n\nBallarin L, Dall'Oro M, Bertotto D, et al.: Haematological parameters in Umbrina cirrosa (Teleostei, Sciaenidae): a comparison between diploid and triploid specimens. Comp Biochem Physiol A Mol Integr Physiol. 2004; 138(1): 45–51. PubMed Abstract | Publisher Full Text\n\nHarikrishnan R, Balasundaram C: Antimicrobial activity of medicinal herbs in vitro against fish pathogen, Aeromonas hydrophila. Fish Pathol. 2005; 40(4): 187–189. Publisher Full Text\n\nLie Ø, Evensen Ø, SØrensen A, et al.: Study on lysozyme activity in some fish species. Dis Aquat Organ. 1989; 6: 1–5. Publisher Full Text\n\nSakai M: Current research status of fish immunostimulants. Aquaculture. 1999; 172(1–2): 63–92. Publisher Full Text\n\nSharp G, Secombes C: Observations on the killing of Aeromonas salmonicida by rainbow trout (Oncorhynchus mykiss, Walbaum) macrophages. Diseases of Asian Aquaculture. 1992; 1: 379–389.\n\nSharp G, Secombes C: The role of reactive oxygen species in the killing of the bacterial fish pathogen Aeromonas salmonicida by rainbow trout macrophages. Fish Shellfish Immunol. 1993; 3(2): 119–129. Publisher Full Text\n\nSaurabh S, Sahoo P: Lysozyme: an important defence molecule of fish innate immune system. Aquac Res. 2008; 39(3): 223–239. Publisher Full Text\n\nSiwicki A, Studnicka M: The phagocytic ability of neutrophils and serum lysozyme activity in experimentally infected carp, Cyprinus carpio L. J Fish Biol. 1987; 31(sA): 57–60. Publisher Full Text\n\nDotta G, de Andrade JI, Gonçalves EL, et al.: Leukocyte phagocytosis and lysozyme activity in Nile tilapia fed supplemented diet with natural extracts of propolis and Aloe barbadensis. Fish Shellfish Immunol. 2014; 39(2): 280–284. PubMed Abstract | Publisher Full Text\n\nJian J, Wu Z: Influences of traditional Chinese medicine on non-specific immunity of Jian Carp (Cyprinus carpio var. Jian). Fish Shellfish Immunol. 2004; 16(2): 185–191. PubMed Abstract | Publisher Full Text\n\nJian J, Wu Z: Effects of traditional Chinese medicine on nonspecific immunity and disease resistance of large yellow croaker, Pseudosciaena crocea (Richardson). Aquaculture. 2003; 218(1–4): 1–9. Publisher Full Text\n\nNugroho RA: Borneo Herbal Plant Extracts as a Natural Medication for Prophylaxis and Treatment of Aeromonas Hydrophila and Pseudomonas Fluorescens Infection in Tilapia (Oreochromis Niloticus). OSF. 2018. http://www.doi.org/10.17605/OSF.IO/A42JB"
}
|
[
{
"id": "41177",
"date": "28 Nov 2018",
"name": "Vishnu K. Venugopal",
"expertise": [
"Reviewer Expertise Clinical biochemistry",
"Lipid chemistry",
"Bioactive compounds characterization"
],
"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\nAfter checking the research article “Borneo herbal plant extracts as a natural medication for prophylaxis and treatment of Aeromonas hydrophila and Pseudomonas fluorescens infection in tilapia (Oreochromis niloticus)” by Dr. Rudy et al., I reached the following suggestions to be made for the article’s acceptance:\n\nThe overall structure of the manuscript is satisfactory, though some changes are recommended.\n\nIn the introduction the mechanism of action of plant extracts and its medical importance could have been added.\n\nThe authors didn’t mention the composition of feed.\n\nThere is a possibility of residual ethanol in the sample. How can you conclude the results with this concern?\n\nThe nature and source of chemicals (Materials) used in this experiment are not mentioned.\n\nFootnotes can be much clearer and the legends used in the figure should be mentioned properly. Also, in some graphs standard deviation is missing.\n\nGive enough information about the figures in figure legends.\n\nThe Discussion part can be much stronger.\n\nIn conclusion, the content of the manuscript has value for indexing. The mentioned suggestions can be considered and resubmitted.\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?\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": "4384",
"date": "29 Jan 2019",
"name": "Rudy Nugroho",
"role": "Author Response",
"response": "Dear Dr. Vishnu Venugopal,First of all, we would like to say thank you for your valuable review and comment. We have revised our article according to your review. We have also made some responses to your review - for the details of our responses, please see the link below:http://osf.io/vzsqe/download [direct download link]"
}
]
},
{
"id": "41178",
"date": "13 Dec 2018",
"name": "Alim Isnansetyo",
"expertise": [
"Reviewer Expertise Immunology",
"microbiology",
"fish diseases",
"natural products"
],
"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 presented the finding of Indonesian herbal extracts for preventing and treating fish diseases. This article might be indexed after several major issues are addressed:\nAre Boesenbergia pandurata (BP), Solanum ferox (SF) and Zingiber Zerumbet (ZZ) typical plants in Borneo? Are the plants not found in the other parts of Indonesia? If yes, please replace the \"Borneo\" with \"Indonesia\" in the title and throughout the article.\n\nWrite one sentence of background in the Abstract.\n\nDescribe systematically in the Abstract: how to prepare the extracts, design experiment, feed preparation, infection, data collecting (hematology, non-specific immune etc.) and data analysis.\n\nWrite the exact concentration for the extract in mg/kg feed instead of ml/kg. Using units of ml/kg feed is not appropriate as the exact concentrations are not known.\n\nWrite systematically the results in the Abstract as described in the Methods.\n\nThe units are written inconsistently: format (.../...., .... per ...., .... ....-1 ).\n\n“Antibody titre” is a term to evaluate the effect of vaccines. To evaluate the effect of immunostimulants, we should use the term \"Natural Agglutination\" as we are not evaluating the specific antibody. No data are presented for Antibody titre/Natural Agglutination, even though this parameter is described in the Methods.\n\nThe authors are confused by the terms of phagocytic activity and phagocytic index. Phagocytic index is not described before either in the Abstract or Materials and Methods. However, the authors describe phagocytic activity in Materials and Methods. Phagocytic index and phagocytic activity are two different parameters. Please refer to some of the recommended references. Add deviation standard for each bar in all graphs.\n\nAdd the notation in each bar of all graphs and values in tables to show insignificant or significant difference based on DMRT test results.\n\nDiscussion: please interpret properly and add additional explanation about why the extracts affect the immune system of fish and increase the SR and RPS. Describe the possible constituents in the extracts by citing the previous publications.\n\nAdd these references in the Introduction, Materials and Methods, and Discussion: Yudiati et al. (20161), Isnansetyo et al. (20162) and Isnansetyo et al. (20153).\n\nSome grammatical errors were found.\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?\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": "4385",
"date": "29 Jan 2019",
"name": "Rudy Nugroho",
"role": "Author Response",
"response": "Dear Dr. Alim Isnansetyo,Thank you for your valuable comment on our article. We have added some important information based on your comment and suggestions. We really appreciate it. For further details of our responses to your comment, please download the details of our responses from this link below:https://osf.io/qwy8z/download [direct download link]"
}
]
}
] | 1
|
https://f1000research.com/articles/7-1847
|
https://f1000research.com/articles/8-67/v1
|
17 Jan 19
|
{
"type": "Research Article",
"title": "A physiological examination of perceived incorporation during trance",
"authors": [
"Helané Wahbeh",
"Cedric Cannard",
"Jennifer Okonsky",
"Arnaud Delorme",
"Cedric Cannard",
"Jennifer Okonsky",
"Arnaud Delorme"
],
"abstract": "Background: Numerous world cultures believe channeling provides genuine information, and channeling rituals in various forms are regularly conducted in both religious and non-religious contexts. Little is known about the physiological correlates of the subjective experience of channeling. Methods: We conducted a prospective within-subject design study with 13 healthy adult trance channels. Participants alternated between 5-minute blocks of channeling and no-channeling three times while electroencephalography (EEG), electrocardiography (ECG), galvanic skin response (GSR), and respiration were collected on two separate days. Voice recordings of the same story read in channeling and no-channeling states were also analyzed.\n\nResults: The pre-laboratory survey data about demographics, perception of the source, purpose and utility of channeled information reflected previous reports. Most participants were aware of their experience (rather than in a full trance) and had varying levels of perceived incorporation (i.e. control of their body). Voice analysis showed an increase in voice arousal and power (dB/Hz) differences in the 125 Hz bins between 0 and 625 Hz, and 3625 and 3875 Hz when reading during the channeling state versus control. Despite subjective perceptions of distinctly different states, no substantive differences were seen in EEG frequency power, ECG measures, GSR and respiration. Conclusions: Voice parameters were different between channeling and no-channeling states using rigorous controlled methods, but other physiology measure collected were not. Considering the subjective and phenomenological differences observed, future studies should include other measures such as EEG connectivity analyses, fMRI and biomarkers.",
"keywords": [
"trance channeling",
"mediumship",
"anomalous information reception",
"electroencephalography",
"electrocardiography",
"galvanic skin response",
"voice analysis",
"spirit possession"
],
"content": "Introduction\n\nChanneling has been defined as: “The communication of information to or through a physically embodied human being, from a source that is said to exist on some other level or dimension of reality than the physical as we know it, and that is not from the normal mind (or self) of the channel”1. Numerous world cultures believe channeling provides genuine information, and channeling rituals in various forms are regularly conducted in both religious and non-religious contexts2–4. Research suggests that channeling-related phenomena continue to be prevalent in contemporary cultures5. In a recent survey of 899 people in the United States, 19.6% of respondents endorsed that they “Had a non-physical source from a different level or dimension of reality use your body as an instrument for communication?”6. Trance channeling can be understood as a form of channeling in which an individual willingly enters degrees of trance-like states of consciousness whereby the channel connects with sources of information that appear to exist outside of their ego-awareness. Trance channels use their body as a “vehicle” for the purported disincarnate “being” to incorporate into and to communicate directly via speaking, writing, or movement. Religious groups such as the Spiritists in Brazil7 and Spiritualists in the United Kingdom8 engage in full-trance channeling as part of their traditions and provide training programs on how to channel.\n\nThere is a paucity of scientific information on channeling, what it is, and how it works. This may be due in part to a number of challenges to studying channeling, such as variability in channeling type, information source, and content. One of the most comprehensive works on the topic is by Jon Klimo who describes some of these variable components1. For example, he describes many types of subjective channeling experiences:\n\nmental - intuitive, telepathy, clairaudience, clairvoyance, clairsentience\n\nautomatism - a variant of conscious, but which includes kinesthetic expressions of automatic writing, Ouija board movement, or pendulum movement\n\nfull-trance – purported disincarnate being incorporates into channel’s body to communicate\n\nsleep and dream – channeling occurs during sleep and channel recalls information\n\nWhile the purported source and content of channeled information are variable, common sources and overarching themes have been noted1. Purported sources include the “higher self,” deceased human beings, gods and/or God, a universal mind, collective unconscious, group beings, Jesus Christ, angels, devas/elementals, plants or animals, extraterrestrials, or earthbound spirits. Deceased humans are the largest reported category of source information. Common content themes are “ageless wisdom,” guidance and personal messages, descriptions of life in non-physical realms, the past and/or future, artistic/creative or scientific/technological material, health and healing, and information from or about deceased humans9. A recent qualitative study of channeled material collected during a focus group of trance channels revealed five common content themes similar to the themes mentioned in Klimo’s book: 1) mechanisms of channeling; 2) the need to awaken humanity and methods by which to do so; 3) the nature of reality; 4) descriptions of multi-dimensional beings and worlds; and 5) suggestions for advancing channeling research10.\n\nWell-known trance channels primarily channel one being (e.g. Esther Hicks and Abraham, Jane Roberts and Seth). However, not all trance channels report channeling only one purported source. A recent preliminary study observed multiple types of channeling in five trance channels with various purported beings communicating through the same channel. The most common was the traditional subjective experience where one person channeled one “being.” There were also several instances where multiple channels (up to 4) channeled different purported “beings” simultaneously. In these cases, more than one channel expressed that they were going to begin channeling and would then do so simultaneously. The purported incorporated “beings” would then have a conversation with each other and with the group. There were two attempts at materialization of a higher dimensional “being” and two instances where multiple channels channeled the same “being” sequentially10.\n\nResearch on channeling has not yet revealed whether channeling is a unique state of consciousness (pathological or otherwise). Three types of dissociative disorders expressing these phenomena have been defined by the Diagnostic and Statistical Manual of Mental Disorders 5th edition (DSM-V)11: Dissociative amnesia, depersonalization disorder and dissociative identity disorder (DID). DID is defined as a personality disorder, when two or more distinct identities or personalities are present, each with its own pattern of perceiving, relating to and thinking about the environment and self11. Pathological dissociation is often associated with historic physical, emotional and sexual abuse12–14. Dissociative states are also prevalent in a number of other psychiatric disorders, such as Post-traumatic stress disorder (PTSD)12,15, Attention deficit disorder12, schizophrenia, and anxiety disorders16, and are more prevalent in nonclinical populations at younger ages15. Mixed results associate dissociative possession experiences with childhood or adult trauma; some preliminary studies show that the prevalence of possession and paranormal experiences is related to childhood trauma17, whereas others do not18. However, dissociative states exist on a continuum in the general population19–21, from non-pathological expressions such as highway hypnosis and day-dreaming, to pathological states of derealization (surrealness), and depersonalization (absence of identity)14. Although trance states are currently considered a symptom of dissociation16,20,22, trance mediums do not show higher rates of pathological dissociation than the general population23,24. Most researchers would define mediums as people who communicate with discarnate or deceased personalities on a regular basis25 while trance mediums do so while under trance. Roxburgh and Roe surveyed 233 mediums and spiritualists in the United Kingdom and found no significant difference between mediums and non-medium spiritualists on the Dissociative Experience Scale. In fact, mediums scored significantly higher on psychological well-being and lower on psychological distress in comparison to non-mediums8. Although, Negro, Palladino-Negro and Louza surveyed 110 participants of a Kardecist Center in Brazil and reported that mediumship was associated with dissociation, but not with high-level pathological dissociation7. In general, most mediums and channels do not have dissociative symptoms in their daily lives above clinical cut-off scores8,10,23,24,26.\n\nA few studies have evaluated the veracity, information source and physiology of mental channels rather than trance channels27–30. Mental channels report that they communicate with deceased human beings or other discarnate entities mentally rather than the purported being using their body to communicate directly. Decreased frontal electroencephalography (EEG) power was associated with accurately discerning between photographs of deceased or living humans where visual processing as indexed by the early visual activity in the parieto-occipital right cortex differed between correct versus incorrect responses31. While we may never know if the mental channel is actually contacting deceased humans or tapping into some telepathic reservoir of knowledge32, we can investigate the cognitive and physiological processes by which people access information that is not available to them through conventional means.\n\nCan research reveal how channeling could work even if one hypothesized that channels were in fact relaying accurate information from purported discarnate “beings”? Very few studies have evaluated trance channels as opposed to mental channels. As far as we know, the first recorded trance channel study with EEG was by Hughes et al. in 1990. They found large, statistically significant increases in the amount and percentage of beta, alpha, and theta brainwave activity during channeling sessions in 10 trance channels33. The famous channel JZ Knight was evaluated before, during, and after channeling. The researchers found large increases from baseline in electromyography, skin temperature, heart rate, blood volume pulse, and electrodermal activity; and increased variability in sympathetic activation1. The details of this study’s methods and analysis are missing so it is uncertain how reliable the findings are. A recent systematic review evaluated trance channels during sessions and found an increase in noradrenaline levels, muscle tone, heart rate, and in spectral power in alpha, beta, and theta EEG frequency bands26. Another study examined EEG before, during, and after channeling in 10 experienced full-trance channels and 10 non-channel controls, all recruited from the same Spiritist cultural community25. The study found that channels compared to the controls had greater beta EEG power in all phases of the experiment, greater theta power on one electrode out of 22 while communicating, and greater alpha power on one electrode during the post-communication phase. This result was not corrected for multiple comparisons and no within-group differences were noted. Another study of possession trances in Bali, Indonesia found that theta, alpha 1 and 2, and beta signals were significantly increased compared to control periods34. Thus, while reported changes noted in EEG measures may be spurious, they are useful as preliminary efforts to study physiological differences between channeled trance-state and non-channeled states. There have been limited studies on other physiological measures. Bastos and colleagues conducted another study evaluating blood pressure, heart rate variability (HRV), and other neuroendocrine markers. All these measures except for the neuroendocrine markers increased immediately post-channeling35. Studies of dissociative symptoms and physiology contribute to our understanding of channeling in general and to the unique nature of the channeling state specifically, yet more research is needed.\n\nAnecdotally, many channels have reported unusual sensory or energetic sensations during channeling sessions. Thus, another approach to studying channeling is to incorporate objective measures that may be sensitive to subtle environmental effects associated with shifts in consciousness. In this regard, random number generators (RNG) have been used to study intentional and attentional-related effects in many laboratory and real-world experiments36–38. Our recent study found a statistically different RNG measure during channeling sessions compared to the control periods10.\n\nThese limited preliminary studies have begun to examine the veracity of channeled information, the source of the channeled information, and channeling physiology, but more research is needed on these and other aspects of channeling. Because these experiences have been so foundational to massive movements in human society, such as religions and spiritual traditions, involved in major discoveries and insights, and are relied on by many people to guide everyday decisions (such as listening to God, guardian angels, spirit guides), the mechanisms underlying the subjective experience of channeling are worth investigating. Regardless of the source, altered states of consciousness may provide a pathway for receiving information from beyond the traditional five senses that we could access more reliably if we better understood the neurophysiological pathways mediating them.\n\nThe overall goal of this project was to assess neurophysiological correlates associated with the process of channeling. With a prospective within-participant design, the study’s objective was to evaluate neurophysiological measures in trance channels before, during and after channeling sessions to characterize correlates involved in the process of going from a baseline state to a channeling state and back again. We hypothesized that differences in EEG frequency measures would be observed between the channeling and no-channeling state, distinguishing these two states as distinct. We also hypothesized that differences in the autonomic nervous system measures of heart rate and variability measures, respiration, and electrodermal activity would be observed between the channeling and no-channeling state in the direction of sympathetic activation during the channeling state, also distinguishing the two states as distinct.\n\n\nMethod\n\nWe conducted a prospective within-subject design study with 13 adult trance channels at the Institute of Noetic Sciences laboratory. The study began January, 2018 and ended July, 2018. All participants completed a screening and data collection survey (Extended data 139), submitted a video of themselves channeling, and came to the laboratory for two separate sessions of alternating channeling and no-channeling states while connected to electroencephalography (EEG), electrocardiography (ECG), temperature, and electrodermal electrodes. The study was approved through the Institute of Noetic Sciences Institutional Review Board (IRB Protocol Number WAHH_2014_01). The grant protocol submitted to the BIAL Foundation No 72/16 represents the pre-registration of analyses included in this manuscript.\n\n13 trance channels were recruited for this study with the following inclusion and exclusion criteria (assessed by questionnaires described in Procedures below). Inclusion: Adults aged 18 years and older; self-defined full trance channels who have a consensual working relationship with disincarnate beings; able to initiate connection at will; and able to remain still while channeling. Trance channeling was defined as “The channel goes into a trance state at will (the depth of the trance may vary) and the disincarnate entity/spirit uses the channel’s body with permission to communicate directly through the channel's voice, body movements, etc. (rather than the channel receiving information mentally or otherwise and then relaying what is being received).”\n\nExclusion: Significant acute medical illness that would decrease likelihood of study completion; psychotic symptoms; Community Assessment of Psychic Experiences-Positive Scale (CAPE-P 15) - Item 3, 4, 12, or 7 >040; Dissociative Experiences Scale-T - Score ≥2041; mediums/channels who mentally relay information communicated by spirit. Because of the sensitivity of the physiological recording equipment, only participants who could remain still while channeling were recruited. Trance channels were used for this study because it was hypothesized that physiology would be affected more than in mental channeling. Exclusion criteria were chosen to ensure the participants were healthy, well-adjusted adults.\n\nParticipants completed an online survey (Extended data 139) administered with the SurveyMonkey platform and developed for this project between January 2018 and May 13, 2018. A consent was embedded into the survey. The survey collected demographic data (race, gender, socioeconomic status42,43), inclusion/exclusion criteria information, frequency and characteristics of channeling experiences, paranormal beliefs, and traits relevant to channeling experiences. These questionnaires were chosen to evaluate qualities relevant to channeling experiences. At the end of the survey, volunteers had an opportunity to fill in their contact information if they were interested in participating in the study’s next steps.\n\nThe following measures were administered as part of the inclusion/exclusion criteria assessment:\n\nDissociation Experiences Scale Taxon (DES-T) indexes pathological dissociation and has been shown to differentiate between psychiatric presentations that contain dissociative symptoms and those that do not44. The DES-T is an eight-item subscale of the full-scale DES45, Cronbach ɑ of 0.75 and is significantly correlated to the larger DES scale scale (r = 0.79)46. Respondents selected a percentage number (e.g., 0% to 100%) indicating the frequency that they experienced the dissociative symptom. Each item was then scored on a scale from 1 to 100 and the overall score was the mean of the eight items. The DES-T distinguishes pathological dissociation more accurately than does the full-scale DES, with a cutoff score of 20 capturing nearly 90% of cases of pathological dissociation.\n\nThe Community Assessment of Psychic Experiences-Positive Scale (CAPE-P15)47 is a self-screening questionnaire to address subclinical positive psychotic symptoms in community contexts. It is valid, reliable, has the same three‐factor structure as the lifetime version consisting of persecutory ideation, bizarre experiences and perceptual abnormalities, and is highly predictive of generalized distress (r = .52)40.\n\nThe following measures were administered to evaluate against population and clinical norms. That is, do the trance channels we recruited for our experiment differ significantly from the general or clinical populations in anxiety, depression, personality, empathy, sensitivity, absorption levels, and paranormal beliefs and experiences? These measures were also collected in order to evaluate them as predictors or moderators of any physiological changes we may have observed between channeling and no-channeling states.\n\nPatient Health Questionnaire-4 (PHQ-4)48 is a 4-item inventory rated on a 4-point Likert-type scale that is a very brief and accurate evaluation of depression and anxiety with established internal reliability, construct validity, and factorial validity48.\n\nBig Five Inventory-10 (BFI-10)49 is a ten-item scale with personality categories of extraversion, agreeableness, conscientiousness, neuroticism, and openness. Each within each category are averaged to derive category scores (Cronbach α range from 0.74 - 0.89).\n\nMultidimensional Personality Questionnaire Absorption Scale50,51 is one of the 11 component scales of the larger Multidimensional Personality Questionnaire. It has 34 true/false self-report items that assess an individual’s openness to experience, emotional and cognitive alterations across a variety of situations. Summed scores on the instrument are calculated by identifying true responses as 1 and false responses as 0, creating a possible range of 0 to 34, with higher scores indicating stronger trait absorption. Tellegen reported high levels of internal reliability (r = 0.88) and high levels of test– retest reliability (r = 0.85)52.\n\nEmpathy Quotient53 is an 8-item questionnaire that was developed to evaluate empathy in the respondent. Each item is rated on a four-point scale (Strongly disagree = 0; Slight disagree = 0; Slightly agree = 1; Strongly agree = 2) with items 5-8 being reverse-scored. The summed value is then multiplied by 2.75 for comparison to original 22-item instrument53,54.\n\nHighly Sensitive Person Scale is an 6-item self-report questionnaire that evaluates people sensitive to environment and emotions55. The unidimensional scale has alphas of 0.65–0.85 across numerous samples and high test–retest reliability. Sample items include “Are you easily overwhelmed by things like bright lights, strong smells, coarse fabrics or sirens close by?” and “Do other people's moods affect you?” Respondents reply on a 7-item Likert scale with 1 anchored at Not at All and 7 anchored at Extremely. Items are averaged for a total score.\n\nParanormal Belief and Experience Scale was developed by the Institute of Noetic Sciences because of confounds in many paranormal belief scales of belief and experience. The instrument was developed to explicitly separate paranormal belief from perceived paranormal experience. Like all explicit questionnaires, bias exists and would be best administered with implicit measures as well. The scale contained ten statements about the belief in intuition, out of body experiences, extraterrestrials, precognition, near death experience, mediumship, clairvoyance, psychokinesis, telepathy, automatism which the participant rated on a slider from Disagree Strongly (0) to Agree Strongly (100). For each of the ten items, participants also answered whether they had ever personally experienced the phenomenon on a slider scale from Never (0) to Always (100). The sliders allow for gradation of belief and experience rather than binary or Likert responses as are often included in these types of questionnaires. Six of the 10 items were taken from the Australian Sheep-Goat Scale, three exact (#’s 9, 10, 11), and three modified (#’s 4, 5, 14)56. The Cronbach α’s of the belief and experiences components are 0.90 and 0.93 respectively.\n\nOnce participants completed the survey, the team reviewed the data and identified participants who met the inclusion and exclusion criteria and expressed interest in the laboratory phase of the channeling study. Potential eligible participants were asked via email to share a five-minute video of their channeling. A member of the research team further screened the potential participants during a telephone screen to ensure they identified as a trance channel as we defined it, and that they could remain still during their channeling session. The study team chose qualified trance channels after a review of the survey data, telephone screening information, and video. Eligible participants received study instructions and summary via email. A second consent for the in-lab portion of the study was sent via email through DocuSign (DocuSign, Inc. San Francisco, CA).\n\nParticipants returned the completed consent and study staff made their travel arrangements. Travel, meals and lodging were paid for by the study. In addition, participants received a $100 gift card at the end of their visit to compensate for their time.\n\nParticipants were greeted on the day of the study and oriented to the IONS laboratory. Before being connected to equipment, the participants were asked if they needed to use the restroom, given a printed summary of the laboratory paradigm and informed that they could stop the experiment at any time. They were then seated in a 10 ft. cubed double steel-walled electromagnetically shielded room. Electrophysiological-monitoring devices were then attached to the participants.\n\nOnce all the equipment was set up and high signal quality was obtained from all sensors, the laboratory paradigm proceeded as presented in Figure 1. Paradigm A was used for Day 1 and a counter-balanced Paradigm B was used for Day 2. A counterbalanced design was used to ensure that any changes in physiology were not created by the regular alternating of the states or independent biological oscillatory components. Day 1 and Day 2 recordings were always at the same time of day for each participant. Each session consisted of two stories (one channeling, one no-channeling), three 5-minute channeling periods (Channel), three 5-minute mind-wandering periods (No Channeling), and two breaks. Sessions were always scheduled between 10 am and 5pm. First, participants read a story (Extended data 139) presented to them on a laminated paper sheet by the research assistant, and their voice was recorded using an Apple iPod (Cupertino, CA). They read “The North Wind” on the first day and the “Rainbow” on the second day. Stories were read for the voice analysis to ensure standardized content during the recordings. Participants were then instructed to keep their eyes closed and to remain as still as possible during the physiological recordings. The same story was then read again while channeling. The participants were then asked if the purported discarnate being had a message they wanted to share, which was also audio recorded and then transcribed (qualitative analysis of these transcripts are in progress and will be reported elsewhere, transcripts available as Extended data 139).\n\nChanneling and no-channeling states were counterbalanced within sessions and across days. Black dots represent research assistant trigger of marker for transition. Gray dots and dash lines represent participant trigger of marker to confirm completed transition into that state or beginning of immersion into the new state (e.g. after a break). The period between the Black and Gray dots represent transitions into and out of the channeling state. Chanl = Channeling; No Chanl = No Channeling. Note that the 5-minute timing began only after the participant triggered they had achieved the state.\n\nIn order to clearly distinguish physiology for the channeling and no-channeling states, a series of markers were inserted into the physiology file by the research assistant and the participant (see black and gray dots in Figure 1). The research assistant would verbally alert the participant to enter the channeling state. After the participant transitioned into the channeling incorporation state, they would press a button on a keyboard which would trigger a marker to note the channeling state had begun. We are calling this an incorporation state since the channel is not speaking but is purportedly incorporating a discarnate being in their body. The research assistant then asked the participant to exit the channeling state. The participant would press the button on the keyboard to trigger a marker noting that they had transitioned out of the channeling state.\n\nChanneling condition – The instructions for the channeling state were for them to go into their channeling state as they normally would except to remain as still and quiet as possible once they achieved that state.\n\nControl condition – The instructions for the no-channeling control condition were for the participant to actively think about one of three following things for each no-channeling period respectively: 1. Think about all the places you have traveled or want to travel in your life; 2. Think about walking through the grocery store for one of your usual shopping visits (i.e. imagine walking through aisles, picking out items, placing them in your cart, checking out); 3. Think about all the steps required to plant a garden and all the things you would plant. This mind-wandering task was chosen to give the participants an active non-channeling task to focus on.\n\nBreaks – During the breaks, the research assistant would enter the electromagnetically shielded box and ask the participant two questions (Figure 2): 1. How would you rate the level of consciousness during the last session? The participant would then mark on a 10 cm line anchored by 0 for “Fully conscious” to 10 for “Fully Unconscious” (knowing that the level of trance during channeling could vary by session and by channel); and 2) How would you rate the level of perceived incorporation? The participant would again mark on a 10 cm line anchored by “No incorporation, my body is not being directly used at all” to “Full incorporation, my whole body is being directly used.” The research assistant then measured the value of the mark and recorded it into the study database.\n\nTemperature and humidity collection: Ambient temperature and humidity of the room were measured with an EIVOTOR Temperature Hygrometer and recorded at the beginning of the experiment, during the breaks and at the end of the experiment.\n\nPhysiological Data Collection: All physiological data were recorded at 1024 Hz with an ActiveTwo Biosemi system (Biosemi, Amsterdam, Netherlands) that integrated EEG, ECG, and galvanic skin response (GSR).\n\nEEG data were recorded through 64 active electrode channels placed in an electrode cap at standardized electrode positions (20 international 10–20 system locations and 12 intermediate ‘‘10–10’’ locations). SignaGel (Parker Laboratories, Inc.) was used to increase impedance for each electrode and was injected at the surface of the scalp manually. Active electrode offsets (rather than impedance as used with standard electrodes) were kept below manufacturer guidelines (i.e. -/+ 20 mV).\n\nECG data were collected with electrodes placed bilaterally just inferior to the clavicles.\n\nGalvanic skin response (GSR) data were collected with one electrode on the palmar thenar eminence and the other on the dorsal aspect between the first and second digits. The GSR used two passive Nihon Kohden electrodes to induce an oscillator signal synchronized with the sample-rate. Because the BioSemi GSR uses \"Lock-in detection\", the stimulus-current can be as low as 1uA. The low-current and synchronized oscillator ensure that the biopotential measurements (ECG, EEG) are not corrupted by the GSR oscillator signal.\n\nBody temperature was collected by an electrode placed on the right index finger of the right hand (palmar side).\n\nFor all physiological measures, data was segmented as follows and averages over these time periods were calculated: Day 1 – Chanl.1_1, Chanl.1_2, Chanl.1_3, NoChanl.1_1, NoChanl.1_2, NoChanl.1_3; Day 2 - Chanl.2_1, Chanl.2_2, Chanl.2_3, NoChanl.2_1, NoChanl.2_2, NoChanl.2_3.\n\nEEG - The data was downsampled to 512 Hz with the Biosemi Decimator for analysis. All processing were conducted with the EEGLAB toolbox v. 1457 in Matlab 2014b (The MathWorks, Inc, Natick, MA, USA). After removing auxiliary channels (processed and analyzed separately) and the offset, high-pass and low-pass filters were applied (i.e. 1 and 50 Hz, respectively). The toolbox Cleanline v. 1.03 was used to filter line-noise artifacts at 16 and 48 Hz. Bad channels and portions of data were removed by visual inspection of the continuous data by experimenter CC. Extended Infomax Independent Component Analysis (ICA) was then used to identify ocular and muscle artifacts (e.g. eye blinks, lateral eye movements). We used the IClabel v. 0.3EEGLAB plugin to classify components, and selected those components which maximum probability lied in the eye or muscle artifactual component category. The continuous data was then segmented into 1-second epochs that were 1-second away from each other (to decrease correlation between neighboring epochs). Traditional frequency analysis was performed using hamming tapered FFT with 1-Hz resolution.\n\nECG - ECG data was extracted from the Biosemi BDF files using EEGLAB in Matlab and processed as previously described58. Briefly, ECG data was then imported into Kubios HRV Premium v 3.1.0 (University of Kuopio, Kuopio, Finland) to generate RR intervals, ECG-derived respiration, heart rate, very low, low, and high frequency domain measures (VLF: 0-0.04; LF: 0.04-0.15 Hz; HF: 0.15-0.4 Hz), frequency peak, absolute and normalized amplitude, and LF/HF ratio and nonlinear measures. HRV analysis parameters included a 100 s window width, 50 % window overlap; autoregressive spectrum model order = 16 with no factorization, and interpolation rate = 4 Hz. No participants were on medications that affected ECG (e.g., beta blockers and calcium channel blockers)59,60.\n\nRespiration – ECG-derived respiration was calculated with Kubios HRV Premium (University of Kuopio, Kuopio, Finland)61.\n\nGSR and body temperature – The body temperature data was down-sampled at 32 Hz and the GSR data was down-sampled to 10 Hz. After removing the offset, artifacts were rejected manually (electrode disconnecting briefly due to a movement), and the data was segmented into different conditions for each subject from the markers embedded in the data. The standard deviation of each epoch was then computed in Matlab for statistical analysis62–64.\n\nVoice analysis - Recordings were reviewed to ensure that only the participant reading the story was included in the analysis (i.e. not any conversation with the research staff). The Voice parameters were calculated with the Beyond Verbal emotion analytics application performing interface v. 5 (Beyond Verbal Communication, LTD), resulting in three measures (temper, valence, and arousal). Temper reflects a speaker’s temperament or emotional state, ranging from gloomy or depressive at the low end, to embracive and friendly in the mid-range, and confrontational or aggressive at the high end of the scale. Valence is an output that measures the speaker’s level of negativity at the lower end of the scale to positive attitude at the higher end. Arousal is an output that measures a speaker’s degree of energy, ranging from tranquil, bored, or sleepy at the lower end of the scale to excited and highly energetic at the higher end. A spectral power analysis of the entire recording period was then calculated. First an autocorrelation linear predictive coding was conducted with the following settings (Predictive Order = 16; Window Length = 0.025s; Time Step 0.005s). Then the power values were output at each 125 Hz spectral bin in dB/Hz using Praat voice analysis software (Praat 6.0.43, Phonetic Sciences, University of Amsterdam Netherlands). This method is comparable to the long-time average spectrum voice analysis that had been done to discriminate between male and female voices65.\n\nDemographics and survey data are described with means and standard deviations for continuous variables and percent endorsed for categorical variables. For temperature and humidity, values for day 1 and 2 were evaluated for statistical differences and collapsed. Values for before and after laboratory activities were evaluated with a Student’s paired t-test.\n\nFor the level of consciousness and incorporation, sessions were evaluated for statistical differences. If the same, then measure was collapsed across sessions by calculating the average. Then, the values for each Day were tested for statistical difference with a Student’s paired t-test. A test-retest parameter for Day 1 and Day 2 using Pearson’s correlation was calculated (i.e. do the measures remain the same on different days?). Day 1 and Day 2 values were then collapsed if statistically the same. All statistical analyses except the EEG were conducted in Stata 12.0 (StataCorp LLC, College Station, Texas).\n\nFor the EEG analyses, statistical tests were computed using the LIMO v. 2.0 plugin of EEGLAB, with sessions (1, 2, 3), day (day 1 and day 2), and condition (channeling vs control) as independent variables. We used general linear model weighted least square optimization between 1 and 40 Hz frequency band and used threshold-free cluster enhancement correction for multiple comparisons. For all other physiology measures, each participant’s average values were calculated for each session (1, 2, 3) on both days (Day 1, 2) such that there were the following segment data for all measures: Day 1 – Chanl.1_1, Chanl.1_2, Chanl.1_3, NoChanl.1_1, NoChanl.1_2, NoChanl.1_3; Day 2 - Chanl.2_1, Chanl.2_2, Chanl.2_3, NoChanl.2_1, NoChanl.2_2, and NoChanl.2_3. A repeated measures analysis of variance was conducted with each measure as the dependent variable, Condition (Channeling, No Channeling) as a factor, and Day (1, 2) and Session (1, 2, 3) as repeated variables. False discovery rate was used to adjust for multiple comparisons66. Scripts used for analysis are available as Extended data 267.\n\n\nResults\n\n155 persons consented to the study and 91 identified as trance channels as defined within the survey. 58 were excluded, 57 (98%) of which was due to not giving direct or indirect permission. However, this exclusion overlapped with additional exclusion criteria (CAPE, 36%; Dissociative scale, 24%; Could not sit still, 15%; Not able to initiate, 12%; and 5% did not provide contact info). 33 were invited via email to submit a video, 12 did not reply to the invitation and 21 were reviewed. Of the 21 videos the research team reviewed, six were excluded; three due to inability to comply with study design, two were not in a trance state; and one due to a scheduling conflict. Two declined to participate; one of the reasons provided was her guides did not feel comfortable participating. 13 were invited and completed the study (Figure 3).\n\nThe thirteen trance channels met all the inclusion/exclusion criteria and completed the study. They were mostly Caucasian, older women with a college education (Table 1). Dissociation, psychotic, anxiety and depression symptoms did not meet clinical cutoffs for any pathology. No participants were currently diagnosed with a psychiatric illness and only one was on a psychoactive medication. The following symptoms were endorsed for previous lifetime mental health history by the number of participants in the parentheses: major depression (1), posttraumatic stress disorder (2), anxiety disorder including phobia and panic attacks (3).\n\nMean ± standard deviation; %, percentage.\n\nThe age channeling started was varied (mean age 39 ± 21, range 4 - 65). How the experiences started also varied (spontaneously - 46.2%, received training - 30.8%, trained myself - 23.8%). Most participants did not endorse other family members having similar skills (No or unknown - 62.5%). Participants perception of the average level of consciousness was 4.7 ± 3.8 (range 0-10; 0 = Fully Conscious) and average level of incorporation was 7.1 ± 2.7 (range 0-10; 0 = No Incorporation). The impact of channeling on the participant’s life was very positive (96.1 ± 7.1; 0 - very negative to 100 - very positive). All participants gave permission for the channeling, could tell when it began, and have told others about the experience. Most initiated the channeling session. The perceived source of information was from their higher self or group beings (a group entity or group mind is described as a coherent bundle of still-individual or once-individual beings who communicate as from a single integrated source.) Guidance and personal messages was the most common purpose of the channeling and it was most often utilized by recording it. See Table 2 for percentages endorsed for each category.\n\nThere was one open text field in the survey asking if the participant could tell when the channeling was starting and how they could tell. All 13 channels entered text into this field. The following themes were reported with the number of participants who endorsed this concept after each theme (see Extended data for full text and coding): physical sensation (9); “being” initiates contact (5); channel initiates contact (3); mental information (2); emotional change (2); and energetic sensation (2).\n\n%, percent of participants who endorsed item.\n\nTemperature and humidity in the electromagnetically shielded room. Temperature and humidity were similar on Day 1 and 2 so they were combined. The temperature in the electromagnetically shielded room changed from the beginning to the end of the experiment (Before - 70.6 ± 1.4; After - 72.2 ± 1.2; t = -8.4, p < 0.00005). Humidity in the electromagnetically shielded room did not change (Before - 44.6 ± 9.8; After - 45.3 ± 10.1; t = -1.6, p = 0.14).\n\nLevel of channel’s consciousness and incorporation during channeling. The perceived level of consciousness and incorporation are listed in Table 3. Participants were in a deeper trance for each subsequent session on both days (F(2,48) = 4.37; p= 0.018). The depth of trance was not significantly different on the two days but marginally so (F(1,24) = 4.04; p= 0.056). There was no interaction of Day and Session. The Pearson’s correlation session by session between Day 1 and Day 2 (i.e. test-retest reliability) was high (r = 0.87). The perceived level of incorporation was similar across sessions and days (non-significant Day, Session, Day x Session). The Pearson’s correlation session by session between Day 1 and Day 2 (i.e. test-retest reliability) was high (r = 0.79). 7/13 participants perceived channeling the same purported being for all the sessions and six had different purported beings for the sessions.\n\nMean ± standard deviation. Consciousness scale was anchored by 0 for “Fully Conscious” to 10 for “Fully Unconscious”. Incorporation scale was anchored by 0 for “No incorporation, my body is not being directly used at all” to 10 for “Full incorporation, my whole body is being directly used.”\n\nEEG. There were no significant differences between channeling and no-channeling condition in the traditional frequency bands (theta 3–7 Hz; alpha 8–12 Hz; beta 13–20 Hz and low gamma 21–40 Hz) across the 64 channels. We observed what might be a trend in the uncorrected statistical maps at 13 Hz, but it did not resist correction for multiple comparisons (Figure 4; raw data files available from figshare, Dataset 168).\n\nECG. The average data length was 302.2 ± 18.2 seconds. The ECG data was exceptionally clean. Only 47 of the 156 segments had any artifact heartbeats. The average percentage of data with artifact heartbeats was only 2.9% ± 2.4. ECG measures (mean heart rate, SDNN, very low frequency, low frequency, high frequency) and respiration were the same in the channeling and no-channeling condition (all model p’s greater than 0.05).\n\nGSR and body temperature. There was no significant difference in the standard deviation of the body temperature or GSR between channeling and no channeling conditions (GSR: channeling 244.4 ± 134.3; no-channeling 335.8 ± 202.5; Body Temperature: channeling 3.2 ± 2.5; no-channeling 3.5 ± 4.4; all model p’s greater than 0.05).\n\nVoice. The story reading during the channeling state was significantly slower than during the no-channeling state (Story 1: no-channel - 38.2 s ± 2.5; channel - 56.0 s ± 18.6; Story 2: no-channel - 36.8 s ± 4.1; channel - 67.3 s ± 32.6; F(1,51) = 15.44, p = 0.0006). Valence was significantly lower during channeling compared to no-channeling (Table 4). There was no significant difference in arousal or temper.\n\nMean ± standard deviation; F, F-statistic; p, probability.\n\nThere were significant power (dB/Hz) differences between the story read during channeling compared to no-channeling in the 125 Hz bins between 0 and 625 Hz, and 3625 and 3875 Hz (Figure 5); see data available for values at each frequency bin; p-values for these bins varied from .03- 0.0007 and remained significant with False Discovery Rate correction).\n\n* = significantly different between channeling and no-channeling readings with False Discovery Rate correction.\n\n\nDiscussion\n\nRecruitment for the study was feasible. From the 155 persons that consented to the survey originally, thirteen met all the inclusion/exclusion criteria and completed the study (21% of those who identified as trance channels and 39% of those who were invited to submit a video). Participant demographic characteristics were similar to other studies on anomalous information reception, namely Caucasian, educated older women6,23,24. The participants also did not have any psychiatric pathologies evident through self-report of diagnoses or medications, which is similar to other studies7,8,10,23,24. All dissociative and psychotic symptoms scores were less than commonly used clinical cutoffs for pathological dissociation44 and psychosis69. Participants personality values were similar to other general healthy adults70,71. Absorption was higher than a large college-student samples which had an average of about 20 (compared to our 30)72, but lower than Brazilian spiritists73. Empathy score means were higher than other population norms53,54. Sensitivity scores reflected the channels being highly sensitive people (as opposed to non-sensitives)55,74. Paranormal beliefs and experiences were much higher than observed in 350 general population respondents (beliefs - 59.3 ± 21.7; experience - 43.7 ± 25.3) as one would expect considering their channeling experiences75.\n\nThe mean age at which participants had their first channeling experience was older than other types of anomalous information reception experiences6 and other self-reported trance channels where such data were reported10,25,35 although the standard deviation and range of the starting age was large. Participants reported a medium level of trance (4.7 out of 10) during channeling which was not as deep one might imagine because of ethnographic reports of deep trances during possession rituals76. As far as we know, no other trance channeling study has formally evaluated level of incorporation. Incorporation level was moderately high. The impact of channeling on the participants lives was very positive as has been noted in other channeling studies8,10,26 and other anomalous information reception studies6,23,24,77. The survey data reflected similar responses about channels characteristics, perception of source, purpose and utilization to other trance channel studies1,10 and reflected consensual experiences with varied purported sources and content.\n\nTemperature was approximately 1.2 degree higher after the experiment than when it began. This is likely due to the fact that people were present in the electromagnetically shielded room and resulted in a slight increase in ambient temperature. There were no humidity changes.\n\nMost channels were aware of their experience (rather than in a full trance) and reported varying levels of perceived incorporation. Trance depth increased over the sessions each day, but was not different across days and test-retest reliability of consciousness level was high. There were no differences in level of incorporation. While micro-variations may have occurred, our values reflect consistent levels of consciousness and incorporation during our laboratory paradigm for these particular participants on a statistical level. Like other trance channel studies, participants reported a variety of purported sources being incorporated10.\n\nOur hypothesis that channeling and no-channeling states would be reflected with distinct EEG measures was not demonstrated. Bastos et al. also did not find any EEG differences in beta, theta and alpha frequency bands before, during, and after mediumistic communication in ten Spiritist mediums in Brazil25. They did find EEG differences when comparing trance channels to controls that were collected simultaneously in a group setting. Increased beta was found in four electrodes pre-communication, two electrodes during communication, and one electrode post-communication, increased theta in one electrode during communication, and increased alpha power in one electrode post-communication for trance channels compared to gender and age matched controls. Bastos et al. compared the two groups by time-point (i.e. rather than using a repeated-measures model) and did not adjust for multiple comparisons. Also, the authors state that only Fp1, Fp2, F3, F4, F7, and F8 were used in the final analysis although they collected data from 22 electrodes because these areas are “widely accepted as being involved with spiritual experiences”25. Hughes et al. found greater theta, alpha and beta power in the trance state compared to pre-trance, and greater beta and alpha comparing the trance state to post-trance for ten trance channels33. The randomization test for matched pairs was used to analyzed the data rather than a repeated measures model and no multiple comparison corrections were made. The EEG was recorded during \"talking baseline\" before and after the trance period which “consisted of talking, listening and answering questions in what might be described as an \"intellectual\" or \"philosophical conversation\" mode of discourse” we assume with eyes open. This is a more naturalistic setting for the trance channels especially considering may of our participants stated that it felt strange and was unique that they would incorporate their “beings” and not say anything. We chose to have the participant stay silent with their eyes closed to ensure a clean EEG signal. Our study also had the unique design of all recordings being collected in an electromagnetically shielded environment which may have reduced additional outside “noise” to the EEG signal. Bastos et al. dealt with this issue in a different way25. Participants were asked to stay seated with eyes closed and avoid blinking during EEG collection. They did communicate verbally during the channeling period but then only EEG epochs free of muscular artifacts were used in the analysis25. Two other studies have evaluated possession trances in a naturalistic setting in Bali, Indonesia34,78 and found differences in the EEG in possession trance ceremonies compared to control conditions and/or control participants. These studies are unique in that they record the EEG during a sacred ceremony while the participant is moving. Oohashi et al. accurately summarize three main issues with in situ recordings of this type: 1) the often sacred nature of the process precludes invasive physiology recording equipment; 2) finding reliable portable devices (although this issue has improved greatly with the invention of increasingly more mobile EEG equipment); and 3) contamination of the data by artifacts. There are only so many post-recording processes that can be done to preserve high quality EEG when the participant is moving and/or receiving or processing different stimuli than are present during the control conditions (e.g. removal of artifacts and Independent Component Analysis (ICA)57,79). However, even with gross artifact removal, filtering of the EEG, and ICA, the naturalistic setting does not control for the other types of activities that could cause EEG differences during the communication such as eyes open versus eyes closed, and auditory or visual stimuli and processing. Collecting and analyzing EEG signals in trance channels is certainly a challenge and each paradigm has it benefits and limitations in adding to our understanding of how channeling may work. We can continue to work towards improving our methods in order to gain more information about the channeling phenomenon and our future studies can consider the following EEG design elements and caveats depending on the specific research question: within- versus between-participant analyses including baseline correction for differences between groups, and repeated measures statistical models, naturalistic recordings versus laboratory recordings ensuring attention to eyes open versus eyes closed states and avoidance of cross-comparison; and speaking versus being in stillness yet incorporated.\n\nOur study did not find any changes in heart rate or heart rate variability measures during the channeling state compared to a control condition. Two studies have found increased heart rate from before to after a trance channeling35,80 and one found in heart rate variability compared to controls35. Similarly, respiration and galvanic skin response were the same across conditions. Autonomic nervous system function did not seem to be altered in our study although it was in others and additional study is warranted.\n\nAs far as we know, this is the first study to objectively evaluate voice measures in trance channeling. Voice valence was significantly lower when read during channeling compared to no-channeling, arousal trended towards being significantly lower and there was no difference in temper. Valence measures the speaker’s level of negativity at the lower end of the scale to positive attitude at the higher end. Arousal is an output that measures a speaker’s degree of energy ranging from tranquil, bored or sleepy to excited and highly energetic. To the subjective listener, the channeled readings were softer in volume and slower in pace, which likely resulted in lower valence and arousal measures. We also observed power differences at specific frequencies using long-time average spectrum voice analysis, one method that has been used to discriminate gender and individuals65. There is the possibility that participants impersonated a different voice during the channeled reading periods, however, the participants exhibited no signs of multiple personality disorder, dissociative identity disorder, or psychotic symptoms reflecting the pathology or motivation to do so.\n\nThere are number of limitations to this study that should be considered when interpreting its results. The study was designed as a within-participant controlled paradigm with multiple state shifts and test-retest data collections for multiple reasons. The repeated measures on Day 1 and Day 2 were done to evaluate test-retest reliability of the measures across days. The repeated sessions were done to evaluate repeatability within the same day and to increase data-points collected. However, the alternating between channeling and no-channeling states may have created a carryover effect, although we gave ample time for the transition states, and the mind-wandering task was perceived as subjectively different than the channeling. Because the trials were alternating often and lasted only five minutes, it is conceivable that the participants were not completely switching between states and that they were still partially in a trance state during some control trials. Or even if they were, it is possible that physiological systems take longer to go back to baseline (e.g. HRV), therefore decreasing the difference between the two conditions. Perhaps five minutes of data in each condition was not long enough to generate changes and longer segments need to be used. We purposefully kept the segment times at 5 minutes to reduce participant burden for the laboratory visit. Future studies could use longer segment periods but fewer conditions switches. Also, the channels baseline state may not be different from the channeling state just like many expert meditators baseline states are altered from their extensive meditation practice. Novice meditators often show more dramatic changes in neurophysiology when meditating versus when they are not meditating because these states are markedly different for novices. However, expert meditators may have a baseline non-meditating state that is not as different than their meditating state (i.e. their meditating and non-meditating states are more similar to each other compared to a novice’s)81–83. Perhaps future studies could include participants who have just begun channeling and those who have years of experience to explore differences in their physiology.\n\nIt is possible that the small sample size did not allow the analyses to detect changes between the two conditions. While the repeated measures design attempted to mitigate the smaller participant number, future studies could include more participants.\n\nIn order to increase the EEG data quality, all frequencies above 40 Hz were filtered out. We can only conclude that there were no differences in the channeling versus no-channeling conditions in the frequency bands that we measure. Future secondary analyses of this data set will include higher frequency bands with sophisticated ICA methods to remove electromyographic sources from the signal that are associated with higher frequencies.\n\nFurthermore, some participants reported different purported beings incorporating during each session which could trigger different neurophysiological signatures. It is possible that no homogenous states were present across all channeling trials, and could consequently appear as noise. However, one would anticipate that this would result in increased variation between the channeling sessions which was not observed. Secondary analyses will include additional within subject analysis for participants who reported different beings incorporating during the sessions. Future studies may also consider instructing participants to channel the same purported “being” for each trial.\n\nAdditionally, the channels reported that they normally speak when channeling as the purpose is to deliver a message, and that holding the incorporation of the purported being without speaking was unusual for them. This was done to ensure a clean EEG signal of high quality but perhaps future technological advances will allow for clean EEG collection while the channel is speaking which may more clearly delineate EEG frequency differences or less optimally select out non-speaking segments as others have done25. As is common with all studies that attempt to replicate real-world experiences in the laboratory, there is the consideration that the laboratory environment was unique to the channels. Thus, they may not have been as comfortable as they normally are when they channel, which could have interfered with their performance although the participants did not comment on this.\n\nPrevious qualitative reports of trance channeled content suggest a concept that may provide context why no physiological differences were seen across conditions10. One concept is that trance channels are channeling an aspect of themselves or higher self, that is, that the incorporated “beings” are actually multi-dimensional aspects of the channel themselves. These multi-dimensional aspects then need to match their energetic vibrational frequency of the channel to be able to use their body for communication. No measures are available at this time to confirm or deny this theory but perhaps measures will be developed to assess this in the future. It is also possible that differences do exist between channeling and no-channeling states but that the measures we used were not able to adequately capture them. For example, we did not do an EEG coherence or synchrony evaluation that may reflect changes in connectivity. Secondary analyses for these measures are planned.\n\nFurther analysis will examine the specific purported beings channeled for each segment compared to its control, transition periods between channeling and no-channeling states, EEG connectivity, magnetometer data, and qualitative evaluation of the channeled content. A composite measure of autonomic nervous system values and EEG values may also be explored. Based on our results on voice valence and power spectrum, future analyses could include pitch and intensity in order to study this effect in more detail.\n\nThis study has moved channeling research forward by evaluating physiological measure differences between the channeling and no-channeling states using rigorous controlled methods. Despite subjective perceptions of distinctly different states, no substantive differences were seen in EEG frequency power, ECG measures, and respiration. However, the voice measure of valence and spectral power in specific frequencies were lower in the channeling state while reading a story.\n\n\nData availability\n\nUnderlying data including voice analysis data, survey data, lab temperature and humidity, raw physiology data files, and transcripts of open-ended questions are available from Figshare\n\nFigshare: Dataset 1. Wahbeh_Data for Trance Channel Study. https://doi.org/10.6084/m9.figshare.7355027.v268\n\nThe online survey, stories read by the participants, transcripts and scripts for process the data are available as Extended data from figshare.\n\nFigshare: Extended data 1. Wahbeh_Additional Data for Trance Channel Study. https://doi.org/10.6084/m9.figshare.7423454.v139\n\nFigshare: Extended data 2. Wahbeh_Scripts from Trance Channel study. https://doi.org/10.6084/m9.figshare.7469972.v167\n\nAll underlying and extended data are available under a CC0 1.0 Universal Public Domain Dedication.",
"appendix": "Grant information\n\nThis work was supported by FUNDAÇÃO BIAL (grant number No. 72/16). This work was also supported by the Ray Benton Foundation and the Federico and Elvia Faggin Foundation.\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 would like to thank the Institute of Noetic Sciences, the participants of the study, the Ray Benton Foundation, the Federico and Elvia Faggin Foundation and FUNDAÇÃO BIAL for their support of this and other projects in this research program.\n\n\nReferences\n\nKlimo J: Channeling: Investigations on receiving information from paranormal sources. Berkeley, CA: North Atlantic Books; 1998. Reference Source\n\nCrook JH: The indigenous psychiatry of Ladakh, Part I: Practice theory approaches to trance possession in the himalayas. Anthropol Med. 1997; 4(3): 289–307. Publisher Full Text\n\nHageman JH, Peres JFP, Moreira-Almeida A, et al.: The Neurobiology of Trance and Mediumship in Brazil. In: Krippner SC, Friedman HL, eds. Mysterious Minds: The Neurobiology of Psychics, Mediums, and Other Extraordinary People. USA: Praeger; 2009. Reference Source\n\nKua EH, Sim LP, Chee KT: A Cross-Cultural Study of the Possession-Trance in Singapore. Aust N Z J Psychiatry. 1986; 20(3): 361–4. PubMed Abstract | Publisher Full Text\n\nHaraldsson E, Houtkooper JM: Psychic experiences in the multinational human values study: Who reports them. J Am Soc Psych Res. 1991; 85(2): 145–165. Reference Source\n\nWahbeh H, Radin D, Mossbridge J, et al.: Exceptional experiences reported by scientists and engineers. Explore (NY). 2018; 14(5): 329–341. PubMed Abstract | Publisher Full Text\n\nNegro PJ Jr, Palladino-Negro P, Louzã MR: Do religious mediumship dissociative experiences conform to the sociocognitive theory of dissociation? J Trauma Dissociation. 2002; 3(1): 51–73. Publisher Full Text\n\nRoxburgh EC, Roe CA: A survey of dissociation, boundary-thinness, and psychological wellbeing in spiritualist mental mediumship. J Parapsychol. 2011; 75(2): 279. Reference Source\n\nKlimo J: Channeling research: A progress report with recommendations. Paper presented at: International Conference on Paranormal Research; Fort Collins, CO. 1988.\n\nWahbeh H, Carpenter L, Radin D: A mixed methods phenomenological and exploratory study of channeling. Journal of the Society for Psychical Research. 2018; 82(3): 129–148. Reference Source\n\nAmerican Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders. (5th ed.). Washington D.C.: American Psychiatric Publishing Inc. 2013. Reference Source\n\nCoons PM: Confirmation of childhood abuse in child and adolescent cases of multiple personality disorder and dissociative disorder not otherwise specified. J Nerv Ment Dis. 1994; 182(8): 461–464. PubMed Abstract | Publisher Full Text\n\nOgawa JR, Sroufe LA, Weinfield NS, et al.: Development and the fragmented self: longitudinal study of dissociative symptomatology in a nonclinical sample. Dev Psychopathol. 1997; 9(4): 855–879. PubMed Abstract | Publisher Full Text\n\nStolovy T, Lev-Wiesel R, Witztum E: Dissociation: adjustment or distress? Dissociative phenomena, absorption and quality of life among Israeli women who practice channeling compared to women with similar traumatic history. J Relig Health. 2015; 54(3): 1040–1051. PubMed Abstract | Publisher Full Text\n\nRoss CA, Joshi S, Currie R: Dissociative experiences in the general population. Am J Psychiatry. 1990; 147(11): 1547–1552. PubMed Abstract | Publisher Full Text\n\nSeligman R: Distress, dissociation, and embodied experience: Reconsidering the pathways to mediumship and mental health. Ethos. 2005; 33(1): 71–99. Publisher Full Text\n\nSar V, Alioğlu F, Akyuz G: Experiences of possession and paranormal phenomena among women in the general population: are they related to traumatic stress and dissociation? J Trauma Dissociation. 2014; 15(3): 303–318. PubMed Abstract | Publisher Full Text\n\nSchaffler Y, Cardeña E, Reijman S, et al.: Traumatic Experience and Somatoform Dissociation Among Spirit Possession Practitioners in the Dominican Republic. Cult Med Psychiatry. 2016; 40(1): 74–99. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKihlstrom JF: Dissociative disorders. Annu Rev Clin Psychol. 2005; 1: 227–253. PubMed Abstract | Publisher Full Text\n\nSeligman R, Kirmayer LJ: Dissociative experience and cultural neuroscience: narrative, metaphor and mechanism. Cult Med Psychiatry. 2008; 32(1): 31–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpitzer C, Barnow S, Freyberger HJ, et al.: Recent developments in the theory of dissociation. World Psychiatry. 2006; 5(2): 82–86. PubMed Abstract | Free Full Text\n\nCastillo RJ: Trance, functional psychosis, and culture. Psychiatry. 2003; 66(1): 9–21. PubMed Abstract | Publisher Full Text\n\nWahbeh H, Radin D: People reporting experiences of mediumship have higher dissociation symptom scores than non-mediums, but below thresholds for pathological dissociation [version 3; referees: 2 approved, 1 not approved]. F1000Res. 2017; 6: 1416. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWahbeh H, McDermott K, Sagher A: Dissociative Symptoms and Anomalous Information Reception. Act Nerv Super (Praha). 2018; 60(3–4): 75–85. Publisher Full Text\n\nBastos MA, Bastos PR, Osório IH, et al.: Frontal electroencephalographic (EEG) activity and mediumship: a comparative study between spiritist mediums and controls. Arch Clin Psychiat. 2016; 43(2). Publisher Full Text\n\nBastos MA, Bastos PR, Goncalves LM, et al.: Mediumship: review of quantitatives studies published in the 21st century. Arch Clin Psychiat. 2015; 42(5): 129–138. Publisher Full Text\n\nBeischel J, Boccuzzi M, Biuso M, et al.: Anomalous information reception by research mediums under blinded conditions II : replication and extension. Explore (NY). 2015; 11(2): 136–142. PubMed Abstract | Publisher Full Text\n\nRock AJ, Beischel J, Boccuzzi M, et al.: Discarnate readings by claimant mediums: Assessing phenomenology and accuracy under beyond double-blind conditions. J Parapsychol. 2014; 78(2): 183–194. Reference Source\n\nBeischel J, Schwartz GE: Anomalous information reception by research mediums demonstrated using a novel triple-blind protocol. Explore (NY). 2007; 3(1): 23–27. PubMed Abstract | Publisher Full Text\n\nDelorme A, Beischel J, Michel L, et al.: Electrocortical activity associated with subjective communication with the deceased. Front Psychol. 2013; 4: 834. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDelorme A, Pierce A, Michel L, et al.: Intuitive Assessment of Mortality Based on Facial Characteristics: Behavioral, Electrocortical, and Machine Learning Analyses. Explore (NY). 2018; 14(4): 262–267. PubMed Abstract | Publisher Full Text\n\nBeischel J, Rock AJ: Addressing the survival versus psi debate through process-focused mediumship research. J Parapsychol. 2009; 73: 71. Reference Source\n\nHughes DJ, Melville NT: Changes in brainwave activity during trance channeling: a pilot study. J Transpersonal Psy. 1990; 22(2): 175–189. Reference Source\n\nKawai N, Honda M, Nishina E, et al.: Electroencephalogram characteristics during possession trances in healthy individuals. Neuroreport. 2017; 28(15): 949–955. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBastos MAV Jr, Bastos PRHO, Osório IHS, et al.: Physiologic correlates of culture-bound dissociation: A comparative study of Brazilian spiritist mediums and controls. Transcult Psychiatry. 2018; 55(2): 286–313. PubMed Abstract | Publisher Full Text\n\nCardeña E, Palmer J, Marcusson-Clavertz D: Parapsychology: A handbook for the 21st century. McFarland; 2015. Reference Source\n\nNelson RD: Implicit Physical Psi: The Global Consciousness Project. In: Cardena E, Palmer J, Marcusson-Clavertz D, eds. Parapsychology: A Handbook for the 21st Centry. Jefferson, North Carolina: McFarland & Company, In., Publishers; 2015; 1. : 282–292. Reference Source\n\nVarvoglis M, Bancel PA: Micro-psychokinesis. In: Cardena E, Palmer J, Marcusson-Clavertz D, eds. Parapsychology: A Handbook for the 21st Century. Jefferson, North Carolina: McFarland & Company, Inc., Publishers; 2015; 1. : 266–281. Reference Source\n\nWahbeh H, cannard C, Delorme A, et al.: Wahbeh_Additional Data for Trance Channel Study. 2018. http://www.doi.org/10.6084/m9.figshare.7423454.v1\n\nCapra C, Kavanagh DJ, Hides L, et al.: Current CAPE-15: a measure of recent psychotic-like experiences and associated distress. Early Interv Psychiatry. 2015; 11(5): 411–417. PubMed Abstract | Publisher Full Text\n\nChu JA: Rebuilding shattered lives: treating complex PTSD and dissociative disorders. 2nd ed. Hoboken, N.J.: John Wiley & Sons; 2011. Publisher Full Text\n\nBarratt W: The Barratt Simplified Measure of Social Status (BSMSS) measuring SES. 2006; Reference Source\n\nCirino PT, Chin CE, Sevcik RA, et al.: Measuring socioeconomic status: reliability and preliminary validity for different approaches. Assessment. 2002; 9(2): 145–155. PubMed Abstract | Publisher Full Text\n\nWaller N, Putnam FW, Carlson EB: Types of dissociation and dissociative types: A taxometric analysis of dissociative experiences. Psychol Methods. 1996; 1(3): 300–321. Publisher Full Text\n\nWaller NG, Ross CA: The prevalence and biometric structure of pathological dissociation in the general population: taxometric and behavior genetic findings. J Abnorm Psychol. 1997; 106(4): 499–510. PubMed Abstract | Publisher Full Text\n\nZingrone NL, Alvarado CS: The Dissociative Experiences Scale-II: Descriptive Statistics, Factor Analysis, and Frequency of Experiences. Imagin Cogn Pers. 2001; 21(2): 145–157. Publisher Full Text\n\nCapra C, Kavanagh DJ, Hides L, et al.: Brief screening for psychosis-like experiences. Schizophr Res. 2013; 149(1-3): 104-107. PubMed Abstract | Publisher Full Text\n\nKroenke K, Spitzer RL, Williams JB, et al.: An ultra-brief screening scale for anxiety and depression: the PHQ-4. Psychosomatics. 2009; 50(6): 613-621. PubMed Abstract | Publisher Full Text\n\nRammstedt B, John OP: Measuring personality in one minute or less: A 10-item short version of the Big Five Inventory in English and German. J Res Pers. 2007; 41(1): 203-212. Publisher Full Text\n\nTellegen A: A brief manual for the differential personality questionnaire. Minneapolis: Department of Psychology, University of Minnesota; 1982. Reference Source\n\nTellegen A, Atkinson G: Openness to absorbing and self-altering experiences (\"absorption\"), a trait related to hypnotic susceptibility. J Abnorm Psychol. 1974; 83(3): 268-277. PubMed Abstract | Publisher Full Text\n\nKihlstrom JF, Register PA, Hoyt IP, et al.: Dispositional correlates of hypnosis: a phenomenological approach. Int J Clin Exp Hypn. 1989; 37(3): 249–263. PubMed Abstract | Publisher Full Text\n\nLoewen PJ, Lyle G, Nachshen JS: An eight-item form of the Empathy Quotient (EQ) and an application to charitable giving. 2009. Reference Source\n\nWakabayashi A, Baron-Cohen S, Wheelwright S, et al.: Development of short forms of the Empathy Quotient (EQ-Short) and the Systemizing Quotient (SQ-Short). Pers Individ Dif. 2006; 41(5): 929–940. Publisher Full Text\n\nAron EN, Aron A: Sensory-processing sensitivity and its relation to introversion and emotionality. J Pers Soc Psychol. 1997; 73(2): 345–368. PubMed Abstract | Publisher Full Text\n\nLange R, Thalbourne MA: Rasch scaling paranormal belief and experience: structure and semantics of Thalbourne's Australian Sheep-Goat Scale. Psychol Rep. 2002; 91(3 Pt 2): 1065–1073. PubMed Abstract | Publisher Full Text\n\nDelorme A, Makeig S: EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Methods. 2004; 134(1): 9–21. PubMed Abstract | Publisher Full Text\n\nWahbeh H, Oken BS: Peak high-frequency HRV and peak alpha frequency higher in PTSD. Appl Psychophysiol Biofeedback. 2013; 38(1): 57–69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoodnick PJ, Jerry J, Parra F: Psychotropic drugs and the ECG: focus on the QTc interval. Expert Opin Pharmacother. 2002; 3(5): 479–498. PubMed Abstract | Publisher Full Text\n\nOlgin J, Zipes D: Specific arrhythmias: diagnosis and treatment. In: Libby P, Bonow R, Mann D, Zipes D, eds. Braunwald's Heart Disease: A Textbook of Cardiovascular Medicine. St. Louis, Mo: WB Saunders; 2007.\n\nSchmidt M, Schumann A, Müller J, et al.: ECG derived respiration: comparison of time-domain approaches and application to altered breathing patterns of patients with schizophrenia. Physiol Meas. 2017; 38(4): 601–615. PubMed Abstract | Publisher Full Text\n\nSoleymani M, Chanel G, Kierkels JJ, et al.: Affective characterization of movie scenes based on content analysis and physiological changes. Int J Semant Comput. 2009; 3(02): 235–254. Publisher Full Text\n\nKoelstra S, Yazdani A, Soleymani M, et al.: Single trial classification of EEG and peripheral physiological signals for recognition of emotions induced by music videos. Paper presented at: International Conference on Brain Informatics. 2010. Publisher Full Text\n\nKhalili Z, Moradi MH: Emotion recognition system using brain and peripheral signals: using correlation dimension to improve the results of EEG. Paper presented at: Neural Networks, IJCNN 2009. International Joint Conference on. 2009. Publisher Full Text\n\nMendoza E, Valencia N, Muñoz J, et al.: Differences in voice quality between men and women: use of the long-term average spectrum (LTAS). J Voice. 1996; 10(1): 59–66. PubMed Abstract | Publisher Full Text\n\nBenjamini Y, Hochberg Y: Controlling the false discovery rate: A practical and powerful approach to multiple testing. J Roy Stat Soc B Met. 1995; 57(1): 289–300. Publisher Full Text\n\nWahbeh H, Delorme A, cannard C, et al.: Wahbeh_Scripts from Trance Channel study. 2018. http://www.doi.org/10.6084/m9.figshare.7469972.v1\n\nWahbeh H, Delorme A, okonsky J, et al.: Wahbeh_Data for Trance Channel Study. 2018. http://www.doi.org/10.6084/m9.figshare.7355027.v2\n\nBukenaite A, Stochl J, Mossaheb N, et al.: Usefulness of the CAPE-P15 for detecting people at ultra-high risk for psychosis: Psychometric properties and cut-off values. Schizophr Res. 2017; 189: 69–74. PubMed Abstract | Publisher Full Text\n\nRammstedt B: The 10-item Big Five Inventory: Norm values and investigation of sociodemographic effects based on a German population representative sample. Eur J Psychol Assess. 2007; 23(3): 193. Publisher Full Text\n\nSrivastava S, John OP, Gosling SD, et al.: Development of personality in early and middle adulthood: set like plaster or persistent change? J Pers Soc Psychol. 2003; 84(5): 1041–53. PubMed Abstract | Publisher Full Text\n\nGlisky ML, Tataryn DJ, Tobias BA, et al.: Absorption, openness to experience, and hypnotizability. J Pers Soc Psychol. 1991; 60(2): 263–72. PubMed Abstract | Publisher Full Text\n\nHageman JH, Krippner S, Wickramasekera II I: Across cultural boundaries: psychophysiological responses, absorption, and dissociation comparison between Brazilian spiritists and advanced meditators. Neuroquantology. 2011; 9(1). Publisher Full Text\n\nAron EN, Aron A, Jagiellowicz J: Sensory processing sensitivity: a review in the light of the evolution of biological responsivity. Pers Soc Psychol Rev. 2012; 16(3): 262–282. PubMed Abstract | Publisher Full Text\n\nWahbeh H, Yount G, Delorme A, et al.: The paranormal belief and experience scale: validation and reliability. In preparation. 2018.\n\nSeligman R: From affliction to affirmation: narrative transformation and the therapeutics of Candomblé mediumship. Transcult Psychiatry. 2005; 42(2): 272–294. PubMed Abstract | Publisher Full Text\n\nKennedy J, Kanthamani H, Palmer J: Psychic and spiritual experiences, health, well-being, and meaning in life. J Parapsychol. 1994; 58(4): 353–383. Reference Source\n\nOohashi T, Kawai N, Honda M, et al.: Electroencephalographic measurement of possession trance in the field. Clin Neurophysiol. 2002; 113(3): 435–445. PubMed Abstract | Publisher Full Text\n\nBrown GD, Yamada S, Sejnowski TJ: Independent component analysis at the neural cocktail party. Trends Neurosci. 2001; 24(1): 54–63. PubMed Abstract | Publisher Full Text\n\nKawai N, Honda M, Nakamura S, et al.: Catecholamines and opioid peptides increase in plasma in humans during possession trances. Neuroreport. 2001; 12(16): 3419–3423. PubMed Abstract | Publisher Full Text\n\nCahn BR, Polich J: Meditation states and traits: EEG, ERP, and neuroimaging studies. Psychol Bull. 2006; 132(2): 180–211. PubMed Abstract | Publisher Full Text\n\nFox KC, Nijeboer S, Dixon ML, et al.: Is meditation associated with altered brain structure? A systematic review and meta-analysis of morphometric neuroimaging in meditation practitioners. Neurosci Biobehav Rev. 2014; 43: 48–73. PubMed Abstract | Publisher Full Text\n\nFox KC, Dixon ML, Nijeboer S, et al.: Functional neuroanatomy of meditation: A review and meta-analysis of 78 functional neuroimaging investigations. Neurosci Biobehav Rev. 2016; 65: 208–228. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "43131",
"date": "24 Jan 2019",
"name": "Patrizio Tressoldi",
"expertise": [
"Reviewer Expertise Science of consciousness",
"Methodological and statistical procedures"
],
"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\nAs the authors said \"This study has moved channeling research forwards by evaluating physiological measure differences between the channeling and no-channeling states using rigorous controlled methods\". A further strength of this study is the test of the reliability of their physiological measures.\nMain comments:\nBoth in the introduction and the discussion of their study, it is important to make explicit the exploratory approach for the choice of the variables and their parameters, unless pre-defined in the grant protocol of the funding institution. In this case, the authors should make this document available. The procedure to read stories both in non-channeling and channeling condition is unusual and should comment a little bit further in particular considering how to convince the purported entity to read a story chosen by the experimenters. The data presented in Table 3, shows a clear trend from session 1 to 3 that, in my view, is important to take in consideration. As for all other variables, add the descriptive statistics of the ECG and respiration data. The difference in the GSR variable must be considered given the non trivial difference and not ignored simply because the p value exceed the conventional cutoff of .05. In fact, as acknowledged by the authors, the statistical power with only 13 participants is not sufficient to detect medium to small effect size differences. For the EEG data, I wonder whether it could be more appropriate to cluster the 64 channels in order to investigate specific regions of interest, enhancing the statistical power of the statistics.\nMinor comments:\nChannel or channeler? I wonder whether to use channeler instead of channel could be more precise in conveying the difference between the agent and his/her action, e.g. in the sentence \"Well-known trance channels primarily channel one being\"\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": "4400",
"date": "05 Feb 2019",
"name": "Helané Wahbeh",
"role": "Author Response",
"response": "Thank you for your thoughtful comments about our manuscript. We have incorporated your suggestions into the manuscript as outlined in italics under each point.Main comments: Both in the introduction and the discussion of their study, it is important to make explicit the exploratory approach for the choice of the variables and their parameters, unless pre-defined in the grant protocol of the funding institution. In this case, the authors should make this document available. - The analysis of the study were pre-defined in the grant protocol. The original grant protocol has been added to the extended data stores. In the introduction, we have added a sentence regarding this and it is also already included in the Methods in the Study Overview section. The procedure to read stories both in non-channeling and channeling condition is unusual and should comment a little bit further in particular considering how to convince the purported entity to read a story chosen by the experimenters. - We have added more information about this in the Methods section. The data presented in Table 3, shows a clear trend from session 1 to 3 that, in my view, is important to take in consideration. - Yes we agree. We have added a few sentences about this in the discussion session. We had only asked about these two scales during the channeling condition. Because of this we can’t look at these values as a covariate to see if any of the physiology are different between channeling and no-channeling were influenced by these factors. I did a brief exploratory analysis looking at only the channeling alpha EEG collapsing all channels and there the unconsciousness and incorporation variables were not significant by themselves or as an interaction with session number. It would be important to consider this for future studies and if possible, incorporate it in way so that they could be included in the statistical model. As for all other variables, add the descriptive statistics of the ECG and respiration data. - Descriptive statistics for ECG and respiration data have been added to the text. The difference in the GSR variable must be considered given the non-trivial difference and not ignored simply because the p value exceed the conventional cutoff of .05. In fact, as acknowledged by the authors, the statistical power with only 13 participants is not sufficient to detect medium to small effect size differences. - We have added an exploratory paired t-test comparing the collapsed average values of GSR for channeling and no-channeling. The results were significant. We also added a few sentences about this in the discussion section. For the EEG data, I wonder whether it could be more appropriate to cluster the 64 channels in order to investigate specific regions of interest, enhancing the statistical power of the statistics. - We have taken this advice and clustered the 64 channels by location and conducted exploratory analyses. The descriptive statistics and analyses results are now included in the paper. Minor comments: Channel or channeler? I wonder whether to use channeler instead of channel could be more precise in conveying the difference between the agent and his/her action, e.g. in the sentence \"Well-known trance channels primarily channel one being\" -Channeler has been included to refer to the agent throughout the manuscript."
}
]
},
{
"id": "43134",
"date": "28 Jan 2019",
"name": "Marco Aurélio Vinhosa Bastos Jr.",
"expertise": [
"Reviewer Expertise Psychoneuroendocrinology",
"Investigation of physiologic correlates of anomalous experiences"
],
"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 is a detailed account of an experiment investigating physiological correlates of channeling, a type of mediumistic experience. Specifically, the investigators selected only 'trance channels' as participants, i.e. when discarnate entities allegedly speak directly through the mouth of the 'channel' (medium), rather than those individuals with more mental and intuitive mediumistic experiences. In addition to several psychometric measurements, the method included evaluation of many physiological parameters (EEG, ECG, respiration, 'electrodermal activity', voice analysis), during two days of experiments in the Laboratory, adopting a 'within-subject' design – comparison of the findings in the 'channeling' and 'no-channeling' conditions. Basically, no differences were noted in subjects' physiological parameters between the channeling and no-channeling conditions, except for the voice analysis, which showed lower valence and lower arousal in the channeling condition. The experiment is interesting and seems to have been well executed (e.g. appropriate counterbalanced design of tasks to avoid ‘carry over effects’, clear exposition of recruitment strategy and response-rates, sound statistical analysis), and the paper is well written.\nHowever, I have some specific comments, as follows:\nMajor points:\nIn order to avoid artifacts during EEG recording of the experiment, only mediums that assured being able to stand still during the channeling were selected. Additionally, being able to control when to start and when to terminate the channeling experience and having a ‘consensual working relationship with discarnate beings’ were also inclusion criteria. The authors discuss their results, in view of the few studies on the topic available in the literature. The findings (basically ‘no-differences’) are somewhat incongruent to most of the previous studies (pointed out in the papers' References), that have suggested greater arousal in subjects during the mediumistic experience, both in 'within-subject' and in between-group design (comparing with non-medium subjects from the same cultural contexts) experiments. The authors consider that the explanation for this incongruence to previous studies may be the more rigorous statistical methods they adopted, although recognizing that it may also stem from the un-naturalistic design of the study. In the present study, the channels needed to enter into trance but the communicating entities could not speak, except for the periods when the investigators asked them to read standard texts, so that the medium` voice could be recorded for analysis. This point is approached mainly in the Limitations section (pages 15, last paragraph), but some words about it should be included when discussing ECG data as well. The finding of differences in voice analysis, which is interesting and deserves to be confirmed in other studies, could maybe be tested with a more naturalistic (and perhaps more appropriate) approach, allowing the entity's spontaneous speech through the medium, and then performing transcription of the speech. This could then be read by the same medium at another moment, in the 'no-channeling' condition, comparing voice parameters. It can be anticipated that results of voice analysis might be very different if spontaneous mediumistic speech is allowed. Differently from what the authors state in the Discussion section (page 14, first paragraph), Hughes et al. (1990) have not evaluated theta, alpha and beta POWER pre-, during and post-trance in channels. Instead, they have evaluated “the percentage of time that theta (or alpha, or beta) brainwave activity was registered by the EEG equipment”1. This may have a different physiological meaning.\n\nI agree with Professor Tressoldi, it is important to include the results of ECG data in the article (mean ± standard deviation, statistics). The same for respiration data.\n\nMinor points:\nI wonder whether Figure 4 could be reframed so that it becomes clearer for the non-specialist reader. Last paragraph of the Introduction section: ‘distinguishing these two states as distinct’. This sentence seems overly redundant. Page 6, third paragraph: ‘confounds in many paranormal belief scales of belief’. This is probably a spelling error. Page 14, penultimate line of first paragraph: ‘considering may of our participants stated that it felt strange’. It should read ‘many of our participants stated’.\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": "4401",
"date": "05 Feb 2019",
"name": "Helané Wahbeh",
"role": "Author Response",
"response": "Thank you for your thoughtful comments about our manuscript. We have incorporated your suggestions into the manuscript as outlined in italics under each point. In order to avoid artifacts during EEG recording of the experiment, only mediums that assured being able to stand still during the channeling were selected. Additionally, being able to control when to start and when to terminate the channeling experience and having a ‘consensual working relationship with discarnate beings’ were also inclusion criteria. The authors discuss their results, in view of the few studies on the topic available in the literature. The findings (basically ‘no-differences’) are somewhat incongruent to most of the previous studies (pointed out in the papers' References), that have suggested greater arousal in subjects during the mediumistic experience, both in 'within-subject' and in between-group design (comparing with non-medium subjects from the same cultural contexts) experiments. The authors consider that the explanation for this incongruence to previous studies may be the more rigorous statistical methods they adopted, although recognizing that it may also stem from the un-naturalistic design of the study. In the present study, the channels needed to enter into trance but the communicating entities could not speak, except for the periods when the investigators asked them to read standard texts, so that the medium` voice could be recorded for analysis. This point is approached mainly in the Limitations section (pages 15, last paragraph), but some words about it should be included when discussing ECG data as well. - We have added some language about this point in the ECG discussion section. The finding of differences in voice analysis, which is interesting and deserves to be confirmed in other studies, could maybe be tested with a more naturalistic (and perhaps more appropriate) approach, allowing the entity's spontaneous speech through the medium, and then performing transcription of the speech. This could then be read by the same medium at another moment, in the 'no-channeling' condition, comparing voice parameters. It can be anticipated that results of voice analysis might be very different if spontaneous mediumistic speech is allowed. - Thank you, we have added text about this into our discussion section. Differently from what the authors state in the Discussion section (page 14, first paragraph), Hughes et al. (1990) have not evaluated theta, alpha and beta POWER pre-, during and post-trance in channels. Instead, they have evaluated “the percentage of time that theta (or alpha, or beta) brainwave activity was registered by the EEG equipment”1. This may have a different physiological meaning. - We have corrected this. I agree with Professor Tressoldi, it is important to include the results of ECG data in the article (mean ± standard deviation, statistics). The same for respiration data. - ECG and respiration descriptive statistics have been included. Minor points: I wonder whether Figure 4 could be reframed so that it becomes clearer for the non-specialist reader. - We have added a legend for Figure 4 that helps clarify what it means. Last paragraph of the Introduction section: ‘distinguishing these two states as distinct’. This sentence seems overly redundant. - This phrase has been removed. Page 6, third paragraph: ‘confounds in many paranormal belief scales of belief’. This is probably a spelling error. - Extra belief has been removed Page 14, penultimate line of first paragraph: ‘considering may of our participants stated that it felt strange’. It should read ‘many of our participants stated’. - This has been fixed."
}
]
}
] | 1
|
https://f1000research.com/articles/8-67
|
https://f1000research.com/articles/7-1937/v1
|
16 Dec 18
|
{
"type": "Research Article",
"title": "Nrf2-Mediated Antioxidant Activity of the inner bark extracts obtained from Tabebuia rosea (Bertol) DC and Tabebuia chrysantha (JACQ) G. Nicholson.",
"authors": [
"Sandra C. Garzón-Castaño",
"Iván A. Lopera-Castrillón",
"Francisco J. Jiménez-González",
"Fernando Siller-López",
"Luz A. Veloza",
"Juan Carlos Sepúlveda-Arias",
"Sandra C. Garzón-Castaño",
"Iván A. Lopera-Castrillón",
"Francisco J. Jiménez-González",
"Fernando Siller-López",
"Luz A. Veloza"
],
"abstract": "Background: Several ethnobotanical and ethnopharmacological studies have shown the therapeutic potential of plants from the genus Tabebuia, which have long been used in traditional medicine in rural areas of South America, for the treatment of several human diseases. This study aimed to evaluate the Nrf2-mediated antioxidant activity of the inner bark extracts obtained from Tabebuia rosea and Tabebuia chrysantha. Methods: The antioxidant activity of extracts obtained from the inner bark of T. rosea and T. chrysantha was evaluated using the Oxygen radical absorbance capacity (ORAC) technique. The effect of extracts on the viability of HepG2 cells was determined using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) method. The translocation of Nrf2 to the nucleus after exposure of HepG2 cells to the extracts and controls (α-lipoic acid, curcumin and hydrogen peroxide) was evaluated using the Nrf2 transcription factor kit. Induction of the Nrf2-mediated antioxidant response gene (NQO1) was evaluated by real-time PCR. Results: The ethyl acetate extract obtained from both species displayed the highest ORAC activity (12,523 and 6,325 µmoles Eq Trolox/g extract, respectively). In addition, the extracts had the ability to activate and to translocate Nrf2 to the nucleus, as well as to induce the expression of NQO1. Conclusion: These results indicate that the ethyl acetate extracts obtained from the inner bark of T. chrysantha and T. rosea have an important antioxidant effect mediated by Nrf2 activation, and could be used as a new source of natural antioxidants.",
"keywords": [
"Tabebuia chrysantha",
"Tabebuia rosea",
"Bignoniaceae",
"extracts",
"Nrf2",
"antioxidant agents."
],
"content": "\n\nAbbreviations used: AAPH: 2,2’-Azobis (2-amidinopropane) dihydrochloride; ALA: Alpha-lipoic acid; CUR: Curcumin; ORAC: Oxygen radical absorbance capacity; MTT: 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide.\n\n\nIntroduction\n\nNature’s compounds reveal a great diversity of chemical structures and physicochemical properties, as well as biological ones. Over the years, plants have been used for the treatment of various diseases, including those of inflammatory origin, such as arthritis, obesity, and cancer. Plants of the genus Tabebuia belong to the Bignoniaceae family, which is composed of about 120 genera with 827 species, and is considered the second most diverse family of species of neotropical woody plants in dry forests1. There are reports on the presence of chemical compounds⎯including quinones and phenols, among others⎯in this family2,3. The Tabebuia genus comprises about 100 species of trees and shrubs, mainly distributed from Mexico to several regions of Central and South America, which have been used in traditional medicine. Plants of this genus are an important source of bioactive molecules such as: naphthoquinones; quinones; phenols; and molecules with anti-inflammatory, antioxidant, anti-microbial and anti-proliferative activity4–8.\n\nA large number of chemical compounds exert their antioxidant effects through the activation of key transcriptional regulation mechanisms, such as the transcription factor Nrf2 (nuclear factor erythroid 2-related factor 2)9. Under normal physiological conditions, this factor is in the cytoplasm inhibited by Keap1 (Kelch ECH associating protein 1), which leads to its degradation10. In cells exposed to oxidative stress, Nrf2 is not targeted for ubiquitin-dependent degradation. Instead, it is released and translocated to the nucleus, where it activates its antioxidant response through binding to Antioxidant Response Elements (ARE sites), allowing the coordinated expression of more than 200 detoxifying enzymes and antioxidants, such as NAD(P)H quinone oxidoreductase (NQO1) and heme oxygenase 1 (HO-1), among others11–14. The aim of this study was to evaluate the Nrf2-mediated antioxidant activity of extracts obtained from the inner bark of Tabebuia rosea and Tabebuia chrysantha in HepG2 cells.\n\n\nMethods\n\nThe inner bark from T. rosea (Bertol.) DC and T. chrysantha (JACQ) G. Nicholson were collected at Universidad Tecnológica de Pereira campus in April 2011. The plants were identified at the Colombian National Herbarium (Voucher No. COL 582577 for T. rosea and COL 587611 for T. chrysantha). The collection and processing of the material were covered by the collection permission number 1133/2014, issued by the National Environmental Licensing Authority (ANLA) of Colombia.\n\nExtracts were obtained as previously described6. Plant material (inner bark from T. chrysantha and T. rosea) was dried and macerated in methanol analytical grade (14 L)for 48 h. This was followed by homogenization, filtration, and concentration under vacuum using a vacuum rotary evaporator (Heidolph, Laborota model) at 40 °C to obtain the crude extract. This procedure was repeated three times. Crude extracts were dissolved in 400 mL of distilled water, and underwent liquid–liquid extraction with increasing polarity solvents (solvent volume 1.6 L): n-hexane, chloroform (CHCl3), ethyl acetate (EtOAc), and n-butanol (all solvents were analytical grade). Each extract was vacuum dried by vacuum rotary evaporator obtaining the following mass for each extract: n-hexane (0.3 g), chloroform (1.2 g), ethyl acetate (3.7 g), n-butanol (12.5 g) and aqueous (21.7 g). Endotoxin levels in the extracts were assayed using the Limulus Amebocyte Lysate Test, E-Toxate Kit (Sigma Chemical Co, Saint Louis, MO, USA, Cat No. ET0200-1KT). All samples were negative for the presence of endotoxins (detection limit 0.05–0.1 EU/mL). The extracts were kept refrigerated at 4 °C in amber tubes protected from light, heat, air and moisture. For each one of the biological assays, the extracts were dissolved in 0.1% DMSO (Dimethyl sulfoxide, Merck, Darmstadt, Germany, Cat No. 1029521000).\n\nThe preliminary phytochemical screening was performed using selective derivatization reactions for the characterization of secondary metabolites present in the n-hexane, chloroform, ethyl acetate and n-butanol extracts obtained from inner bark, as previously reported for T. rosea6. The extracts were evaluated using normal and reverse phase thin layer chromatography (TLC). Chromatographic plates were revealed with aluminum chloride (AlCl3, Sigma Chemical Co, Saint Louis, MO, USA, Cat No. 563919-25G) and ferric chloride (FeCl3, Sigma Chemical Co, Saint Louis, MO, USA, Cat No. 157740-1KG ) for detection of flavonoids, phenols and phenolic acids; potassium hydroxide (KOH, Merck, Darmstadt, Germany, Cat No. 1050331000) in analytical grade ethanol for detection of anthrones, quinones and coumarins; oleum (Sigma Chemical Co, Saint Louis, MO, USA, Cat No. 778990-500ML) for detection of sesquiterpenic lactones; and the Liebermann-Burchard reagent for detection of terpenes and steroids.\n\nThe total phenolic content of each extract was determined according to the Folin–Ciocalteu colorimetric method15, using gallic acid as standard. Briefly, Folin-Ciocalteu’s reactive (Merck, Darmstadt, Germany, Cat No. 1090010100) was diluted 10-fold with distilled water. 25 µL of the samples (1 mg/mL) were added to the Folin-Ciocalteu’s reactive. After the addition of Na2CO3 (20%), the reaction was maintained at room temperature (RT) in the dark for 30 min, and absorbance was measured at 760 nm in a Shimadzu UV-1700 spectrophotometer. Gallic acid (0.25–5 mg/mL) was used to generate a standard curve (y=0.101x+0.086; R2=0.996). Results are presented as mg gallic acid equivalents per g of extract (mg GAE/g extract). All experiments were performed in triplicate.\n\nOxygen radical absorbance capacity was determined using the method described by Ou et al, with some modifications16. 2,2’-Azobis (2-amidinopropane) dihydrochloride (AAPH) and sodium fluorescein stock solutions were prepared in a 75 mM, pH 7.0 phosphate buffer solution. 31 μL of each sample were diluted in 187 μL of fluorescein (120 nM) and incubated at 37 °C for 10 min. The reaction was started by the addition of 31 μL of AAPH (143 mM) to reach a final volume of 249 μL per well. Extracts were evaluated at 10, 15, and 20 μg/mL. A Trolox® standard curve was prepared (10, 20, 40, and 60 μM). Changes in fluorescence were measured with a Varian Cary Eclipse 1.1 fluorescence spectrophotometer at 2 min intervals for 120 min with emission and excitation wavelengths of 515 and 493 nm, respectively. The excitation slit was 5 nm, as was the emission slit17,18. The antioxidant capacity was calculated as the area under the curve (AUC)19 and expressed as μmol Trolox® equivalents per g of extract (μmol TE/g of extract).\n\nHuman hepatocarcinoma cell line (HepG2; ATCC; CRL-11997) was purchased from American Type Culture Collection (ATCC, Rockville, MD, USA) and cultured with Dulbecco’s Modified Eagle Medium (DMEM), supplemented with Glutamax (GIBCO/BRL, USA, Cat No. 10564-011) and 10% heat inactivated FBS (GIBCO, Cat No. 16140071), 200 μg/mL Penicillin, 200 μg/mL streptomycin, 400 μg/mL neomycin (GIBCO, Cat No. 15640-055), 5 μg/mL amphotericin, 0.05 mM 2-β-mercaptoethanol, and 1 mM sodium pyruvate (all from Sigma Chemical Co, Saint Louis, MO, USA, Cat No. A9528-50MG, M7522-100ML, S8636-100ML, respectively). Cells were maintained at 37 °C with 5% CO2 atmosphere.\n\nTo determine the effects of extracts on HepG2 cells, cell viability was tested using the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) method20. Cells (5 × 104 cell/well) were treated with a concentration of 2 μg/mL of the extracts (the n-butanol extract from T. chrysantha was used at 1 μg/mL) diluted in DMSO (final concentration 0.1%), and incubated for 24 hours. After treatment, the medium was discarded, 200 μL of DMEM medium containing 0.5 mg/mL MTT (Sigma Chemical Co, Saint Louis, MO, USA, Cat No. M2128-500MG) was then added to each well. The plates were incubated for 4 hours at 37°C in a 5% CO2 atmosphere. The medium was discarded and 100 μL of DMSO was then added to solubilize the formazan crystals. Absorbance was measured in an ELISA microplate reader at 492 nm (ELx800; BioTek Instruments Inc., USA). Viability percentage was calculated based on the non-treated control. Three independent assays were performed, each one in triplicate.\n\nThe HepG2 cell line (3 × 105 cell/well) was cultured in DMEM medium using a T25 flask. The medium was discarded, and the cells were exposed at two time points (0 and 4 hours) to: Alpha-lipoic acid (ALA, 50 μM, Sigma Chemical Co, Saint Louis, MO, USA, Cat No. T1395-1G), Curcumin (CUR, 15 μM, Sigma Chemical Co, Saint Louis, MO, USA, Cat No. C7727-500MG )21–24, ethyl acetate extracts from T. chrysantha (0.5 μg/mL), ethyl acetate extract from T. rosea (1 μg/mL) and H2O2 (0.6 mM). After exposure, cells were harvested and used for nuclear and cytosolic protein extraction simultaneously, following the specifications included in the Nuclear Extraction Kit (Cayman Chemical, Ann Arbor MI, USA, Item No 10009277). Protein fractions were quantified using the BCA Protein Quantification Kit (Abcam, Cambridge, UK, ab102536). Nrf2 was detected by using the Nrf2 transcription factor assay kit (Cayman Chemical, Ann Arbor MI, USA, Item No 600590), and following the manufacturer’s instructions. The absorbance of each well was measured at 450 nm in an ELx800 BioTek microplate reader.\n\nHepG2 cells (3 × 105 cells/well) were treated with ALA (50 μM), CUR (15 μM, ethyl acetate extract from T. chrysantha (0.5 μg/mL), ethyl acetate extract from T. rosea (1 μg/mL) and H2O2 (0.6 mM) for durations of 0, 6 and 12 hours. After treatment, mRNA extraction was performed using the RNeasy Plus Mini Kit (Qiagen, Maryland, USA, Cat No. 74134). The mRNA was quantified with a NanoDrop 2000c (Thermo Fisher Scientific, Waltham, MA). The expression of the NQO1 gene (an Nrf2-dependent gene containing ARE sequences in its promoter region25,26 that is expressed in response to oxidative stress) was evaluated by qRT-PCR and quantified using the 2-ΔΔCt method, using predesigned TaqMan Gene Expression Assays (code Hs00168547, Applied Biosystems, Foster City, CA, Cat No. 4331182) and the TaqMan® RNA-to-CTTM 1-Step kit (Applied Biosystems, Foster City, CA, Cat No. 4392653). The run method was holding 48 °C and 15 min, 95 °C 10 min and cycling (40 cycles) 95 °C 15 sec, 60 °C 1 min. Beta-actin was used as an endogenous control.\n\nEach experiment was performed at least in duplicate. Results were expressed as the mean ± SD of at least three independent experiments. Statistical analysis was performed using the Kruskal-Wallis test and a p-value less than 0.05 was considered statistically significant. The statistical tests were applied using GraphPad Prism, version 5.02 (GraphPad Software, San Diego, CA, USA).\n\n\nResults\n\nThe preliminary phytochemical screening of the inner bark extracts obtained from T. rosea and T. chrysantha did show the presence of anthrones, quinones and coumarins, as previously reported for T. rosea6 (Table 1). Sesquiterpenic lactones were present in the n-hexane, chloroform and ethyl acetate extracts of T. rosea, but were absent from the n-hexane extract from T. chrysantha. Steroids were identified in chloroform and ethyl acetate extracts of both species. The presence of flavonoids and phenolic acids was observed only in the ethyl acetate extract from T. rosea. The total phenolic content in the extracts was determined by the Folin Ciocalteu colorimetric method. The ethyl acetate extracts obtained from T. rosea and T. chrysantha exhibited the highest total phenolic content (2.18 ± 0.29 and 2.08 ± 0.72 mg GAE/g extract, respectively), whereas the chloroform and aqueous extracts displayed the lowest phenolic content (0.63 ± 0.11, 1.55 ± 0.78, -0.668 ± 0.23 and 0.07 ± 0.03 mg GAE/g extract, respectively). The total phenolic content of the ethyl acetate extract from T. rosea was significantly higher (p <0.05) than its chloroform extract. The order of total phenolic content in both species is: ethyl acetate> n-butanol> chloroform> aqueous (Table 2). The ORAC results indicated that the ethyl acetate extracts from T. rosea, at a concentration of 10 µg/mL, have the best antioxidant activity (12,523.41 ± 840.46 µmol TE/g extract) and even this activity was superior to that obtained with the controls, showing a significant difference (p <0. 05) in the antioxidant activity of quercetin (Table 2). Both the chloroform and n-butanol extracts from T. rosea also showed important activity. Among the T. chrysantha extracts, the ethyl acetate extract displayed the best antioxidant activity (6,325.74 ± 1,057.14 µmol TE/g extract); however, this activity did not exceed that obtained with luteolin and quercetin (Table 2). The MTT assay revealed that neither the extracts from the inner bark of T. rosea nor those from T. chrysantha affected the viability of the cells studied, since viability was greater than 80% after 24 hours of exposure (Table 2).\n\n+: Presence of compounds; -: Absence of compounds; KOH: Potassium hydroxide; AlCl3: Aluminum chloride;\n\nFeCl3: Ferric chloride.\n\nTE: Trolox equivalents; GAE: Gallic acid equivalents; nd: Not determined. All experiments were carried out in triplicate, and the data represent the mean ± SD. Kruskal Wallis * p<0.05, *** p<0.001.\n\nThe Nrf2 detection test allowed for the evaluation of the ability of the extracts to activate and translocate Nrf2 to the nucleus. Nrf2 detection enabled comparisons of the basal Nrf2 status in both the cytosol and the nucleus. It also allowed for comparison of their associated changes after the exposure of HepG2 cells to the ethyl acetate extracts from T. chrysantha (0.5 μg/mL) and T. rosea (1 μg/mL), which displayed the best antioxidant activity in the ORAC assay. As shown in Figure 1, the exposure of HepG2 cells to ALA, CUR, H2O2, and the ethyl acetate extract from both T. chrysantha and T. rosea, decreased the Nrf2 levels in the cytoplasm after 4 hours of exposure. This decrease was measured in relation to their basal level (non-exposed cells). As expected, an increase in Nrf2 levels in the nucleus was observed after exposure to ALA, CUR, H2O2 and the extracts. However, significant differences were found only after exposure to ALA, CUR and, H2O2 (p<0.01).\n\nKruskal Wallis **p<0.01, *** p<0.001. ALA: Alpha-lipoic acid; CUR: Curcumin.\n\nTranscriptional regulation of antioxidant response genes against oxidative stress represents a defense against cell damage. In this study, we evaluated the expression of the gene coding for the antioxidant enzyme NQO1, which is involved in protection against oxidative stress. The level of expression of the NQO1 gene was evaluated (qRT-PCR) and quantified using the 2-ΔΔCt method. The results indicate that the ethyl acetate extract from both T. chrysantha and T. rosea, as well as the culture of HepG2 cells in the presence of H2O2, significantly increased the expression levels of NQO1 after six hours of exposure (p<0.05), compared to the controls ALA and CUR (Figure 2). The relative expression levels of NQO1 gene decreased significantly after 12 hours post-exposure.\n\nKruskal Wallis * p<0.05. ALA: Alpha-lipoic acid; CUR: Curcumin.\n\n\nDiscussion\n\nOxidative stress is important because of its relation with a wide variety of disorders associated with an increase in the levels of oxidative markers and damaged cellular components, such as Parkinson's disease, Alzheimer's disease, Huntington's disease, amyotrophic lateral sclerosis21, premature aging, inflammatory diseases and cancer27.\n\nPlants are widely used as sources of antioxidants due to their phenolic compounds and ability to scavenge ROS and free radicals, which makes them among the most potent and therapeutically useful biocompounds28. Some studies have evaluated the antioxidant activity in extracts obtained from T. chrysantha and T. rosea6,29,30, demonstrating the potential of these plants in the search for new molecules with significant biological effects. Previous studies have also shown the potential anti-inflammatory activity of T. chrysantha and T. rosea6,30. Such activity can also be associated with and effect Nrf2, a molecule that not only regulates oxidative/xenobiotic stress response, but also represses inflammation by opposing transcriptional upregulation of a number of pro-inflammatory cytokine genes31. It is due to this that the antioxidant activity of the inner bark extracts obtained from T. chrysantha and T. rosea and its association with the activation-dependent translocation of Nrf2 to the nucleus and the induction in the expression of NQO1 gene was evaluated.\n\nThe ethyl acetate extracts from both T. chrysantha and T. rosea displayed strong antioxidant activities due to their oxygen radical absorbance capacity, which could be related to the high total phenol content found with the Folin Ciocalteu method. This capacity could also be related to phenols previously reported in T. rosea, like gentisic acid32 or phenols found in the same genus, such as α-tocopherol and γ-tocopherol33. The results of the present study are in agreement with those previously reported for T. rosea6. This is the first study involving the ethyl acetate extract from T. chrysantha. An evaluation of the scavenging hydroxyl radical capacity of the methanolic and aqueous extracts from T. chrysantha revealed significant scavenging of the hydroxyl radical (80 and 83%, respectively), and reductions in the production of the peroxyl radical30.\n\nGiven that ethyl acetate extracts were the most active of those produced, and that they did not affect the viability of HepG2 cells, these extracts were used to evaluate effects on Nrf2 translocation and expression of antioxidant response genes. We compared basal Nrf2 levels in both the cytosol and the nucleus, as well as the changes associated with exposure to the extracts. As expected, after 4 hours of exposure of HepG2 cells to the extracts, Nrf2 translocate from the cytoplasm to the nucleus; however, this effect was more pronounced after exposure to ALA, CUR, and H2O2. Flavonoids found in the ethyl acetate extract of T. rosea6 could be related to its ability to induce the activation and translocation of Nrf2, as they possibly have the same action mechanism as resveratrol, whose ability to activate Nrf2 translocation to the nucleus in astrocytes after 2.5 hours of exposure has been demonstrated previously34. ALA may induce Nfr2 translocation to the nucleus after 1 hour of treatment35. A recent study did show that pau d’arco (an extract from the inner bark of T. impetiginosa) has the capacity to activate and translocate Nrf2 to the nucleus via a MEK (MAPK/ERK kinase)-independent mechanism36. The results show that the ethyl acetate extracts obtained from T. chrysantha and T. rosea did induce the nuclear translocation of Nrf2 in HepG2 cells. Therefore, the upregulated expression of the NQO1 gene by the ethyl acetate extracts is due to the stabilization and nuclear accumulation of Nrf2.\n\nAntioxidant activity through upregulated expression of NQO1 gene has been reported for β-lapachone37, a quinone that has previously been isolated from T. chrysantha38. It is also possible that steroids found within ethyl acetate extracts of T rosea and T. chrysantha during the Lieberman-Burchard test could be responsible for the overexpression of NQO1 gene, as that has been reported for steroids like 17β-estradiol on CCD841CoN cell line39. The Nrf2 pathway is considered the most important in the cell for protection against oxidative stress, which is generated by an accumulation of ROS and/or electrophiles, leading to the oxidation of biomolecules, membrane damage, DNA adduct formation, and mutagenicity. All these changes lead to degeneration of tissues, premature aging, apoptotic cell death and the development of cancer13.\n\nNrf2 is a major activator of the phase II antioxidant genes such as SOD, CAT, GST, HO-1 and NQO140. Our results demonstrated that the ethyl acetate extracts from T. chrysantha and T. rosea significantly increased the expression of NQO1 in HepG2 cells after six 6 hours of exposure compared to ALA and CUR. It has been shown that overexpression of Nrf2, by Nrf2-cDNA, upregulates the expression and induction of the NQO1 gene in response to antioxidants and xenobiotics41. In addition, Nrf2-null mice exhibited a marked decrease in NQO1 expression and induction, indicating that Nrf2 plays an essential role in the in vivo regulation of NQO1 in response to oxidative stress13. NQO1 overexpression is also considered a cytoprotective mechanism after exposure to toxic metals42.\n\n\nConclusion\n\nThe present study indicates that the ethyl acetate extracts obtained from the inner bark of T. chrysantha and T. rosea have promising antioxidant activity, as measured by the ORAC method. Both biocompounds have the ability to activate and translocate the Nrf2 transcription factor, inducing the expression of the NQO1 gene. These results reinforce the importance of these plants in the search for new antioxidant molecules.\n\n\nData availability\n\nUnderlying data is available from Open Science Framework.\n\nOSF: Dataset 1. Nrf2-Mediated Antioxidant Activity Tabebuia. https://doi.org/10.17605/OSF.IO/9WZ8643\n\nLicence: CC0 1.0 Universal (CC0 1.0) Public Domain Dedication",
"appendix": "Grant information\n\nThis study was supported by Patrimonio Autónomo Fondo Nacional de Financiamiento para la Ciencia, la Tecnología y la Innovación, Francisco José de Caldas, Contrato RC-0572-2012-Bio-Red-Co-CENIVAM, Universidad Tecnológica de Pereira (Project 9-14-5), Fundación Universitaria Autónoma de las Américas and Sistema General de Regalías de Colombia (Código BPIN 2012000100050).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nBittencourt NS Jr, Pereira EJ Jr, de Souza São-Thiago P, et al.: The reproductive biology of Cybistax antisyphilitica (Bignoniaceae), a characteristic tree of the South American savannah-like “Cerrado” vegetation. Flora - Morphology, Distribution, Functional Ecology of Plants. 2011; 206(10): 872–86. Publisher Full Text\n\nDíaz F, Medina JD: Furanonaphthoquinones from Tabebuia ochracea ssp. neochrysanta. J Nat Prod. 1996; 59(4): 423–4. Publisher Full Text\n\nGómez Castellanos JR, Prieto JM, Heinrich M: Red Lapacho (Tabebuia impetiginosa)--a global ethnopharmacological commodity? J Ethnopharmacol. 2009; 121(1): 1–13. PubMed Abstract | Publisher Full Text\n\nGentry AH: A Synopsis of Bignoniaceae Ethnobotany and Economic Botany. Ann Mo Bot Gard. 1992; 79(1): 53–64. Publisher Full Text\n\nGómez-Estrada H, Díaz-Castillo F, Franco-Ospina L, et al.: Folk medicine in the northern coast of Colombia: an overview. J Ethnobiol Ethnomed. 2011; 7: 27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJimenez-Gonzalez FJ, Vélez-Gomez JM, Melchor-Moncada JJ, et al.: Antioxidant, anti-inflammatory, and antiproliferative activity of extracts obtained from Tabebuia Rosea (Bertol.) DC. Pharmacogn Mag. 2018; 14(55): 25–31. Publisher Full Text\n\nJiménez-González FJ, Veloza LA, Sepúlveda-Arias JC: Anti-infectious activity in plants of the genus Tabebuia. Univ Sci (Bogota). 2013; 18(3): 257–67. Publisher Full Text\n\nSichaem J, Kaennakam S, Siripong P, et al.: Tabebuialdehydes A-C, cyclopentene dialdehyde derivatives from the roots of Tabebuia rosea. Fitoterapia. 2012; 83(8): 1456–9. PubMed Abstract | Publisher Full Text\n\nBryan HK, Olayanju A, Goldring CE, et al.: The Nrf2 cell defence pathway: Keap1-dependent and -independent mechanisms of regulation. Biochem Pharmacol. 2013; 85(6): 705–17. PubMed Abstract | Publisher Full Text\n\nTaguchi K, Motohashi H, Yamamoto M: Molecular mechanisms of the Keap1–Nrf2 pathway in stress response and cancer evolution. Genes Cells. 2011; 16(2): 123–40. PubMed Abstract | Publisher Full Text\n\nDhakshinamoorthy S, Long Ii DJ, Jaiswal AK: Antioxidant regulation of genes encoding enzymes that detoxify xenobiotics and carcinogens. In: Earl RS Chock PB, editors. Current Topics in Cellular Regulation. Academic Press; 2001; 36: 201–16. Publisher Full Text\n\nJaiswal AK: Nrf2 signaling in coordinated activation of antioxidant gene expression. Free Radic Biol Med. 2004; 36(10): 1199–207. PubMed Abstract | Publisher Full Text\n\nKaspar JW, Niture SK, Jaiswal AK: Nrf2:INrf2(Keap1) Signaling in Oxidative Stress. Free Radic Biol Med. 2009; 47(9): 1304–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKlaassen CD, Reisman SA: Nrf2 the rescue: effects of the antioxidative/electrophilic response on the liver. Toxicol Appl Pharmacol. 2010; 244(1): 57–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi AN, Li S, Li HB, et al.: Total phenolic contents and antioxidant capacities of 51 edible and wild flowers. J Funct Foods. 2014; 6: 319–30. Publisher Full Text\n\nOu B, Hampsch-Woodill M, Prior RL: Development and Validation of an Improved Oxygen Radical Absorbance Capacity Assay Using Fluorescein as the Fluorescent Probe. J Agric Food Chem. 2001; 49(10): 4619–26. PubMed Abstract | Publisher Full Text\n\nAlarcón E, Campos AM, Edwards AM, et al.: Antioxidant capacity of herbal infusions and tea extracts: A comparison of ORAC-fluorescein and ORAC-pyrogallol red methodologies. Food Chem. 2008; 107(3): 1114–9. Publisher Full Text\n\nAtala E, Vásquez L, Speisky H, et al.: Ascorbic acid contribution to ORAC values in berry extracts: An evaluation by the ORAC-pyrogallol red methodology. Food Chem. 2009; 113(1): 331–5. Publisher Full Text\n\nCao G, Prior RL: [5]Measurement of oxygen radical absorbance capacity in biological samples. Methods Enzymol. Academic Press; 1999; 299: 50–62. Publisher Full Text\n\nMosmann T: Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods. 1983; 65(1): 55–63. Publisher Full Text\n\nHsu WH, Lee BH, Huang YC, et al.: Ankaflavin, a novel Nrf-2 activator for attenuating allergic airway inflammation. Free Radic Biol Med. 2012; 53(9): 1643–51. PubMed Abstract | Publisher Full Text\n\nShay KP, Michels AJ, Li W, et al.: Cap-independent Nrf2 translation is part of a lipoic acid-stimulated detoxification stress response. Biochim Biophys Acta. 2012; 1823(6): 1102–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBalogun E, Hoque M, Gong P, et al.: Curcumin activates the haem oxygenase-1 gene via regulation of Nrf2 and the antioxidant-responsive element. Biochem J. 2003; 371(Pt 3): 887–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYap WH, Khoo KS, Ho ASH, et al.: Maslinic acid induces HO-1 and NOQ1 expression via activation of Nrf2 transcription factor. Biomedicine & Preventive Nutrition. 2012; 2(1): 51–8. Publisher Full Text\n\nKoriyama Y, Kamiya M, Takadera T, et al.: Protective action of nipradilol mediated through S-nitrosylation of Keap1 and HO-1 induction in retinal ganglion cells. Neurochem Int. 2012; 61(7): 1242–53. PubMed Abstract | Publisher Full Text\n\nHuerta-Olvera SG, Macías-Barragán J, Ramos-Márquez ME, et al.: Alpha-lipoic acid regulates heme oxygenase gene expression and nuclear Nrf2 activation as a mechanism of protection against arsenic exposure in HepG2 cells. Environ Toxicol Pharmacol. 2010; 29(2): 144–9. PubMed Abstract | Publisher Full Text\n\nSosa V, Moliné T, Somoza R, et al.: Oxidative stress and cancer: an overview. Ageing Res Rev. 2013; 12(1): 376–90. PubMed Abstract | Publisher Full Text\n\nJe JY, Lee DB: Nelumbo nucifera leaves protect hydrogen peroxide-induced hepatic damage via antioxidant enzymes and HO-1/Nrf2 activation. Food Funct. 2015; 6(6): 1911–8. PubMed Abstract | Publisher Full Text\n\nFranco Ospina LA, Castro Guerrero JP, Ocampo Buendía YC, et al.: Actividad antiinflamatoria, antioxidante y antibacteriana de dos especies del género Tabebuia. Rev Cubana Plant Med. 2013; 18(1): 34–46. Reference Source\n\nOspina GLF, Aragón NDM, Vergel NE, et al.: Anti-inflammatory and antioxidant activities of Phenax rugosus (POIR.) WEDD and Tabebuia chrysanta G. NICHOLSON. Vitae. 2011; 18(1): 49–55. Reference Source\n\nKobayashi EH, Suzuki T, Funayama R, et al.: Nrf2 suppresses macrophage inflammatory response by blocking proinflammatory cytokine transcription. Nat Commun. 2016; 7: 11624. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGriffiths LA: On the Distribution of Gentisic Acid in Green Plants. J Exp Bot. 1959; 10(3): 437– 42. Publisher Full Text\n\nPires TC, Dias MI, Calhelha RC, et al.: Bioactive Properties of Tabebuia impetiginosa-Based Phytopreparations and Phytoformulations: A Comparison between Extracts and Dietary Supplements. Molecules. 2015; 20(12): 22863–71. PubMed Abstract | Publisher Full Text\n\nErlank H, Elmann A, Kohen R, et al.: Polyphenols activate Nrf2 in astrocytes via H2O2, semiquinones, and quinones. Free Radic Biol Med. 2011; 51(12): 2319–27. PubMed Abstract | Publisher Full Text\n\nKoriyama Y, Nakayama Y, Matsugo S, et al.: Protective effect of lipoic acid against oxidative stress is mediated by Keap1/Nrf2-dependent heme oxygenase-1 induction in the RGC-5 cellline. Brain Res. 2013; 1499: 145– 57. PubMed Abstract | Publisher Full Text\n\nRichter M, Winkel AF, Schummer D, et al.: Pau d'arco activates Nrf2-dependent gene expression via the MEK/ERK-pathway. J Toxicol Sci. 2014;39(2): 353–61. PubMed Abstract | Publisher Full Text\n\nPark JS, Lee YY, Kim J, et al.: β-Lapachone increases phase II antioxidant enzyme expression via NQO1-AMPK/PI3K-Nrf2/ARE signaling in rat primary astrocytes. Free Radic Biol Med. 2016; 97: 168–78. PubMed Abstract | Publisher Full Text\n\nBurnett AR, Thomson RH: Naturally occurring quinones. Part XII. Extractives from Tabebuia chrysantha nichols and other bignoniaceae. J Chem Soc. 1968; 850–3. Publisher Full Text\n\nSon HJ, Kim N, Song CH, et al.: 17β-Estradiol reduces inflammation and modulates antioxidant enzymes in colonic epithelial cells. Korean J Intern Med. 2018. PubMed Abstract | Publisher Full Text\n\nChen B, Lu Y, Chen Y, et al.: The role of Nrf2 in oxidative stress-induced endothelial injuries. J Endocrinol. 2015; 225(3): R83–R99. PubMed Abstract | Publisher Full Text\n\nVenugopal R, Jaiswal AK: Nrf1 and Nrf2 positively and c-Fos and Fra1 negatively regulate the human antioxidant response element-mediated expression of NAD(P)H:quinone oxidoreductase1 gene. Proc Natl Acad Sci U S A. 1996; 93(25): 14960–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAmara IE, El-Kadi AO: Transcriptional modulation of the NAD(P)H:quinone oxidoreductase 1 by mercury in human hepatoma HepG2 cells. Free Radic Biol Med. 2011; 51(9): 1675–85. PubMed Abstract | Publisher Full Text\n\nSepúlveda-Arias JC: Nrf2-Mediated Antioxidant Activity Tabebuia. 2018. http://www.doi.org/10.17605/OSF.IO/9WZ86"
}
|
[
{
"id": "42264",
"date": "09 Jan 2019",
"name": "Jaime Martin‐Franco",
"expertise": [
"Reviewer Expertise Natural Products and free radical"
],
"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\nTabebuia is the largest genus of the Bignoniaceae family and several bioactive compounds such as: naphthoquinones; quinones and phenols have been extracted so far, due to the traditional use of numerous species of this genus. Very few studies have evaluated the anti-oxidant activity of extracts obtained from both T. rosea and T. chrysantha and only one study has evaluated the effect of T. impetiginosa extracts on Nrf-2 activation and translocation.\nThe study from Garzón-Castaño et al. describes for the first time the effect of the inner bark extracts from T. rosea and T. chrysantha on the Nrf-2-mediated antioxidant activity. The best antioxidant activity (evaluated using the ORAC Technique) was displayed by the ethyl acetate extracts from both species. This activity was clearly associated with the activation and translocation of Nrf-2 to the nucleus and the induction in the expression of the NQO1 gene in HepG2 cells. This study contributes to the knowledge on the biological activity of plants from the genus Tabebuia that have not been studied extensively and highlight its importance on the search of new molecules with antioxidant activity.\nFinally, the research paper is very clear and presents results of great expectation.\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": "42265",
"date": "23 Jan 2019",
"name": "Silvia Quesada",
"expertise": [
"Reviewer Expertise Biological activities of plants and fruits extracts",
"especially anti-inflammatory",
"antioxidant and cytotoxic activities."
],
"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 the preliminary phytochemical screening, total phenolic content and ORAC activity of different extracts obtained from the inner bark of Tabebuia rosea and T. chrysantha. The goal of the article is to study if the antioxidant activity of ethyl acetate extracts is mediated by Nrf2 nuclear translocation and its induction of antioxidant response gene (NQO1).\nThe methodology as the results are clearly presented with enough detail to allow replication by others. Both ethyl acetate extracts present similar total phenolic content, however T. rosea ORAC activity is almost twice T. chrysantha activity. Also both extracts showed a similar preliminary phytochemical screening, with a little difference; T. rosea showed a positive FeCl3 test but a negative AlCl3, otherwise T. chrysantha presented a negative FeCl3 test but a positive AlCl3. Is this difference important?\nIn the tests: Nrf2 activation and expression of NQO1, I think you should explain the purpose of using H2O2.\nPlease find an annotated copy of the article here for further suggestions and comments.\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": "4408",
"date": "06 Feb 2019",
"name": "Juan Carlos Sepúlveda-Arias",
"role": "Author Response",
"response": "Dear Dr. Quesada. Thank you very much for your evaluation. We included in the manuscript that H2O2 was used as an oxidative stress inductor. We also included in the new version the suggestions you highlight in the PDF."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1937
|
https://f1000research.com/articles/7-1988/v1
|
31 Dec 18
|
{
"type": "Review",
"title": "Epidemiology of zoonotic tick-borne diseases in Latin America: Are we just seeing the tip of the iceberg?",
"authors": [
"Alfonso J. Rodriguez-Morales",
"D. Katterine Bonilla-Aldana",
"Samuel E. Idarraga-Bedoya",
"Juan J. Garcia-Bustos",
"Jaime A. Cardona-Ospina",
"D. Katterine Bonilla-Aldana",
"Samuel E. Idarraga-Bedoya",
"Juan J. Garcia-Bustos",
"Jaime A. Cardona-Ospina"
],
"abstract": "Ticks are responsible for transmission of multiple bacterial, parasitic and viral diseases. Tick-borne diseases (TBDs) occur particularly in tropical and also subtropical areas. The frequency of these TBDs has been increasing and extending to new territories in a significant way, partly since ticks’ populations are highly favored by prevailing factors such as change in land use patterns, and climate change. Therefore, in order to obtain accurate estimates of mortality, premature mortality, and disability associated about TBDs, more molecular and epidemiological studies in different regions of the world, including Latin America, are required. In the case of this region, there is still a limited number of published studies. In addition, there is recently the emergence and discovering of pathogens not reported previously in this region but present in other areas of the world. In this article we discuss some studies and implications about TBDs in Latin America, most of them, zoonotic and with evolving taxonomical issues.",
"keywords": [
"Tick-borne disease",
"zoonoses",
"Anaplasma",
"Babesia",
"Borrelia",
"Ehrlichia",
"Rickettsia",
"epidemiology",
"public health"
],
"content": "\n\nOver the past decades there have been significant achievements in the understanding of tick borne diseases (TBDs), which are mostly zoonoses and classed as neglected diseases1–5. Their occurrence is significant in tropical and subtropical areas, leading to an important impact on public health as well as the economy, as they affect humans, domestic animals and livestock, among others6. Knowledge of the occurrence of these diseases in animal species is of utmost importance for the understanding of the risk for human infection7. Ticks, and animals, including human beings, interact with nature, and their environmental and ecological interactions regulate the populations of ticks and vertebrates, determining their contact rates and the circulation of the diseases8. Moreover, although most TBDs are tick-borne, other ways of transmission can occur. For instance, bovine anaplasmosis, caused by Anaplasma marginale, can be spread through mechanical means, e.g. biting arthropods or by contaminated fomites, like needles, ear tagging, dehorning and castration equipment9. A. marginale can be also transmitted transplacentally, which has contributed to the occurrence of bovine anaplasmosis in some areas10.\n\nIn Latin America, there is a lack of studies about TBDs. However, data from Panama, Brazil, Mexico, Peru, Colombia, and Venezuela clearly show that these pathogens are prevalent when they are assessed. Recently, two lethal cases of rickettsiosis caused by Rickettsia rickettsii were reported in rural and urban of Panama, probably transmitted by the tick Rhipicephalus sanguineus s.l., accounting for the first molecular detection of this bacteria in this tick in Panama and Central America11. TBDs caused by rickettsial species are life-threatening infections that in the tropical Americas have an emerging and reemerging trend. Until some years ago, R. rickettsia was the only tick-borne species of rickettsia present in Latin America. Nowadays, multiple other species, such as R. parkeri and R. massiliae, are causing infections in humans in this region. In Peru, different ectoparasite samples have tested positive for a Rickettsia genus-specific qPCR, with strong evidence that active searching of TBDs etiological agents is required in order to improve reporting and detection12. Additional, new species are being reported; although their pathogenicity has not been definitely confirmed, they should be considered as potential pathogens13.\n\nEhrlichiosis is another TBD caused by rickettsial organisms of the genus Ehrlichia14,15. Canine monocytic ehrlichiosis caused by Ehrlichia canis apparently is highly endemic in Brazil and has been also detected in Mexico, where it is closely related to strains from the USA14. In Brazil, E. canis is the principal Ehrlichia specie found in canines and has been also detected in felines, although the prevalence has not been estimated yet. In addition, in Brazil, E. ewingii has been recently detected, as well as E. chaffeensis in marsh deers, and there is immune-epidemiological evidence that suggests the occurrence of ehrlichiosis in humans16,17, but its etiologic agent has not yet been established. Improved molecular diagnostic resources for laboratory testing will allow better identification and characterization of ehrlichial organisms associated with human ehrlichiosis in Brazil16.\n\nIn Colombia, three outbreaks of human rickettsiosis have been reported in the Northwestern region during 2006–2008, with a lethality up to 54%18. And later, in 2010–2011 a cross-sectional study revealed the presence of three different Rickettsia species: R. felis in fleas, and R. bellii and Rickettsia sp. strain Atlantic rainforest, both in Amblyomma ovale ticks18. Additionally, in Venezuela, detection of Anaplasma platys has been described in humans. Those patients were bitten by Rhipicephalus sanguineus and suffered chronic non-specific clinical signs, including headaches and muscle pains, supporting A. platys as a zoonotic tick-borne pathogen19. Other studies from Venezuela have found a high proportion of positivity of antibodies against Babesia caballi and Theileria equi in horses and other animals20, suggesting that probably the real frequency and importance of these hemoparasites are overlooked.\n\nBabesiosis is caused by any of a group of vector-borne, protozoal hemoparasites of the phylum Apicomplexa. There are more than 110 described Babesia spp. worldwide, identified from mammalian and also avian hosts21,22. This group of TBDs is transmitted by ixodid ticks and they infect a wide variety of vertebrates that maintain transmission cycles. Till today, there are multiple species of the genus Babesia that can infect people, and the regional distribution of Babesia microti is the most prevalent. In the USA, babesiosis is caused primarily by B. microti, whereas cases in European countries are commonly caused by B. divergens22. In countries such as the USA its incidence has increased 260% between 2005–201023, with a proportion of 40% of cases of Lyme disease reporting co-infections with Babesia. In Latin American countries, particularly Colombia, the seroprevalence of Babesia has been reported has high as 30% from people of urban and rural locations24. Additionally, recently, new Borrelia burgdorferi sensu lato strains or new related species have been described in countries such as Uruguay, Brazil and Chile25.\n\nIn the USA, some authors suggest that approximately 95% of ~50 thousand cases of locally acquired vector-borne diseases in humans reported annually to the Centers for Disease Control and Prevention are caused by organisms that were vectorized by ticks (Table 1)6,26. Beyond the Americas, in other regions of the world, like in Europe, ticks are the main vectors of animal and human organisms. Ticks transmit several viral agents, called tick-borne viruses (TBV), such as tick-borne encephalitis virus and Crimean-Congo hemorrhagic fever virus, which have reemerged in multiple areas of the world27. TBV have a natural cycle between ticks and wild animals in nature, with humans as accidental hosts27,28. Emerging TBVs are continually discovered, probably related to the increase of tick populations in different regions of the planet and invasion of human beings into areas infested by ticks27,28.\n\nModified from Eisen et al.6.\n\nDetection and sentinel surveillance of TBD require molecular tools for diagnosis29, for example, serological tests have proven to be inconclusive in diagnose Lyme disease30. The use of molecular biology tests in recent years has increased the sensitivity and specificity of the diagnosis of infections in the group of Rickettsiales. Molecular diagnosis enables the accurate identification not only at genus level, but species, providing additional characterization on the epidemiology and the evolution of clinical disease. Furthermore, PCR as well enzyme restriction tests of the vector blood meal can be employed to analyze their feeding source and possibly identify the ecological reservoir of the organisms31. Etiological agents of the group of Rickettsial, including those in the genuses Anaplasma, Neorickettsia, Ehrlichia, and Rickettsia, are relevant and often vector-borne organisms of canines and felines, but also of bovine, livestock and other animals, which appears to be a wide range of hosts10,32.\n\n\nConclusions\n\nBesides the limited number of studies in Latin America on TBDs, the prevalence of these diseases is increasing, triggered by globalization, as well the impact climate change and variability. Tick and TBDs investigators, vet doctors, medical and public health practitioners should work to share their expertise on different aspects of TBDs, such as tick ecology, disease transmission, diagnostics, and treatment, in order to face the challenges of scientific, political, and public engagement for TBD research and control in this region33. Systematic reviews as well as observational analyses are necessary in order to understand the current situation of TBDs. Molecular tools can provide valuable information for understanding the evolution of their etiological agents, as well as provide insights into host-pathogen-vector-environment interactions. Probably, what we have seen till now is just the tip of an iceberg and there is a need for more studies in Latin America about TBDs.\n\n\nData availability\n\nNo data is associated with this article.",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nTalagrand-Reboul E, Boyer PH, Bergstrom S, et al.: Relapsing Fevers: Neglected Tick-Borne Diseases. Front Cell Infect Microbiol. 2018; 8: 98. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMolina-Hoyos K, Montoya-Ruiz C, Díaz F, et al.: Enfermedades virales transmitidas por garrapatas. scielo.org.co. 2018; 31(1): 36–50. Publisher Full Text\n\nTirosh-Levy S, Gottlieb Y, Apanaskevich DA, et al.: Species distribution and seasonal dynamics of equine tick infestation in two Mediterranean climate niches in Israel. Parasit Vectors. 2018; 11(1): 546. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCheng TY, Chen Z, Li ZB, et al.: First Report of Ixodes nipponensis Infection in Goats in China. Vector Borne Zoonotic Dis. 2018; 18(10): 575–8. PubMed Abstract | Publisher Full Text\n\nYousefi A: Phylogenetic analysis of Anaplasma marginale and Anaplasma ovis isolated from small ruminant based on MSP4 gene in western regions of Iran. Comp Clin Path. 2018; 27(5): 1161–1165. Publisher Full Text\n\nEisen RJ, Kugeler KJ, Eisen L, et al.: Tick-Borne Zoonoses in the United States: Persistent and Emerging Threats to Human Health. ILAR J. 2017; 58(3): 319–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAndersson MO, Marga G, Banu T, et al.: Tick-borne pathogens in tick species infesting humans in Sibiu County, central Romania. Parasitol Res. 2018; 117(5): 1591–7. PubMed Abstract | Publisher Full Text\n\nEstrada-Pena A, de la Fuente J: Host Distribution Does Not Limit the Range of the Tick Ixodes ricinus but Impacts the Circulation of Transmitted Pathogens. Front Cell Infect Microbiol. 2017; 7: 405. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOstfeld RS, Brunner JL: Climate change and Ixodes tick-borne diseases of humans. Philos Trans R Soc Lond B Biol Sci. 2015; 370(1665): pii: 20140051. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAubry P, Geale DW: A review of bovine anaplasmosis. Transbound Emerg Dis. 2011; 58(1): 1–30. PubMed Abstract | Publisher Full Text\n\nMartínez-Caballero A, Moreno B, González C, et al.: Descriptions of two new cases of Rocky Mountain spotted fever in Panama, and coincident infection with Rickettsia rickettsii in Rhipicephalus sanguineus s.l. in an urban locality of Panama City, Panama. Epidemiol Infect. 2018; 146(7): 875–8. PubMed Abstract | Publisher Full Text\n\nFlores-Mendoza C, Florin D, Felices V, et al.: Detection of Rickettsia parkeri from within Piura, Peru, and the first reported presence of Candidatus Rickettsia andeanae in the tick Rhipicephalus sanguineus. Vector Borne Zoonotic Dis. 2013; 13(7): 505–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbarca K, Oteo JA: [Clinical approach and main tick-borne rickettsiosis present in Latin America]. Rev Chilena Infectol. 2014; 31(5): 569–76. PubMed Abstract | Publisher Full Text\n\nAlmazán C, González-Álvarez VH, Fernández de Mera IG, et al.: Molecular identification and characterization of Anaplasma platys and Ehrlichia canis in dogs in Mexico. Ticks Tick Borne Dis. 2016; 7(2): 276–83. PubMed Abstract | Publisher Full Text\n\nHerrmann JA, Dahm NM, Ruiz MO, et al.: Temporal and Spatial Distribution of Tick-Borne Disease Cases among Humans and Canines in Illinois (2000-2009). Environ Health Insights. 2014; 8(Suppl 2): 15–27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVieira RF, Biondo AW, Guimarães AM, et al.: Ehrlichiosis in Brazil. Rev Bras Parasitol Vet. 2011; 20(1): 1–12. PubMed Abstract | Publisher Full Text\n\nPaddock CD, Childs JE: Ehrlichia chaffeensis: a prototypical emerging pathogen. Clin Microbiol Rev. 2003; 16(1): 37–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLondoño AF, Acevedo-Gutiérrez LY, Marín D, et al.: Wild and domestic animals likely involved in rickettsial endemic zones of Northwestern Colombia. Ticks Tick Borne Dis. 2017; 8(6): 887–94. PubMed Abstract | Publisher Full Text\n\nArraga-Alvarado CM, Qurollo BA, Parra OC, et al.: Case report: Molecular evidence of Anaplasma platys infection in two women from Venezuela. Am J Trop Med Hyg. 2014; 91(6): 1161–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMujica FF, Perrone T, Forlano M, et al.: Serological prevalence of Babesia caballi and Theileria equi in horses of Lara State, Venezuela. Vet Parasitol. 2011; 178(1–2): 180–3. PubMed Abstract | Publisher Full Text\n\nPastor A, Milnes E: 92 - Babesiosis in Cervidae. In: Mille RE, Lamberski N, Calle PP, editors. Fowler's Zoo and Wild Animal Medicine Current Therapy. W.B. Saunders, 2019; 9: 647–55.\n\nHomer MJ, Aguilar-Delfin I, Telford SR 3rd, et al.: Babesiosis. Clin Microbiol Rev. 2000; 13(3): 451–69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nApostolou A, Sorhage F, Tan C: Babesiosis surveillance, new jersey, USA, 2006-2011. Emerg Infect Dis. 2014; 20(8): 1407–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBuelvas F, Alvis N, Buelvas I, et al.: [A high prevalence of antibodies against Bartonella and Babesia microti has been found in villages and urban populations in Cordoba, Colombia]. Rev Salud Publica (Bogota). 2008; 10(1): 168–77. PubMed Abstract\n\nRobles A, Fong J, Cervantes J: [Borrelia Infection in Latin America]. Rev Invest Clin. 2018; 70(4): 158–63. PubMed Abstract | Publisher Full Text\n\nAdams DA, Thomas KR, Jajosky RA, et al.: Summary of Notifiable Infectious Diseases and Conditions - United States, 2014. MMWR Morb Mortal Wkly Rep. 2016; 63(54): 1–152. PubMed Abstract | Publisher Full Text\n\nFailloux AB, Bouattour A, Faraj C, et al.: Surveillance of Arthropod-Borne Viruses and Their Vectors in the Mediterranean and Black Sea Regions Within the MediLabSecure Network. Curr Trop Med Rep. 2017; 4(1): 27–39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrackney DE, Armstrong PM: Transmission and evolution of tick-borne viruses. Curr Opin Virol. 2016; 21: 67–74. PubMed Abstract | Publisher Full Text\n\nFaccini-Martinez ÁA, Hidalgo M: [Speaking of Latin American guidelines for the diagnosis of tick-transmitted rickettsiosis]. Rev Chilena Infectol. 2014; 31(3): 354. PubMed Abstract | Publisher Full Text\n\nLeeflang MM, Ang CW, Berkhout J, et al.: The diagnostic accuracy of serological tests for Lyme borreliosis in Europe: a systematic review and meta-analysis. BMC Infect Dis. 2016; 16: 140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuillemi EC, Tomassone L, Farber MD: Tick-borne Rickettsiales: Molecular tools for the study of an emergent group of pathogens. J Microbiol Methods. 2015; 119: 87–97. PubMed Abstract | Publisher Full Text\n\nAllison RW, Little SE: Diagnosis of rickettsial diseases in dogs and cats. Vet Clin Pathol. 2013; 42(2): 127–44. PubMed Abstract | Publisher Full Text\n\nSakamoto JM: Progress, challenges, and the role of public engagement to improve tick-borne disease literacy. Curr Opin Insect Sci. 2018; 28: 81–9. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "42312",
"date": "30 Jan 2019",
"name": "Lidia Chitimia-Dobler",
"expertise": [
"Reviewer Expertise Ticks and tick-borne 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\nTitle and content of this work do not fit together.\nWhat about the taxonomical issues stated at the end of the abstract? It is not discussed in the paper.\nFor a review, it is not very comprehensive and does not focus on Latin America. The structure of the manuscript is not logical, starting with Rickettsia, continue with Ehrlichia, coming back to Rickettsia. There is too much content about the USA for a short review referring to Latin America.\nIn the table, there are a number of tick-borne diseases which do not occur in Latin America E.g. Colorado tick fever, Heartland virus diseases or Powassan encephalitis. It is not acceptable to simply transfer data from USA to Latin America.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? No\n\nAre all factual statements correct and adequately supported by citations? No\n\nIs the review written in accessible language? Yes\n\nAre the conclusions drawn appropriate in the context of the current research literature? No",
"responses": [
{
"c_id": "4396",
"date": "04 Feb 2019",
"name": "Alfonso Rodriguez-Morales",
"role": "Author Response",
"response": "Dear Dr. Chitima-DoblerThanks for your valuable comments. We want to answer all your comments, as we improved significantly our manuscript based on yours as well as on the other reviewer.Title and content of this work do not fit together.Well, we have significantly improved the manuscript, which was originally intended as an Opinion Article. As an Advisor of this Gateway, called, Disease Outbreaks, I should explain to you that at this open publishing platform where there are Opinion Articles, Review Articles and Systematic Review Articles.This article is NOT a Systematic Review, it was originally submitted as an Opinion Article, as an invitation from F1000Research as being F1000Research Disease Outbreaks Gateway Advisor, but later classified as Review (narrative review). As you well read, we only referred to some examples of studies in Latin America, about tick-borne diseases, that illustrate the problem, in terms of a neglected group of conditions in the region, wherein most of the countries are not under surveillance, and there is still a lack of studies, but even more actions for effective control.What about the taxonomical issues stated at the end of the abstract? It is not discussed in the paper.We have improved on the updated taxonomy.For a review, it is not very comprehensive and does not focus on Latin America. The structure of the manuscript is not logical, starting with Rickettsia, continue with Ehrlichia, coming back to Rickettsia. There is too much content about the USA for a short review referring to Latin America.As we mentioned, this is not a Systematic Review, is an opinion article, published by decision of F1000Research as a Review (narrative). We focused now more in Latin America.In the table, there are a number of tick-borne diseases which do not occur in Latin America E.g. Colorado tick fever, Heartland virus diseases or Powassan encephalitis. It is not acceptable to simply transfer data from USA to Latin America. We significantly changed and corrected the table."
}
]
},
{
"id": "42313",
"date": "30 Jan 2019",
"name": "Joyce M Sakamoto",
"expertise": [
"Reviewer Expertise Ticks and microbiology"
],
"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 entitled \"Epidemiology of zoonotic tick-borne diseases in Latin America: Are we just seeing the tip of the iceberg?\" suggests that more studies on TTBD (ticks and tick-borne diseases) are needed worldwide, and particularly in Latin America, but there is a dearth of available data published.\n\nBecause the authors emphasized that their manuscript was not intended to be an extensive review article, but was originally presented as an opinion piece, I reviewed this manuscript as the latter. There is a lack of cohesiveness and transition between subjects within the manuscript and an overall lack of a common theme. There are also some minor errors in terminology and italicization throughout, which I will point out in each paragraph. In summary, while I appreciate the fact that this paper was originally intended to be a superficial review intended to emphasize the lack of research literature on ticks and TBD in Latin America, it does not do so in a cohesive manner, nor does it make a clear case for the need for further resources. Nevertheless, many Latin American countries would benefit from more resources dedicated to TBD research and surveillance. Should the authors reorganize and focus their attention to a specific objective, this might make for a stronger case.\n\nSpecific Comments:\nThe 1st paragraph gives a brief introduction to the impacts of TTBD on human and animal health, particularly in tropical and subtropical areas of the world. While it is interesting to point out that some TBD can be transmitted in other ways besides via ticks, I'm not sure that really supports the point of your manuscript. Make sure to be consistent in whether you use \"tick-borne\" or \"tick borne\" throughout the manuscript.\nThe 2nd paragraph begins with an introduction to Latin America, but then goes directly into rickettsioses, followed by a paragraph each on ehrlichioses, babesioses, and a brief mention of borrelioses. This feels much like a listing of diseases that have no connectivity with the overall purpose of the article. To help this, a sentence or two to introduce what will be discussed in the coming paragraphs might help to outline what a reader might expect to see. Maybe something like…”Here we will briefly review the known literature and highlight the increasing incidence/discovery/etc of tick-borne pathogens…”\nThere seems to be a missing word in \"Rickettsia rickettsii were reported in rural and urban of Panama\" \"Until some years ago, R. rickettsia was the only tick-borne species of rickettsia present in Latin America. Nowadays\" - 'Nowadays' is too colloquial. Perhaps \"Presently\" or \"Currently\" are better alternatives? There is what looks like an autocorrect error (\"R. rickettsia\" should be \"R. rickettsii\"). Question: Is it possible that other Rickettsia spp. were present already, but the older serological diagnostics that identified RMSF (R. rickettsii) were cross-reacting with them and were all recorded as RMSF? Many Rickettsia spp. are non-pathogenic and are instead obligate symbionts of the ticks in which they reside.\n3rd paragraph: \"specie\" should be \"species\". The genus \"Ehrlichia\" should be italicized.\n4th paragraph: I refer you to \"Rickettsia species\". Rickettsia capitalized refers to the genus and it should be italicized. If referring to the colloquial term used to refer to Rickettsiales, then it should be lowercase and not italicized (rickettsias). I believe the former applies in this sentence. \"sp.\" indicates that the species designation is not known and is not italicized.\n5th paragraph:\n\"Till\" should be \"Until\". The paragraph is almost entirely about babesioses worldwide, yet the last sentence briefly mentions Borrelia and does not flow at all. Borreliosis should be its own section, with a discussion on both Lyme borreliosis and relapsing fever variants. Perhaps the authors could mention the role that soft ticks play in relapsing fever, particularly since this is a problem in Latin America.\n6th paragraph:\nIf the purpose of this paper is to focus on Latin America and the paucity of work on ticks and tickborne disease relative to other parts of the world, there should be more focus on Latin America. The discussion of other countries detracts from this message. The only exception is in the context of potential TTBD flow between these countries via trade, human or animal migration, and impacts of climate change. Further, there is a sudden switch mid-paragraph to tickborne viruses (TBV). Is the purpose of mentioning TBV to say that TBV exist in Latin America, but no one has looked hard enough? There are a few review articles that may provide support for this, but I suggest that as it is here, there is no context and it feels like it was just added after-the-fact. Replace “vectorized” with “transmitted”. Note “Transmit” = verb, “vector” = noun. \"Vectorized” isn’t really a word, or at least not commonly used North American medical entomology.\n7th paragraph: I'm unclear on the purpose of this paragraph. Is it to highlight the advances in diagnostics that make it possible to detect TBD? If so, how does this support the overall theme of this article?\nThe last sentence of the 7th paragraph is not clear and has several errors. \"Etiological agents of the group of Rickettsial, including those in the genuses Anaplasma, Neorickettsia, Ehrlichia, and Rickettsia, are relevant and often vector-borne organisms of canines and felines, but also of bovine, live-stock and other animals, which appears to be a wide range of hosts\". Could this be moved somewhere else to make a transition or taken out completely? “group of “Rickettsial” should be either “rickettsial pathogens in the…” OR “Etiological agents in the Order Rickettsiales” \"live-stock\" should be \"livestock\" “Genuses” should be “genera\" The Order Rickettsiales is an Order and not italicized. Admittedly this order's taxonomy is problematic, but if you are going to refer to it, don't italicize it.\nTable 1:\nYour table is entitled “Examples of selected tick-borne diseases in Latin America,” yet contains tick species mostly localized to North America (D. andersoni, D. variabilis, D. occidentalis, I. pacificus, and I. scapularis). There was a review by Esteve-Gassent et al (2014)1 that suggests that some of these species could potentially spill over the Mexico-USA border and therefore these species could potentially warrant further study. Your use of this table, however, does not provide any context and feels out-of-place and irrelevant. There have been several articles detailing known hard and soft tick species and their epidemiological significance from many different Latin American countries (e.g. Rivera-Páez in 2018 gives updates to Colombian Ixodidae2; Mastropaolo 2014 reviewed both hard and soft ticks of Bolivia3; Lopes in 2016 of Belize4, and Witter in 2016 from wild animals of Brazil5, just to name a few). There have also been reviews on tick species found in the Caribbean, Cuba, and Mexico. If you can obtain it, a valuable and comprehensive resource is “The Hard Ticks of the World (Acari: Ixodida: Ixodidae)”, by Guglielmone et al (2014)6, which contains summaries of all known hard tick species worldwide, including host associations and geographic distributions.\n\nConclusion: I am afraid I don't see the relevance of this conclusion section to the rest of the article. What you need is to tie together what you have written and make a conclusion.\n\nWhat is your conclusion? Do we need more surveillance? More research? Better diagnostics? Better identification approaches? What are you trying to state that leads to your final concluding statement that this is \"just the tip of the iceberg\"? To say there is not enough data is sort of vague. More data is always better, but I myself have at least 100 references for literature on TTBD from Central and South America, and my list is not extensive. I would posit that you may need to rethink the purpose of your article. To state that there insufficient work on this topic in Latin America borders on insulting those who have spent careers working on these exact topics. There is an element that you have not discussed: the lack of infrastructure and/or funding to support continued vector surveillance studies. Are there reports comparing the estimated costs of these types of studies (surveillance as well as diagnostics)? If so, how does that compare to the estimated proportion of the national budgets that are specifically earmarked for vector-borne surveillance and public health efforts, and what part of that is allotted toward TTBD research? Perhaps this would strengthen your case for a call to action.\n\nIs the topic of the review discussed comprehensively in the context of the current literature? No\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? Partly",
"responses": [
{
"c_id": "4395",
"date": "04 Feb 2019",
"name": "Alfonso Rodriguez-Morales",
"role": "Author Response",
"response": "Dear Dr. SakamotoThanks for your valuable comments. We want to answer all your comments, as we improved significantly our manuscript based on yours as well as on the other reviewer.This manuscript entitled \"Epidemiology of zoonotic tick-borne diseases in Latin America: Are we just seeing the tip of the iceberg?\" suggests that more studies on TTBD (ticks and tick-borne diseases) are needed worldwide, and particularly in Latin America, but there is a dearth of available data published.Certainly, there are more available data published, but our original intention was present an Opinion Article. As an Advisor of this Gateway, called, Disease Outbreaks, I should explain to you that at this open publishing platform where there are Opinion Articles, Review Articles and Systematic Review Articles.This article is NOT a Systematic Review, it was originally submitted as an Opinion Article, as an invitation from F1000Research as being F1000Research Disease Outbreaks Gateway Advisor, but later classified as Review (narrative review). As you well read, we only referred to some examples of studies in Latin America, including yours (cited, now 36. Sakamoto JM: Progress, challenges, and the role of public engagement to improve tick-borne disease literacy. Curr Opin Insect Sci. 2018;28:81–9. 30551772 10.1016/j.cois.2018.05.011), about tick-borne diseases, that illustrate the problem, in terms of a neglected group of conditions in the region, wherein most of the countries are not under surveillance, and there is still a lack of studies, but even more actions for effective control.Because the authors emphasized that their manuscript was not intended to be an extensive review article, but was originally presented as an opinion piece, I reviewed this manuscript as the latter. There is a lack of cohesiveness and transition between subjects within the manuscript and an overall lack of a common theme. There are also some minor errors in terminology and italicization throughout, which I will point out in each paragraph. In summary, while I appreciate the fact that this paper was originally intended to be a superficial review intended to emphasize the lack of research literature on ticks and TBD in Latin America, it does not do so in a cohesive manner, nor does it make a clear case for the need for further resources. Nevertheless, many Latin American countries would benefit from more resources dedicated to TBD research and surveillance. Should the authors reorganize and focus their attention to a specific objective, this might make for a stronger case. We have significantly improved the grammatical and orthography of this paper. We modified the paper considering your comments.Specific Comments:The 1st paragraph gives a brief introduction to the impacts of TTBD on human and animal health, particularly in tropical and subtropical areas of the world. While it is interesting to point out that some TBD can be transmitted in other ways besides via ticks, I'm not sure that really supports the point of your manuscript. Make sure to be consistent in whether you use \"tick-borne\" or \"tick borne\" throughout the manuscript.Well, that's a comment, additionally to the main point of this article. Regard your comment of consistency in the use of tick-borne, we have checked and only left \"tick-borne\".The 2nd paragraph begins with an introduction to Latin America but then goes directly into rickettsioses, followed by a paragraph each on ehrlichioses, babesioses, and a brief mention of borrelioses. This feels much like a listing of diseases that have no connectivity with the overall purpose of the article. To help this, a sentence or two to introduce what will be discussed in the coming paragraphs might help to outline what a reader might expect to see. Maybe something like…”Here we will briefly review the known literature and highlight the increasing incidence/discovery/etc of tick-borne pathogens…” We have restructured the paragraph in that way. There seems to be a missing word in \"Rickettsia rickettsii were reported in rural and urban of Panama\". Corrected. \"Until some years ago, R. rickettsia was the only tick-borne species of rickettsia present in Latin America. Nowadays\" - 'Nowadays' is too colloquial. Perhaps \"Presently\" or \"Currently\" are better alternatives? Corrected. There is what looks like an autocorrect error (\"R. rickettsia\" should be \"R. rickettsii\"). Corrected. Question: Is it possible that other Rickettsia spp. were present already, but the older serological diagnostics that identified RMSF (R. rickettsii) were cross-reacting with them and were all recorded as RMSF? Included. Many Rickettsia spp. are non-pathogenic and are instead obligate symbionts of the ticks in which they reside. Considered. 3rd paragraph: \"specie\" should be \"species\". The genus \"Ehrlichia\" should be italicized.Done. All that should be italicized, has been done.4th paragraph: I refer you to \"Rickettsia species\". Rickettsia capitalized refers to the genus and it should be italicized. If referring to the colloquial term used to refer to Rickettsiales, then it should be lowercase and not italicized (rickettsias). I believe the former applies in this sentence. \"sp.\" indicates that the species designation is not known and is not italicized.Done. Capitalized where corresponded.5th paragraph: \"Till\" should be \"Until\". Ok. The paragraph is almost entirely about babesioses worldwide, yet the last sentence briefly mentions Borrelia and does not flow at all. Borreliosis should be its own section, with a discussion on both Lyme borreliosis and relapsing fever variants. Perhaps the authors could mention the role that soft ticks play in relapsing fever, particularly since this is a problem in Latin America. This has been modified. We discussed also relapsing fever. 6th paragraph: If the purpose of this paper is to focus on Latin America and the paucity of work on ticks and tickborne disease relative to other parts of the world, there should be more focus on Latin America. The discussion of other countries detracts from this message. The only exception is in the context of potential TTBD flow between these countries via trade, human or animal migration, and impacts of climate change. We have improved this aspect and more focused on Latin America. Further, there is a sudden switch mid-paragraph to tickborne viruses (TBV). Is the purpose of mentioning TBV to say that TBV exist in Latin America, but no one has looked hard enough? There are a few review articles that may provide support for this, but I suggest that as it is here, there is no context and it feels like it was just added after-the-fact. We have corrected this. Replace “vectorized” with “transmitted”. Note “Transmit” = verb, “vector” = noun. \"Vectorized” isn’t really a word, or at least not commonly used North American medical entomology. Corrected. 7th paragraph: I'm unclear on the purpose of this paragraph. Is it to highlight the advances in diagnostics that make it possible to detect TBD? If so, how does this support the overall theme of this article? We restructured the paragraph. The last sentence of the 7th paragraph is not clear and has several errors. \"Etiological agents of the group of Rickettsial, including those in the genuses Anaplasma, Neorickettsia, Ehrlichia, and Rickettsia, are relevant and often vector-borne organisms of canines and felines, but also of bovine, live-stock and other animals, which appears to be a wide range of hosts\". Could this be moved somewhere else to make a transition or taken out completely? Corrected. “group of “Rickettsial” should be either “rickettsial pathogens in the…” OR “Etiological agents in the Order Rickettsiales”. Corrected. \"live-stock\" should be \"livestock\". Done. “Genuses” should be “genera\". Done. The Order Rickettsiales is an Order and not italicized. Admittedly this order's taxonomy is problematic, but if you are going to refer to it, don't italicize it. Ok. Table 1: Your table is entitled “Examples of selected tick-borne diseases in Latin America,” yet contains tick species mostly localized to North America (D. andersoni, D. variabilis, D. occidentalis, I. pacificus, and I. scapularis). There was a review by Esteve-Gassent et al (2014)1 that suggests that some of these species could potentially spill over the Mexico-USA border and therefore these species could potentially warrant further study. Your use of this table, however, does not provide any context and feels out-of-place and irrelevant. We have extensively changed the table. There have been several articles detailing known hard and soft tick species and their epidemiological significance from many different Latin American countries (e.g. Rivera-Páez in 2018 gives updates to Colombian Ixodidae2; Mastropaolo 2014 reviewed both hard and soft ticks of Bolivia3; Lopes in 2016 of Belize4, and Witter in 2016 from wild animals of Brazil5, just to name a few). There have also been reviews on tick species found in the Caribbean, Cuba, and Mexico. If you can obtain it, a valuable and comprehensive resource is “The Hard Ticks of the World (Acari: Ixodida: Ixodidae)”, by Guglielmone et al (2014)6, which contains summaries of all known hard tick species worldwide, including host associations and geographic distributions. We have corrected that. Conclusion: I am afraid I don't see the relevance of this conclusion section to the rest of the article. What you need is to tie together what you have written and make a conclusion. What is your conclusion? Do we need more surveillance? More research? Better diagnostics? Better identification approaches? We included those considerations. What are you trying to state that leads to your final concluding statement that this is \"just the tip of the iceberg\"? To say there is not enough data is sort of vague. More data is always better, but I myself have at least 100 references for literature on TTBD from Central and South America, and my list is not extensive. I would posit that you may need to rethink the purpose of your article. To state that there insufficient work on this topic in Latin America borders on insulting those who have spent careers working on these exact topics. Now we have considered that and discussed in the Conclusions. There is an element that you have not discussed: the lack of infrastructure and/or funding to support continued vector surveillance studies. Are there reports comparing the estimated costs of these types of studies (surveillance as well as diagnostics)? If so, how does that compare to the estimated proportion of the national budgets that are specifically earmarked for vector-borne surveillance and public health efforts, and what part of that is allotted toward TTBD research? Perhaps this would strengthen your case for a call to action. We mentioned all of that now and tried to make a more deep call for action."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1988
|
https://f1000research.com/articles/8-169/v1
|
08 Feb 19
|
{
"type": "Research Article",
"title": "Early assessment of antiretroviral efficacy is critical to prevent the emergence of resistance mutations in HIV-tuberculosis coinfected patients: a substudy of the CARINEMO-ANRS12146 trial",
"authors": [
"Elisabeth Baudin",
"Nilesh Bhatt",
"Christine Rouzioux",
"Micaela Serafini",
"Lucas Molfino",
"Ilesh Jani",
"Anne-Marie Taburet",
"Maryline Bonnet",
"Alexandra Calmy",
"CARINEMO Study Group",
"Nilesh Bhatt",
"Christine Rouzioux",
"Micaela Serafini",
"Lucas Molfino",
"Ilesh Jani",
"Anne-Marie Taburet",
"Maryline Bonnet",
"Alexandra Calmy"
],
"abstract": "Background: In the CARINEMO ANRS 12146 clinical trial, HIV-tuberculosis co-infected patients in Mozambique were randomized to nevirapine (NVP) or to efavirenz (EFV)-based antiretroviral therapy to compare these two non-nucleoside reverse transcriptase inhibitors (NNRTIs) in treatment naïve patients. Methods: In this sub study, we explored the relationship of NNRTI concentrations with virological escape and the possible emergence of resistance mutations at week 48. The virological escape was defined as an HIV-RNA above 400 copies/m at week 48. Results: Among the 570 randomized patients, 470 (82%) had an HIV-RNA result at week 48; 54 (12.1%) patients had a viral escape and 35 patients had at least one major resistance mutation detected. Low drug concentration at weeks 12 and 24 (below the 10th percentile) were independently associated with virologic escape at week 48 (adjusted odds ratio [aOR]=2.9; 95% CI: 1.1 -7.2; p=0.0312 and aOR=4.2; 95% CI: 1.8-9.8; p=0.0019, respectively), and independently associated with an increased risk of emergence of resistance mutation (aOR=4.5; 95% CI: 1.8-14.6; p=0.009 at week 12; aOR=5.1; 95% CI: 1.8-14.6 at week 24). Receiver operating characteristic curves analyses indicated a better predictability of the mid-dose concentration and of the HIV-1 RNA values on resistance mutations in contrast to virological escape. Conclusions: Very low drug plasma concentrations early after treatment initiation (week 12) were predictive factors of virological escape and the emergence of resistance mutations at week 48, and early monitoring of drug intake may prevent the occurrence of late virological escape and the selection of vial resistance mutations.",
"keywords": [
"HIV/TB coinfection",
"NNRTI concentrations",
"drug–drug interactions",
"antiretroviral therapy",
"resistance",
"virological escape"
],
"content": "Introduction\n\nAntiretroviral therapy (ART) aims to sustain virological suppression, which is associated with a clinical benefit and immune recovery. It also prevents HIV transmission and limits the emergence of antiretroviral (ARV) drug resistance. In a recent meta-analysis, Gupta et al. reported that East Africa had the highest estimated rate of drug-resistance mutations (29% per year) since the roll-out of ART, with an estimated prevalence of ARV drug resistance of 7.4% at 8 years after rollout1.\n\nIn 2016, 80% of the worldwide prescription of ART-included efavirenz (EFV), a non-nucleoside reverse-transcriptase inhibitor (NNRTI)-class drug2. Efavirenz-based ART is also recommended in the context of tuberculosis (TB) coinfection, as drug-drug interactions with rifampicin, a cornerstone anti-TB drug, are limited. However, the risk of central nervous toxicity with EFV may lead to altered adherence to ARTs. Thus, it is important to identify early markers predicting the emergence of new resistance mutations in patients on NNRTI-based ART.\n\nThe phase 3 CARINEMO randomized clinical trial enrolled 570 HIV-TB coinfected patients in Mozambique, Africa, and compared the efficacy and safety of two NNRTIs (nevirapine [NVP] and EFV) for ART-naïve patients3. In the intent-to-treat population, 64.6% (95% confidence interval (CI): 58.7-70.1%) of patients who received NVP achieved virological suppression at week 48 (defined as HIV-1 RNA <50 copies per ml), compared with 69.8% (95% CI: 64.1-75.1%) of those who received EFV. The evolution of plasma concentrations of NVP and EFV during and after anti-TB therapy, as well as its association with toxicity and virological suppression, has been previously described4. The emergence of ARV-resistance mutations was observed during the trial and briefly described, but the relationship between NNRTI plasma concentrations and the emergence of resistance was not investigated. Here, we analyzed subgroup datasets from the CARINEMO trial, which provided a unique opportunity to explore the factors associated with viral replication and the emergence of resistance mutations while on ART. These data also offered the possibility to assess the relationship between viral replication, ARV plasma concentrations and the emergence of resistance mutations. The identification of risk factors of virological escape at week 48 in a well-characterized and homogeneous population is critical to prevent treatment failure in settings where the best timing for routine HIV-RNA still needs to be assessed.\n\n\nMethods\n\nThe CARINEMO trial (ClinicalTrials.gov identifier: NCT00495326) was conducted in three health centers located in Maputo, Mozambique, from 2007 to 2011; a full description of the trial is available from Bonnet et al.3. Participants were randomized to NVP or EFV (without lead-in dose) and received either a fixed-dose combination of NVP (400 mg/day), lamivudine and stavudine (Triomune®) or EFV (600 mg/day) plus lamivudine and stavudine started 4 weeks after anti-TB treatment initiation and for a duration of 48 weeks. In August 2010, stavudine was replaced by zidovudine. For TB, all patients received a fixed-dose combination of isoniazid (H), rifampicin (R), ethambutol (E) and pyrazinamide (Z) for 2 months, followed by 4 months of isoniazid/rifampin.\n\nFour ethics committees approved the study protocol: the Comite Nacional de Bio-Etica para a Saude (Maputo, Mozambique), the Medecins Sans Frontieres Ethics Review Board (Zurich, Switzerland), the Comite de Protection des Personnes (Saint Germain-en-Laye, France), and the Columbia University ethics review committee (New York, NY, USA). All participants provided signed informed consent.\n\nPlasma HIV-RNA levels were measured at inclusion and then at weeks 12, 24, 36 and 48 using the Roche Cobas Amplicor HIV-1 Monitor Test v1.5 (Roche Diagnostics, Basel, Switzerland) at the molecular biology laboratory of the Instituto Nacional de Saúde, Maputo, Mozambique. Resistance mutations to NRTI and NNRTI were determined in all patients with plasma HIV-1 RNA >400 copies/ml at week 48 by sequencing the reverse transcriptase gene using the consensus technique of the AC11 ANRS Resistance Group (www.hivfrenchresistance.org) at the Department of Virology, Necker Hospital (Paris, France). A patient was defined as having an emergence of resistance mutations at week 48 if at least one (N)NRTI resistance mutation was detected at any level.\n\nAdherence counseling on both ART and anti-TB therapies was provided by the study team at the clinics. At each follow-up visit, adherence to both ART and anti-TB treatment was monitored using an analog visual scale, standardized questionnaire administered by a nurse and pill counts. Adherence to ART was calculated for each time point using pills counts only. The number of returned doses during the last 3 months prior to weeks 12, 24, 36 and 48 were compared to the number of doses prescribed and refills. An indicator of compliance was defined by classifying adherence with a threshold of 95%.\n\nPre-dose concentrations of NVP and 12 h after the evening intake of EFV were measured at weeks 12, 24, 36 and 48. Patients for whom the measured concentrations were below the limit of quantification at each measurement were removed from the analysis, assuming the ART was not taken at all.\n\nVirological suppression was defined as an HIV-RNA below 400 copies/ml at week 48 and virological escape as an HIV-RNA above 400 copies/ml. Patients switched during follow-up to another ART regimen were excluded from the analysis. Percentiles (P) of drug concentrations were provided for each NNRTI at each time point. The P10, P25, P50, P75 and P95 were calculated and used to categorize drug concentrations. The P10 value was used to classify patients as having low drug concentrations (threshold below which 10% of drug concentrations were measured). Mean changes in HIV-1 RNA values after log transformation at each time point vs. baseline values were compared between patients with and without the emergence of resistance mutations by performing an analysis of covariance at each time point with HIV-1 RNA baseline values (log transformed) and treatment as covariates. Univariate and multivariate logistic regression models were fitted to assess the associations between virological escape and the emergence of resistance mutations at week 48 with drug concentrations at weeks 12 and 24, adherence to ART and other patient-associated factors, such as body mass index, sex, age, CD4 cell counts, as well as the HIV-1 RNA and ART regimen at treatment initiation. For both outcomes, factors associated with a P-value <0.20 in univariate analysis were selected for the initial multivariate analysis and a manual backward stepwise approach was used to obtain the final multivariate model. Only factors significantly associated (P<0.05) with the outcomes remained in the model and the importance of each in the final model was tested with a likelihood ratio test at the same level of significance (5%). The area under the receiver operating characteristics (ROC) curve was computed to assess the prediction of the low drug concentration on the risk of virological escape and the emergence of resistance mutation. The same analysis was repeated to evaluate the prediction of the HIV 1 RNA at weeks 12 and 24 on the risk of virological escape at week 48 and the emergence of resistance mutation. Other statistical comparisons were performed using the Chi-square test, Fisher’s exact test or Student’s t-test as appropriate. A P-value of ≤0.05 was considered statistically significant. Tests were performed with Stata 14 (StataCorp LP, College Station, TX).\n\n\nResults\n\nOf the 570 patients randomized in the CARINEMO trial, 470 had available measurement of HIV-RNA at week 48. Among these, 446 had at least one measure of detectable drug plasma concentrations without being switched during follow-up to another ART regimen (Figure 1). Demographic data and clinical characteristics at baseline and during the 48-week follow-up are summarized in Table 1. De-identified raw data for each patient is available on figshare5.\n\nAmong the 446 patients an adherence rate less than 95% was observed among 7 (1.6%) patients from enrolment up to week 12, in 11 (2.5%) patients from weeks 12 up to 24, in 11 (2.5%) patients between weeks 24 and 36, and 8 (1.8%) patients between weeks 36 and 48.\n\nAmong the 446 patients, 54 (12.1%) presented a virological escape; 48 patients (10.8%) had a genotype performed and 35 (7.8%) had at least one major resistance mutation detected on the reverse transcriptase gene. The decrease in HIV-1 RNA levels from baseline was significantly slower in patients in whom resistance mutations were identified at week 48 compared with those with no occurrence of resistance (Table 3).\n\nPercentile values for NVP and EFV drug concentrations at each time points are presented in Table 2. Values of P10, P25 and P50 at week 12 were 1253 ng/ml, 1784 ng/ml and 2786 ng/ml, respectively, for EFV, and 1893 ng/ml, 2996 ng/ml and 4095 ng/ml, respectively, for NVP. The distribution of drug concentrations using these percentile categories differed statistically between patients with virological suppression and those with virological escape at week 48. At week 12, 28.2% (11/39) of patients with a plasma concentration of the NNRTI-component within the P10 failed to suppress their viral load at week 48 compared with 10.1% (35/348) in those with higher concentrations (p=0.001) similar to week 24 (35% [14/40] vs. 9.2% [32/348], respectively; p<0.001). Among these patients, median concentrations were lower in those with virological escape compared to cases with virological suppression at week 12 for the NVP group and in both the EFV and NVP groups at week 24 (Figure 2, p=NS). The same differences were observed in the distribution of drug concentrations between patients with or without the emergence of resistance mutations. At week 12, 21.2% (8/37) of patients presenting plasma drug concentrations of the NNRTI component within the P10 had resistance mutations at week 48, compared with 5.8% (20/344) in those with higher concentrations (p<0.001), similar to week 24 (26.3% [10/38] vs. 5.2% [18/345], respectively).\n\nData given as mean (standard deviation).\n\nNS, not significant.\n\nMultivariate analyses showed that plasma drug concentrations below the P10 threshold at weeks 12 and 24 were independently associated with virological escape at week 48 (adjusted odds ratio [aOR]=2.9; 95% CI: 1.1 -7.2; p=0.0312 and aOR=4.2; 95% CI: 1.8-9.8; p=0.0019, respectively), as well as adherence below 95% at week 24 (aOR=10.5; 95% CI: 1.2-89.8; p=0.044, respectively) (Table 4). There was no influence of the choice of the NNRTI component or the CD4 cell count at baseline on factors associated with virological escape at week 48. Drug concentrations below the P10 threshold at weeks 12 and 24 (aOR=4.5; 95% CI: 1.8-14.6; p=0.009 at week 12; aOR=5.1; 95% CI: 1.8-14.6 at week 24) were also independently associated with an increased risk of emergence of resistance mutation as well as the ARV treatment received at initiation (aOR=3.2; 95% CI: 1.1-9.1; p=0.0244), for NVP vs. EFV. Adherence below 95% at week 24 was no longer shown to be associated at the significance level of 5% (p=0.0581) (Table 5).\n\nAmong the 345 patients with both mid-dose concentrations at weeks 12 and 24, the ROC analysis showed an area under the curve (AUC) at week 12 of 0.62 (95% CI: 0.52-0.72) and 0.67 (95% CI: 0.65-0.82) at week 24 for virological escape. An AUC of 0.76 (95% CI: 0.66 -0.87) and 0.75 (95% CI: 0.63-0.86) at weeks 12 and 24, respectively, was observed for the emergence of resistance mutations, thus indicating a better predictability of the mid-dose concentration on resistance mutations in contrast to virological escape. When using the HIV-1 RNA values at weeks 12 and 24 to predict the two outcomes, the ROC analysis showed AUCs of 0.69 (95% CI: 0.60-0.77) and 0.66 (95% CI: 0.55-0.76), respectively, for virological escape and 0.72 (95% CI: 0.63-0.80) and 0.75 (95% CI: 0.65-0.86), respectively, for the emergence of resistance mutations. These results indicate a better predictability of the HIV-1 RNA values on resistance mutations in contrast to virological escape.\n\n\nDiscussion\n\nIn the present study, we used the data of a large randomized clinical trial assessing two drugs of the NNRTI class in combination with anti-TB drugs. Our findings showed that very low drug plasma concentrations early after treatment initiation (week 12) were predictive factors of virological escape and the emergence of resistance mutations at week 48. Low drug concentrations may be explained by a suboptimal adherence or a potent drug interaction when patients receive other drugs such as rifampicin. Recently, it was suggested that the wave of ART treatment failure primarily affecting resource-limited countries should be considered as a fourth epidemic6. This epidemic, accompanied by the emergence of ARV drug resistance, could affect 3 to 5 million individuals between 2020 and 20307. Therefore, early predictors of ART failure are critically important. Until low-cost, simple assays for drug monitoring are available, pharmacological drug monitoring cannot be routinely recommended in low-resource settings to trigger drug resistance testing8,9. For this reason, we advocate for the development of easy-to-use point-of-care tests for anti-HIV drugs to help monitoring for adequate drug intake and therefore drug exposure during clinic visits. This would allow reducing unnecessary viral load measurements and viral genotype determination and could prevent unnecessary switches to costly and complex salvage ART in contexts where the preservation of future treatment lines is critical.\n\nAdherence is a complex non-steady phenomenon and there is no gold standard or universal tool at present to detect irregular adherence10–14. This is particularly true during the first months of treatment initiation in a given population for a given ART treatment, taking into account forgiveness of the combined three ARV drugs15. Our study confirms that an adherence rate below 95% is independently associated with an increased risk of virological escape and the emergence of drug resistance16. In the absence of an adequate tool, surveillance of plasma concentrations with a simple assay in a subset of randomly selected patients could be a strategy to monitor a given cohort of patients starting ART. Viral load testing and adequate adherence support from the very first weeks of treatment should also be implemented in these settings17. The World Health Organization recommends performing the first viral load testing after ARV initiation at 24 weeks. However, some field reports have observed an improvement in long-term virological suppression in patients undergoing 12-week viral load testing18. Newly-developed, point-of-care test assays will benefit low-resource settings and help to expand such viral load measurements monitoring17,19,20.\n\nOur results demonstrated that the early detection of low drug plasma levels of the NNRTI component of the treatment regimen was able to discriminate patients who will later develop a resistance mutation. We showed that low to very low drug concentrations (below P10) in the first months after starting ART were significantly associated with the emergence of later virological escape and drug-resistance mutations. We were surprised by the EFV concentration levels in our study, which triggered a signal for viral escape. Indeed, the P10 at week 12 was 1253 ng/ml for the EFV component, whereas concentrations below 1000 ng/ml were sufficient in earlier studies21 to predict treatment failure. Furthermore, the ENCORE1 study showed the efficacy of the 400 mg EFV daily dose, suggesting also that the efficacy cut-off might be lower than 1000 ng/ml22. We hypothesized that the high frequency of CYP2B6 genetic polymorphism in individuals of African descent may explain a population concentration distribution above that observed in Caucasian patients by Marzolini et al.21.\n\nThis study has some limitations. First, in the CARINEMO clinical trial, data were collected at fixed time points and HIV-1 RNA and plasma drug concentrations started to be measured for all patients at week 12. This limited the assessment of earlier effects on virological escape and the emergence of resistance mutations in the very first weeks of treatment initiation. Second, included patients may not be representative of larger coinfected TB/HIV populations. In particular, these patients were closely followed and received support to sustain adherence to the ART and TB drugs. However, the use of pill counts only to calculate the compliance rate may have overestimated adherence. This was shown earlier in other reports16,23 as observed by the proportion of patients with adherence below 95% and low drug concentrations, even though other factors such as drug genetic polymorphism may have influenced the drug concentrations. Third, the nucleoside analog (NRTI) backbone used in this study is no longer recommended (d4T/lamivudine or zidovudine/lamivudine) and the current use of a backbone such as tenofovir disoproxil fumarate or tenofovir alafenamide, with a longer intracellular half-life, may have changed these results. Although NVP is no longer a preferred first-line therapy and many countries have now transitioned to a dolutegravir-based regimen, we believe that our results remain relevant. Dolutegravir has a shorter half-life than NVP and EFV, and assessing early drug exposure is likely to be extremely critical. In addition, when combined with anti-TB drugs, the dose of dolutegravir needs to be doubled, which supports the use of EFV-based ARV in coinfected TB patients. Fourth, the co-administration of anti-TB drugs with ART may have altered drug concentrations24 as rifampicin is a known potent inducer of NNRTI metabolism, in particular for NVP-based ART25. However, although no treatment effect was shown in our findings on virological escape, we observed a significant treatment effect in the multivariate analyses on resistance mutations, similar to previous trials26. ARV concentrations measured 12 h post-dose were previously used to predict virological and resistance outcomes, and were significantly associated with both outcomes at week 4814,27,28, despite the high inter-individual variability. Finally, the analyses were performed post hoc and were not discussed at the time of the initial statistical analysis plan.\n\nIn summary, early monitoring of drug intake may prevent the occurrence of late virological escape and the selection of viral resistance mutations. Adherence measurement using solely pill counts does not allow for such a prediction. Indeed, higher concentrations of NNRTI were associated with better virological outcomes. In low-resource settings, implementing routine 12-week HIV-1 viral load and innovative adherence measurements might ensure long-term treatment success and reduce the possibility of the emergence of drug resistance mutations.\n\n\nData availability\n\nRaw data associated with this study, including basic demographic information and data on viral load, are available on figshare. DOI: https://doi.org/10.6084/m9.figshare.7655630.v15.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Grant information\n\nThis work was supported by UNITAID.\n\nThe funders had no role in experimental design, data analysis and interpretation, or the decision to submit the work for publication.\n\n\nAcknowledgments\n\nWe thank the medical, research team and patients from the trial sites: Centro de Saúde de Alto Maé, Hospital Geral de Mavalane, Hospital Geral de José Macamo in Maputo, Mozambique; the Instituto Nacional de Saúde; Médecins sans Frontières – Switzerland (MSF- CH), Mozambique; the International Care AIDS Program (ICAP) in Maputo and the Agence nationale de recherches sur le sida et les hépatites virales (ANRS) (for the funding of the trial).\n\n\nCARINEMO - ANRS 12146 clinical trial study group\n\nIlesh V. Jani MD PhD, Nádia Sitoe Bsc, Adolfo Vubil Bsc MSc, Maria Nhadzombo, Fernando Sitoe, Delário Nhumaio, Odete Bule (Instituto Nacional de Saúde, Mozambique); Rui Bastos MD and Elizabete Nunes MD (Hospital Central, Maputo, Mozambique); Paula Samo Gudo MD MPH (National Tuberculosis Control Program, Mozambique); Josué Lima MD and Mie Okamura (International Center for AIDS Care and Treatment Programs, Mozambique); Laura Ciaffi MD, Agnès Sobry MD, Mariano Lugli and Bruno Lab (Médecins Sans Frontières - Switzerland, Mozambique); Avertino Barreto MD (Mozambique National AIDS Service Organisation, Mozambique); Christophe Michon MD (Regional Hospital, Annecy, France); Alexandra Calmy MD PhD (Médecins Sans Frontières; Division of Infectious Diseases, Geneva University Hospital, Geneva, Switzerland); Alpha Diallo (ANRS pharmacovigilance unit, France); Christine Rouzioux PharmD PhD (Paris-Descartes University, EA3620, Sorbonne Paris Cite, APHP, Necker Hospital, Paris, France).\n\n\nReferences\n\nGupta RK, Jordan MR, Sultan BJ, et al.: Global trends in antiretroviral resistance in treatment-naive individuals with HIV after rollout of antiretroviral treatment in resource-limited settings: a global collaborative study and meta-regression analysis. Lancet. 2012; 380(9849): 1250–1258. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorld Health Organization: Consolidated Guidelines on the Use of Antiretroviral Drugs for Treating and Preventing HIV Infection: Recommendations for a Public Health Approach. World Heal Organ. 2016; 155. PubMed Abstract\n\nBonnet M, Bhatt N, Baudin E, et al.: Nevirapine versus efavirenz for patients co-infected with HIV and tuberculosis: a randomised non-inferiority trial. Lancet Infect Dis. 2013; 13(4): 303–312. PubMed Abstract | Publisher Full Text\n\nBhatt NB, Baudin E, Meggi B, et al.: Nevirapine or efavirenz for tuberculosis and HIV coinfected patients: exposure and virological failure relationship. J Antimicrob Chemother. 2015; 70(1): 225–232. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBaudin E: data_carinemo.csv. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.7655630.v1\n\nLaborde-Balen G, Taverne B, Ndour CT, et al.: The fourth HIV epidemic. Lancet Infect Dis. 2018; 18(4): 379–380. PubMed Abstract | Publisher Full Text\n\nEstill J, Ford N, Salazar-Vizcaya L, et al.: The need for second-line antiretroviral therapy in adults in sub-Saharan Africa up to 2030: a mathematical modelling study. Lancet HIV. 2016; 3(3): e132–e139. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLamorde M, Fillekes Q, Sigaloff K, et al.: Therapeutic drug monitoring of nevirapine in saliva in Uganda using high performance liquid chromatography and a low cost thin-layer chromatography technique. BMC Infect Dis. 2014; 14: 473. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGandhi M, Yang Q, Bacchetti P, et al.: Short communication: A low-cost method for analyzing nevirapine levels in hair as a marker of adherence in resource-limited settings. AIDS Res Hum Retroviruses. 2014; 30(1): 25–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarcellin F, Spire B, Carrieri MP, et al.: Assessing adherence to antiretroviral therapy in randomized HIV clinical trials: a review of currently used methods. Expert Rev Anti Infect Ther. 2013; 11(3): 239–250. PubMed Abstract | Publisher Full Text\n\nNachega JB, Hislop M, Dowdy DW, et al.: Adherence to nonnucleoside reverse transcriptase inhibitor-based HIV therapy and virologic outcomes. Ann Intern Med. 2007; 146(8): 564–573. PubMed Abstract | Publisher Full Text\n\nMekuria LA, Prins JM, Yalew AW, et al.: Which adherence measure - self-report, clinician recorded or pharmacy refill - is best able to predict detectable viral load in a public ART programme without routine plasma viral load monitoring? Trop Med Int Health. 2016; 21(7): 856–869. PubMed Abstract | Publisher Full Text\n\nKouanfack C, Laurent C, Peytavin G, et al.: Adherence to antiretroviral therapy assessed by drug level monitoring and self-report in cameroon. J Acquir Immune Defic Syndr. 2008; 48(2): 216–9. PubMed Abstract | Publisher Full Text\n\nOrrell C, Cohen K, Leisegang R, et al.: Comparison of six methods to estimate adherence in an ART-naïve cohort in a resource-poor setting: which best predicts virological and resistance outcomes? AIDS Res Ther. 2017; 14(1): 20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nParienti JJ, Bangsberg DR, Verdon R, et al.: Better adherence with once-daily antiretroviral regimens: a meta-analysis. Clin Infect Dis. 2009; 48(4): 484–488. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcMahon JH, Jordan MR, Kelley K, et al.: Pharmacy adherence measures to assess adherence to antiretroviral therapy: review of the literature and implications for treatment monitoring. Clin Infect Dis. 2011; 52(4): 493–506. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBonner K, Mezochow A, Roberts T, et al.: Viral load monitoring as a tool to reinforce adherence: a systematic review. J Acquir Immune Defic Syndr. 2013; 64(1): 74–78. PubMed Abstract | Publisher Full Text\n\nKerschberger B, Boulle AM, Kranzer K, et al.: Superior virologic and treatment outcomes when viral load is measured at 3 months compared to 6 months on antiretroviral therapy. J Int AIDS Soc. 2015; 18(1): 20092. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoberts T, Bygrave H, Fajardo E, et al.: Challenges and opportunities for the implementation of virological testing in resource-limited settings. J Int AIDS Soc. 2012; 15(2): 117324. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCalmy A, Ford N, Hirschel B, et al.: HIV viral load monitoring in resource-limited regions: optional or necessary? Clin Infect Dis. 2007; 44(1): 128–134. PubMed Abstract | Publisher Full Text\n\nMarzolini C, Telenti A, Decosterd LA, et al.: Efavirenz plasma levels can predict treatment failure and central nervous system side effects in HIV-1-infected patients. AIDS. 2001; 15(1): 71–75. PubMed Abstract | Publisher Full Text\n\nDickinson L, Amin J, Else L, et al.: Pharmacokinetic and Pharmacodynamic Comparison of Once-Daily Efavirenz (400 mg vs. 600 mg) in Treatment-Naïve HIV-Infected Patients: Results of the ENCORE1 Study. Clin Pharmacol Ther. 2015; 98(4): 406–416. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartin M, Del Cacho E, Codina C, et al.: Relationship between adherence level, type of the antiretroviral regimen, and plasma HIV type 1 RNA viral load: a prospective cohort study. AIDS Res Hum Retroviruses. 2008; 24(10): 1263–1268. PubMed Abstract | Publisher Full Text\n\nBoulle A, Van Cutsem G, Cohen K, et al.: Outcomes of nevirapine- and efavirenz-based antiretroviral therapy when coadministered with rifampicin-based antitubercular therapy. JAMA. 2008; 300(5): 530–539. PubMed Abstract | Publisher Full Text\n\nUttayamakul S, Likanonsakul S, Manosuthi W, et al.: Influence of Cyp2B6 Haplotype on Plasma Efavirenz and Nevirapine Levels When Co-Administered With Rifampicin in Hiv / Tb Co-Infected Thai Adults. 2010; 2551.\n\nVan den Berg-Wolf M, Hullsiek KH, Peng G, et al.: Virologic, immunologic, clinical, safety, and resistance outcomes from a long-term comparison of efavirenz-based versus nevirapine-based antiretroviral regimens as initial therapy in HIV-1-infected persons. HIV Clin Trials. 2008; 9(5): 324–336. PubMed Abstract | Publisher Full Text\n\nAlexander CS, Asselin JJ, Ting LS, et al.: Antiretroviral concentrations in untimed plasma samples predict therapy outcome in a population with advanced disease. J Infect Dis. 2003; 188(4): 541–548. PubMed Abstract | Publisher Full Text\n\nHarrigan PR, Hogg RS, Dong WW, et al.: Predictors of HIV drug-resistance mutations in a large antiretroviral-naive cohort initiating triple antiretroviral therapy. J Infect Dis. 2005; 191(3): 339–47. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "44251",
"date": "27 Mar 2019",
"name": "Conrad Muzoora",
"expertise": [
"Reviewer Expertise HIV and co-infections: Cryptococol Meningitis and Tuberculosis"
],
"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 sub-study of an already published major study (CARINEMO ANRS 12146) that initially evaluated the efficacy of Niverapine- and Efavirenz-based Antiretroviral therapy (ART) in patients on concomitant anti-tuberculous therapy.\nThis sub-study utilized viral load, resistance and drug level data to answer an important question about the relationship between low drug levels in the early phase of ART and week 48 virological escape/possible emergence of resistance.\nThis is a relatively understudied area and the authors provide very useful data that has otherwise been unavailable in published literature.\nThe manuscript is well written and easy to understand with sound statistics, straight forward results and justifiable conclusions.\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": "46788",
"date": "23 Apr 2019",
"name": "Gary Maartens",
"expertise": [
"Reviewer Expertise I have had experience researching ARV drug concentrations as objective adherence measures & other adherence measures"
],
"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 post hoc study from the CARINEMO trial, which shows the relationship between NNRTI concentrations & virologic failure & resistance, which is an important addition to the literature showing the value of ARV concentrations as objective adherence measures. In general, the article is well written.\nSpecific comments:\n1. Title: The title doesn't reflect what was actually done - I suggest something like: \"Efavirenz and nevirapine concentrations as predictors of virologic failure & resistance: a substudy...\"\n2. Abstract, results: in last sentence \"mid-dose concentration\" is used - this is true for EFV, but not NVP (it's a trough) - rather use C12 throughout. Mid-dose concentration is also incorrectly used elsewhere in the paper. Clarify that in ROC analyses VL was done at week 12 & 24.\n\n3. Introduction: 2nd paragraph, 1st sentence: delete hyphen in \"ART-included\"\n4. Methods:\nIt's clear from the CONSORT diagram that participants with undetectable NNRTI concentrations at all time points were excluded (incidentally the number is missing from the CONSORT diagram, please correct) - this exclusion criterion should be in the methods (I would personally not have excluded them, but no need to re-analyse). Clarify in adherence section that pill counts were used as the measure in analyses, not the other two measures. I object to the categorisation of adherence using the threshold of 95% as this is not evidence-based - an early study (Ann Intern Med. 2000;133:21-301) of mostly unboosted PIs suggested this threshold, but that was a small study (n=99) of ARVs that are no longer relevant - despite this the threshold remains inappropriately used by many researchers. Numerous studies (e.g. Ann Intern Med. 2007;146:564-5732) have shown a relatively smooth dose response relationship between NNRTI adherence & virologic outcomes. I suggest change the analysis plan based on the distribution of their pill count data (e.g. lowest quartile/tertile or the above/below median). The use of the term \"virologic escape\" is eccentric - this term is usually used to reflect detectable virus in a compartment (typically the CNS) despite undetectable plasma VL. I suggest use either virologic failure or categorise as suppressed/unsuppressed. It's unclear if undetectable NNRTI concentrations were included in the P10 group (they should have been). Also, please state how undetectable NNRTI concentrations were handled in the ROC analyses (usual to take mid-point between zero & the limit of quantification of the assay).\n5. Results:\n\nIt's unacceptable to use P = NS; the calculated value should be given. I found the statement below surprising given the findings of the main trial: \"There was no influence of the choice of the NNRTI component or the CD4 cell count at baseline on factors associated with virological escape at week 48.\" The aOR was 1.8 for virologic failure in those randomised to NVP - although 95% CIs did cross 1 & P was >0.05; this does not mean there is no effect - the statement should be modified.\n6. Discussion: Their statement \"Our study confirms that an adherence rate below 95% is independently associated with an increased risk of virologic escape...\" is not borne out by their data as they did not explore other thresholds.\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": []
}
] | 1
|
https://f1000research.com/articles/8-169
|
https://f1000research.com/articles/7-1649/v1
|
16 Oct 18
|
{
"type": "Antibody Validation Article",
"title": "Comparative study of commercially available and homemade anti-VAMP7 antibodies using CRISPR/Cas9-depleted HeLa cells and VAMP7 knockout mice",
"authors": [
"Agathe Verraes",
"Beatrice Cholley",
"Thierry Galli",
"Sebastien Nola",
"Agathe Verraes",
"Beatrice Cholley",
"Sebastien Nola"
],
"abstract": "VAMP7 (vesicle-associated membrane protein) belongs to the intracellular membrane fusion SNARE (Soluble N-ethylmaleimide-sensitive factor attachment protein receptors) protein family. In this study, we used CRISPR/Cas9 genome editing technology to generate VAMP7 knockout (KO) human HeLa cells and mouse KO brain extracts in order to test the specificity and the background of a set of commercially available and homemade anti-VAMP7 antibodies. We propose a simple profiling method to analyze western blotting and immunocytochemistry staining profiles and determine the extent of the antibodies’ specificity. Using this method, we were able to rank the performance of a set of available antibodies and further showed an optimized procedure for VAMP7 immunoprecipitation, which we validated using wild-type and KO mouse brain extracts.",
"keywords": [
"VAMP7",
"SNARE",
"monoclonal",
"polyclonal",
"CRISPR/Cas9",
"KO",
"immunoprecipitation"
],
"content": "Introduction\n\nIntracellular membrane fusion in the secretory and endocytic pathways relies on SNARE proteins (soluble N-ethylmaleimide-sensitive factor attachment protein receptors) for membrane fusion events. In order to allow apposition and fusion between two membranes, vesicular (v)- and target (t)-SNARE form a so-called trans-SNARE complex, or SNAREpin. VAMP7 (vesicle associated membrane protein 7) is a clostridial neurotoxin-insensitive v-SNARE that belongs to the “Longin” subfamily (as opposed to the short “Brevins”, like VAMPs 1-3): it encompasses an amino-terminal extension, the Longin domain, which acts as an auto-regulatory domain1. VAMP7 mediates Golgi-derived, late endosomal and lysosomal and autophagosomal related membrane fusion events2,3 and co-localizes to a large extent with the tetraspanin CD634,5. VAMP7 is involved in exocytosis in several cell types6, including neurons7, in neurotransmitter basal release8,9 and specific brain circuits and functions10. VAMP7 exocytosis was shown to be regulated by an integrin-, FAK-, and Src-dependent mechanism in developing neurons11 and its transport to the cell periphery by VARP, Rab21 and Kif57, while retrograde transport depends on LRRK112. In non-neuronal cells, VAMP7 secretory vesicles release compounds such as ATP13,14 and interleukin-124. In addition, VAMP7 regulates trafficking of membrane proteins, including the tetraspanin CD8215 and the cold-sensing channel TRPM816. VAMP7 plays an essential role in cell migration and invasion17–19. VAMP7 also contributes to the regulation of membrane composition of sphingolipids and GPI-anchored proteins20.\n\nAt present date, several antibodies against VAMP7 are commercially available. However, not all of them have been extensively characterized and many reported studies have been conducted using exogenous expression. This is too little information regarding the sensitivity of these antibodies, and may limit their use for super-resolution imaging or proximity ligation assay, which require detection of endogenous proteins.\n\nIn this survey, we took advantage of the genome editing CRISPR/Cas9 technology to generate VAMP7-knockout (KO) human HeLa cells. This genetically modified cell line allowed us to test the specificity and background of available commercially or homemade VAMP7 antibodies. Here we compared four mouse monoclonal and four rabbit polyclonal antibodies by western blotting and immunofluorescence using standard protocols. We analyzed the data using a simple profiling of both western blots and immunocytochemistry images to extract a specificity index. We further characterized the best ones by immunoprecipitation assays using VAMP7 constructs from different origins and wild type or VAMP7 knockout mouse tissues.\n\n\nMaterial and methods\n\nHeLa and Cos-7 cells (ATCC CCL-2 and CRL-1651, respectively) were maintained at 37°C and 5% CO2 in a humidified incubator, and grown in Dulbecco’s modified Eagle’s medium (DMEM), supplemented with 10% fetal calf serum (FCS), 100 units/ml penicillin, and 100 μg/ml streptomycin (Gibco, Thermo Fisher Scientific). Cells were regularly split using Trypsin-EDTA to maintain exponential growth. Transfection of cells was performed using Lipofectamine 2000 according to the manufacturer's instructions. All culture media reagents were from Thermo Fisher Scientific.\n\nInvalidation of VAMP7 was achieved using CRISPR RNA-CAS9 guide constructs based on a previously published protocol21. Briefly, sequences for sgRNA were preferentially chosen within the first exon region of VAMP7 genomic gene, with the help of the “RGEN Cas designer”22 web-based tool (http://rgenome.net/cas-designer/). To limit off-targets, an oligo sequence with ≤2 putative mismatches throughout the whole genome or an ‘out of frame’ score <66 were excluded. The sgRNA target sequences used are: 5’-caccgAACAGCAAAAAGAATCGCCA-3’ (forward) and 5’-aaacTGGCGATTCTTTTTGCTGTTc-3’ (reverse). Oligonucleotides (10 mM) were heated at 95°C for 5 min and annealed by ramping down the temperature from 95°C to 25°C at 5°C min-1. Annealed primers were ligated into pSpCas9(BB)-2A-Puro (PX459) vector (Addgene) using the BbsI sites. After validation by sequencing, the targeting constructs were transfected into HeLa cells following a previously described protocol. An empty pSpCas9(BB)-2A-Puro (PX459) vector was used to generate “control” cells. At 24 h post-transfection, cells were diluted 1/10 and transfected ones were selected by 1 µg/ml puromycin addition for 72 h. The selected populations were then seeded into a 96-well plate at 1 cell per well. Clones derived from single cells were amplified and screened for deficiency by immunoblotting.\n\nThe human and rat GFP-VAMP7 constructs are the same as those that have been described previously23. Plasmid containing mouse VAMP7 cDNA was a kind gift from Maurizio D’Esposito (IGCB, Naples, Italy). Mouse VAMP7 was amplified by PCR and cloned into pEGFP-C3 (Clontech) using HindIII / BamHI restriction sites.\n\nThe wild-type (WT) and VAMP7 knockout (KO) littermate male cohort was established at the Mouse Clinical Institute animal facility as previously described10. They had a mixed 129/Sv-C57BL/6 genetic background. They were weaned at 4 weeks and housed two to six per M.I.C.E. cage by sex and litter regardless of the genotype under standard conditions and maintained in a room with controlled temperature (21−22°C) under a 12 h light/dark cycle (light on at 7:00 A.M.), with food (standard chow diet, Safe D04) and water available ad libitum. All experiments were performed in accordance with the European Communities Council Directive regarding the care and use of animals for experimental procedures (2010/63/UE) and were approved by the local ethical committee (CEEA40-Comité d’Ethique Buffon). Mice were euthanized by cervical dislocation. All efforts were made to ameliorate the suffering of animals and to reduce their number per experiment (1 animal cortex per condition for immunoprecipitation experiment).\n\nThe cortex from WT and VAMP7 KO 8 weeks-old C57bl/6 mice were isolated according to a previously published protocol10. Cortex were dissociated by pipette trituration in 500 µl TSE (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM EDTA) supplemented with 1% Triton X-100 and cOmplete protease inhibitor tablets (Roche Applied Science) and volume was adjusted in order to get ~2 mg/ml protein final concentration (considering 10 mg of tissue is equivalent to ~1 mg final protein). Lysis was performed by incubation under agitation at 4°C for 30min. After clarification by 16.000 x g centrifugation for 30 min, protein concentration of the supernatant was estimated using Protein Assay Dye Reagent Concentrate (Bio-Rad Laboratories) and immunoprecipitation was carried out (see below).\n\nThe TG50 (for “Thierry Galli #50”) serum raised against VAMP7 was generated by Covalab (Villeurbanne, France; animal house registration number C21 464 04 EA) using immunization of New Zealand white rabbit with GST-VAMP7 (1-188 aa, as previously described for TG1124 and TG1825) and then purified by affinity chromatography. Briefly, serum was clarified and de-lipidated by high-speed centrifugation (70,000 rpm) and applied on a 6xHis-VAMP7 (coiled-coil 1-188 aa) covalently cross-linked HITrap-NHS column (GE Healthcare) overnight at 4°C using a peristaltic pump (0.3 ml/min). We washed the column with filtered and degassed PBS (20 ml, 1 ml/min). Specific antibodies were eluted using 200 mM Glycine/HCl pH 2.2, and collected in tubes containing neutralizing buffer (TBS 10X, 1X final). For each fraction, protein concentration was quantified by optical density (280 nm) measurement.\n\nReferences for all tested antibodies and reagents used for immunoblotting are listed in Table 1 and Table 2, respectively. Cells were washed in cold phosphate buffered saline (PBS) and lysed 20 min in TSE (50mM Tris-HCl, pH 8.0, 150mM NaCl, 1 mM EDTA) supplemented with 1% Triton X-100 and cOmplete protease inhibitor tablets (Roche Applied Science). Lysates were clarified by centrifugation 30 min at 16.000 x g, and protein concentration was estimated using Protein Assay Dye Reagent Concentrate (Bio-Rad Laboratories). Following heat denaturation at 95°C during 5 min, proteins were separated by 15% SDS-PAGE and transferred to a 0.45-µm nitrocellulose filter (Amersham, GE Healthcare) at 40 mA overnight. Membrane was blocked with 5% low-fat milk in TBS buffer for 20 min at room temperature and probed with indicated primary antibodies diluted in 5% skimmed milk in TBS-T, overnight at 4°C. Following primary antibody incubation, the filter was washed 3× 5 min in TBS-T at room temperature, and probed with HRP- or fluorescently-labeled secondary antibodies diluted in TBS-T for 1 h at room temperature. The filter was then washed 3× 5 min in TBS-T at room temperature. All incubation and washes were performed with gentle rocking. Proteins were detected by enhanced chemiluminescence (ThermoFisher Scientific) and imaged using ImageQuant LAS-4000 (Fujitsu Life Sciences), or scanned in an Odyssey infrared imaging system (Li-Cor).\n\nThe following immunofluorescence staining protocol was performed, with all steps carried out at room temperature unless stated otherwise (see Table 3 for reagent details). HeLa cells were washed once in PBS, fixed with 4% paraformaldehyde in PBS for 20 min, quenched for 20 min with 50 mM NH4Cl in PBS and permeabilized by treatment with 0.3% Triton X-100 in PBS for 4 min. After blocking with 10% FCS + 0.3% Triton in PBS for 30 min, cells were incubated overnight with the primary antibodies diluted in 3% FCS + 0.3% Triton in PBS at 4°C. After several washes with 3% FCS, 0.3% Triton in PBS, cells were then incubated with the secondary antibodies in 3% FCS, 0.3% Triton in PBS for 1 h, then washed several times in 0.3% Triton in PBS. Coverslips were partially dried and mounted in Prolong medium (Invitrogen), and then left to set overnight. Fluorescence microscopy and imaging were performed using an upright epifluorescence microscope (DMRA2, Leica Microsystem) equipped with a CMOS camera (Orca Flash 4.0 LT, Hamamatsu) and a HCX PL APO 100x/1.40-0.70 oil CS oil-immersion Leica objective. Dilution of primary antibodies (Table 3) was formerly optimized to get relatively equivalent signal intensity in the WT cells for the same microscope settings (time of exposure, binning, objective, etc.), allowing direct comparison of signal between antibodies.\n\nTransfected Cos7 cells were washed once in PBS 1X then lysed as described in the “immunoblotting” section. Immunoprecipitation experiments were carried out as followed (see Table 4 for reagent details). Briefly, 1 mg of protein extract was submitted to immunoprecipitation by incubation overnight at 4°C with 2.5 µg of antibodies that were pre-coupled with 25 μl magnetic beads (Dynabeads M-280, Invitrogen). Beads were then extensively washed with TSE-1% Triton and beads resuspended in 2X Laemmli buffer. Samples were loaded on 4–12% Bis-Tris NuPAGE (ThermoFisher Scientific) or RunBlue SDS (Expedeon) gels with manufacturer-recommended electrophoresis buffer, processed for western blotting using HRP-coupled secondary antibodies and enhanced chemiluminescence (ThermoFisher Scientific).\n\nFor immunoprecipitation of endogenous VAMP7, 1 mg of mouse cortex extracts (see “Cortex isolation” section) were submitted to immunoprecipitation as for cell extracts, excepted that 5 µg of antibodies, 40 μl magnetic beads (Dynabeads M-280, Invitrogen), fluorescent secondary antibodies and an Odyssey infrared imaging system (LI-COR, Lincoln, Nebraska, USA) were used.\n\nAll quantification analyses was performed using ImageJ (1.49n) and data were computed in Microsoft Excel.\n\nFor western blotting signal quantification, a 20-pixel-wide straight line vertically crossing each lane was drawn to generate an intensity profile (see Figure 1A and B, left panel). Local background correction was performed by manually drawing a segmented line under the peaks representing the bands detected by western blotting (Figure B, right panel). The VAMP7 band was defined as the ~25 kDa band that would disappear or diminish in intensity in the KO extract compared with the WT extract. Areas under all peaks and the VAMP7 one, shown in grey and blue, respectively (Figure 1B, right panel), were measured. In order to estimate the signal-to-noise ratio of each antibody, taking into account the intensity of the band of interest over the intra-lane non-specific ones and the specificity of the signal in the control condition versus the KO ones, a so-called “WB specificity index” was calculated for each lane of the western blot, using the following equation:\n\n\n\n(A) Immunoblot performed on lysates from control (Ctrl) and VAMP7 knockout (KO) HeLa cells. An equal amount of total protein extracts from each condition was run in replicates. Following transfer on nitrocellulose and blocking, membrane was sliced and each piece was probed with indicated anti-VAMP7 antibodies (expected size: ~25 kDa). Time of exposure for each condition is provided. For loading control, membrane was washed and incubated with anti-α-tubulin antibody (expected size: ~50 kDa). (B) Example of quantification of western blotting signal from (A) (see dotted red line). Intensity profile (right panel) was generated from a 20-pixel-wide straight line (yellow, left panel) across each lane. On the intensity profile (right panel), areas corresponding to the VAMP7 signal and the non-specific bands are shown in blue and gray, respectively (see Methods section for details). (C) Quantification of each antibody tested by western blotting shown in (A). The “specificity index” represents the signal-to-noise ratio of the antibodies, reflecting the intensity of the VAMP7 band amongst the overall signal per lane (including non-specific bands) and its specificity in the control condition compared to KO (see Methods section for details). AP, affinity purified.\n\nFor immunocytochemistry, rather than a raw intensity comparison between control and KO conditions, we chose to rely on the subcellular distribution of VAMP7 to reflect the correct staining pattern of the tested antibodies better. This was estimated using a previously described method with slight modifications26, for each antibody, in control and VAMP7 KO cells. A 20-pixel-wide band (Figure 2B) was drawn from the nucleus (defined from the edge of DAPI staining) to the cell leading edge and going through the perinuclear region pointing to the most elongated part of the cell. Length of this band was normalized by binning in order to obtain a 10-pixel-long line from which a fluorescence intensity profile was generated. Profile from control and KO cells were plotted in function of the distance from nucleus to periphery (i.e. 0–100% on the x axis, Figure 2B, lower panel). A so-called “IF specificity index” was calculated by measuring the area between the control and the KO intensity profiles, which can be computed by the following equation:\n\n\n\n(A) Control and VAMP7 knockout (KO) HeLa cells were grown on glass coverslips, fixed in paraformaldehyde, blocked and immunostained with the indicated mouse or rabbit anti-VAMP7 antibodies (green) and DAPI (blue). KO condition allows the estimation of background signal. Samples were imaged with an epifluorescence microscope using a 100X objective. Bars, 25 µm. (B) Example of quantification of immunofluorescence signal from (A). Distribution of VAMP7 fluorescence was measured across a 20-pixel-wide straight line (yellow) drawn from the border of the nucleus towards the plasma membrane in control and VAMP7 KO cells (left panel). Intensity profiles were plotted on the same graph (right panel) and area between control and VAMP7 KO curve (blue) was measured in order to generate a “specificity index” (see Methods section). (C) Quantification of each antibody tested by IF shown in (A). The “specificity index” reflects both pattern of distribution of VAMP7 across the cell and signal-to-noise ratio given by the tested antibodies (see Methods section for details). AP, affinity purified.\n\nAntibodies with a high score exhibit both a more intense signal in the perinuclear region than the peripheral one in the control and a low signal distribution for VAMP7 KO condition compared to control. Poor score corresponds to either a correct distribution of VAMP7 signal in the control but a very high background in the KO condition, or an unspecific/random distribution of VAMP7 signal in the control, or a combination of both cases.\n\n\nResults\n\nIn order to characterize and compare a set of commercially available and homemade (i.e from “Thierry Galli’s lab”, hereafter referred as “TG lab”) anti-VAMP7 antibodies (Table 1), we first generated VAMP7 knockout cells, using CRISPR/Cas9 engineering21, as described in the methods section. HeLa cells were chosen as they express VAMP7 endogenously in a detectable amount by western blotting and immunocytochemistry15.\n\nSeveral of the tested antibodies were described in the provider’s datasheet to only work for immunofluorescence detection (e.g. Cell Signaling, catalogue number D81Y1R) or western blotting (e.g. Cell Signaling, catalogue number D4D5J) and were used accordingly. Our lab generated two antibodies, the mouse monoclonal “158.2”27 and the rabbit polyclonal “TG50”, which are commercially available from Synaptic Systems and Covalab (TG50 as protein A purified serum), respectively. Only the in-house affinity-purified (AP) version of the TG50 antibody was included in this study because we wanted an affinity-purified serum as best possible positive control.\n\nWe compared four monoclonal and four polyclonal rabbit antibodies by western blotting using control or VAMP7-KO HeLa cell extracts (Figure 1). We used an anti-tubulin antibody as loading control. All antibodies tested in the described conditions (Table 2) were sensitive enough to detect a prominent band at the expected molecular weight (~25 kDa) in the control condition. This band was absent in the VAMP7-KO cell lines in all cases. However, some non-specific bands were visible for all the tested antibodies in both control and VAMP7-KO cell lysates, particularly with polyclonal Synaptic Systems (catalogue number 232 003) and Sigma-Aldrich (catalogue number T6074) antibodies. Therefore, all the tested antibodies showed a signal that was specific for VAMP7, but they also showed variable background bands. As assessed by intensity profile analysis (Figure 1B) and our western blotting specificity index (Figure 1C), the TG lab (TG50) antibody showed the best signal-to-noise ratio using this Western blotting conditions, which may not be surprising, because it had been affinity-purified.\n\nIn order to better characterize these antibodies (Table 1), we performed immunostaining in control and VAMP7 KO HeLa cells. For this assay, the rabbit antibody from Cell Signalling Technology clone D8Y1R was used instead of the D4D5J clone, according to the manufacturer’s recommendations. To compare the specificity of the antibodies, we adjusted their dilution (Table 3) to get relatively equivalent signal intensity in the WT cells with the same acquisition time on the microscope. According to this assay, the mouse antibodies from Creative Diagnostics (CABT-37960MH), Synaptic Systems (158.2 –232 011), TG lab (158.2) and the rabbit antibodies Synaptic Systems (232 003) and TG lab (TG50) stained perinuclear membrane structures and vesicles dispersed in the cytoplasm, a typical and already described localization pattern for VAMP7 in HeLa cells10,15,28. However, in the WT HeLa cells, the R&D Systems (MAB6117) antibody gave a homogenous signal, which spread into the nucleus, the Cell Signaling (14811) antibody, seemed to also stain perinuclear ER-like structures and the Sigma-Aldrich (T6074) antibody exhibited a diffuse cytoplasmic pattern with an absence of vesicular staining. All the tested antibodies seemed to show an overall lower-intensity signal in the VAMP7 KO cells compared to control. According to the “secondary-only” condition that reveals the internal background signal of the experiment (Figure 2A, right panel) and the distribution analysis we conducted (Figure 2B, see Methods for details) to compute the IF specificity index (Figure 2C), the most convincing signal-to-noise ratio was observed with the Creative Diagnostics (CABT-37960MH), Synaptic Systems (232 003) and TG lab (158.2 and TG50) antibodies.\n\nTaken together, immunoblot and immunofluorescence assays suggest that the homemade anti-VAMP7 antibodies (158.2 and TG50) showed the best endogenous signal-to-noise ratio in HeLa cells. To check the inter-species specificity of these antibodies, we carried out immunoprecipitation assays in Cos cells overexpressing mouse, rat or human GFP-tagged VAMP7 constructs (Figure 3A). We used 158.2 and TG50 for immunoprecipitation and immunoblot and species-specific IgG and anti-GFP antibodies as negative and positive controls for immunoprecipitation, respectively (Table 4). Both 158.2 and TG50 antibodies exhibited a sharp ~50 kDa band in all immunoprecipitation lanes, demonstrating a relatively equivalent ability to precipitate either mouse, rat or human VAMP7, while a very faint signal was observed in the negative control IgG IP lanes.\n\n(A) Cos cells overexpressing GFP-tagged mouse, rat or human VAMP7 constructs were lysed and VAMP7 was immunoprecipitated using indicated specific antibodies (“IP Antibody”). Normal isotype IgG (“IgGM” or “IgGR”) and GFP antibody were used as negative and positive control, respectively. Supernatants after immunoprecipitation (SN) and immunoprecipitates (IP) were probed with indicated VAMP7 antibodies (WB primary Ab). *VAMP7 band of interest (expected size of GFP-VAMP7 constructs: ~50 kDa). °Absence of band at VAMP7 corresponding size. (B) VAMP7 was immunoprecipitated from wild-type (“WT”) and VAMP7 knock-out (KO) mouse cortex lysates using rabbit anti-VAMP7 TG50 antibody and detected with mouse anti-VAMP7 158.2. Normal isotype IgG (IgM or IgGR) were used as negative control. Inputs (IN), supernatants after immunoprecipitation (SN) and immunoprecipitates (IP) were probed with mouse anti-VAMP7 158.2 antibody. *VAMP7 band of interest (expected size of GFP-VAMP7 constructs: ~50 kDa). °Absence of band at expected size. ~heavy and light chains of the antibody used for immunoprecipitation. AP = affinity purified.\n\nWe next wondered whether or not these antibodies could immunoprecipitate endogenous VAMP7 from tissue and test background signal in KO tissue. To this aim, we chose to perform immunoprecipitation on lysates from WT or VAMP7 KO mouse cortex using the rabbit TG50 antibody because it performed the best in previous assays and to observe immunoprecipitated VAMP7 with the mouse 158.2 antibody (Figure 3B). Although 158.2 antibody showed stronger background and multiple bands compared to the pattern seen in HeLa cells (Figure 1), a clear band at the expected molecular weight (~25 kDa) was present for the WT immunoprecipitation condition but not VAMP7 KO, demonstrating the specificity and low background noise of the TG50 antibody for immunoprecipitation of endogenous VAMP7 in mouse tissue extracts. Altogether, we showed here that both 158.2 and TG50 antibodies were able to immunoprecipitate VAMP7 from different species and that immunoprecipitation of endogenous VAMP7 could be performed in mouse tissue extracts using TG50 for IP and 158.2 for subsequent western blotting.\n\n\nDiscussion\n\nIn this antibody survey, we used genome-edited VAMP7 KO HeLa cells to compare several commercially available and homemade antibodies using standard methods for western blotting and immunocytochemistry. Apart from poor reactivity/quality, high background observed in western blotting or non-specific signal of some tested antibodies in IF could be due to non-fully optimized technical procedures. For example, different blocking agent could be used, such as bovine serum albumin for immunoblot or immunofluorescence assays, or cells could have been fixed differently (methanol, glutaraldehyde). We also cannot totally exclude some batch effect for the poor signal observed with some commercial antibodies. Furthermore, we voluntarily restricted this survey to the newly generated VAMP7-depleted HeLa cell line generated in this study and tissues from KO mice. We chose to evaluate the background signal of these antibodies in KO cells, criteria of choice that is both crucial for any assay and not always convincingly characterized. Conducting the same study in a different cell type or species might have led to different conclusions regarding background and specificity. Although it might have reduced the overall background, particularly in the VAMP7 KO immunofluorescence conditions, we chose to use epifluorescence microscopy rather than confocal imaging in order to give a global overview of the signal obtained with these antibodies and because many studies still largely rely on wide field microscopy. With these words of caution and limitations, we conclude that 158.2 and TG50 antibodies appeared as the best performers in our assays.\n\nThe datasheet provided with the Synaptic Systems (158.2 –232 011) antibody indicates that it is specific for rat and mouse. Here we provide evidence that it is also able to specifically recognize VAMP7 in human HeLa cells, both in immunoblot and immunofluorescence assays. This is in good agreement with the fact that human, rat and mouse VAMP7 protein sequences share more than 94% identity (alignment with www.uniprot.org website). MAb 158.2 also immunoprecipitated mouse, rat and human GFP-tagged VAMP7 (Figure 3A). Altogether, 158.2 thus appeared as a suitable antibody in all species and for all applications. However, this antibody may not be very sensitive, as it did not allow for the endogenous detection of low amounts of the protein, particularly in tissues (Figure 3). More generally, further comparative study should be conducted to formally assess the efficiency of this set of antibodies in non-human cell lines or tissues, particularly in immunocytochemistry. However, immunoprecipitation using TG50 (protein A purified serum available from Covalab, ref. pab01031-P) and 158.2 detection appeared as a valid strategy for specific isolation of the endogenous VAMP7.\n\nFinally, we proposed here an easy profile comparison of WT and KO western blotting and immunocytochemistry signals to rank the quality of antibodies directed against membrane-associated proteins as a decision-making tool for more complex studies.\n\n\nData availability\n\nDataset 1. Raw images of experimental replicates for Figure 1, immunoblotting experiments. This dataset includes uncropped blots for all experimental replicates that are represented in Figure 1. Treatments and immunoblot methods were performed as outlined in Figure 1. Blots were probed with indicated anti-VAMP7 antibodies and anti-α-tubulin antibodies was used as a loading control. (A) Dataset used for Figure 1, with cropped regions in red dashed line. (B) Additional set of raw images of a replicate experiment. Quantification as performed in Figure 1 is shown in lower panel. Note that although signal intensity and background are different within these two replicates, the relative performance of the different tested antibodies remained the same. DOI: https://doi.org/10.5256/f1000research.15707.d22136029.\n\nDataset 2. Raw images of additional experimental replicates for Figure 2, immunofluorescence experiments. This dataset includes additional images from experimental replicates of the images presented in Figure 2. Immunofluorescence staining methods and quantification (lower panel) were performed as described for Figure 2. Images were taken at 40× objective. Bar, 15µm. DOI: https://doi.org/10.5256/f1000research.15707.d22136130.\n\nDataset 3. Raw images of immunoprecipitation experiments for Figure 3, immunoprecipitation. Uncropped data from Figure 3 (A) and replicate (C) for VAMP7 immunoprecipitation from Cos-7 cell lysate overexpressing GFP-tagged mouse, rat or human VAMP7 constructs. Uncropped immunoblotting data from Figure 3 (B) and replicate (D) for VAMP7 immunoprecipitation from WT and VAMP7 KO mouse cortex extracts. Antibodies used for immunoprecipitation and subsequent immunoblotting are indicated. Red dashed lines show GFP-VAMP7 protein and cropped region, respectively. IN=Input (50 µg in A and C, 100 µg in B and D); SN = supernatant after immunoprecipitation; IP = immunoprecipitate; * = GFP-VAMP7; ° = Absence of band at GFP-VAMP7 size (~50 kDa); ~: immunoglobulins. DOI: https://doi.org/10.5256/f1000research.15707.d22136231.",
"appendix": "Grant information\n\nWork in our group was funded by grants from Association Française contre les Myopathies (Research Grant 16612), the French National Research Agency (NeuroImmunoSynapse ANR-13-BSV2-0018-02; MetDePaDi ANR-16-CE16-0012), the Ecole des Neurosciences de Paris (ENP), the Fondation pour la Recherche Médicale (FRM), Prix Coup d’Elan pour la recherche française of the Fondation Bettencourt Schueller, awards of the Association Robert Debré pour la Recherche Médicale to T.G.\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\nWe thank L. Danglot and C. Vannier for helpful discussion. We are grateful to Dr Maurizio D’Esposito (IGCB, Naples, Italy) for mouse VAMP7 plasmid.\n\n\nReferences\n\nDaste F, Galli T, Tareste D: Structure and function of longin SNAREs. J Cell Sci. 2015; 128(23): 4263–4272. PubMed Abstract | Publisher Full Text\n\nProux-Gillardeaux V, Rudge R, Galli T: The tetanus neurotoxin-sensitive and insensitive routes to and from the plasma membrane: fast and slow pathways? Traffic. 2005; 6(5): 366–373. PubMed Abstract | Publisher Full Text\n\nWang Y, Li L, Hou C, et al.: SNARE-mediated membrane fusion in autophagy. Semin Cell Dev Biol. 2016; 60: 97–104. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChiaruttini G, Piperno GM, Jouve M, et al.: The SNARE VAMP7 Regulates Exocytic Trafficking of Interleukin-12 in Dendritic Cells. Cell Rep. 2016; 14(11): 2624–2636. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCoco S, Raposo G, Martinez S, et al.: Subcellular localization of tetanus neurotoxin-insensitive vesicle-associated membrane protein (VAMP)/VAMP7 in neuronal cells: evidence for a novel membrane compartment. J Neurosci. 1999; 19(22): 9803–9812. PubMed Abstract | Publisher Full Text\n\nChaineau M, Danglot L, Galli T: Multiple roles of the vesicular-SNARE TI-VAMP in post-Golgi and endosomal trafficking. FEBS Lett. 2009; 583(23): 3817–3826. PubMed Abstract | Publisher Full Text\n\nBurgo A, Proux-Gillardeaux V, Sotirakis E, et al.: A molecular network for the transport of the TI-VAMP/VAMP7 vesicles from cell center to periphery. Dev Cell. 2012; 23(1): 166–180. PubMed Abstract | Publisher Full Text\n\nScheuber A, Rudge R, Danglot L, et al.: Loss of AP-3 function affects spontaneous and evoked release at hippocampal mossy fiber synapses. Proc Natl Acad Sci USA. 2006; 103(44): 16562–16567. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHua Z, Leal-Ortiz S, Foss SM, et al.: v-SNARE composition distinguishes synaptic vesicle pools. Neuron. 2011; 71(3): 474–487. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDanglot L, Zylbersztejn K, Petkovic M, et al.: Absence of TI-VAMP/Vamp7 leads to increased anxiety in mice. J Neurosci. 2012; 32(6): 1962–1968. PubMed Abstract | Publisher Full Text\n\nGupton SL, Gertler FB: Integrin signaling switches the cytoskeletal and exocytic machinery that drives neuritogenesis. Dev Cell. 2010; 18(5): 725–736. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang G, Nola S, Bovio S, et al.: Biomechanical Control of Lysosomal Secretion Via the VAMP7 Hub: A Tug-of-War between VARP and LRRK1. iScience. 2018; 4: 127–143. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVerderio C, Cagnoli C, Bergami M, et al.: TI-VAMP/VAMP7 is the SNARE of secretory lysosomes contributing to ATP secretion from astrocytes. Biol Cell. 2012; 104(4): 213–228. PubMed Abstract | Publisher Full Text\n\nFader CM, Aguilera MO, Colombo MI: ATP is released from autophagic vesicles to the extracellular space in a VAMP7-dependent manner. Autophagy. 2012; 8(12): 1741–1756. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDanglot L, Chaineau M, Dahan M, et al.: Role of TI-VAMP and CD82 in EGFR cell-surface dynamics and signaling. J Cell Sci. 2010; 123(Pt 5): 723–735. PubMed Abstract | Publisher Full Text\n\nGhosh D, Pinto S, Danglot L, et al.: VAMP7 regulates constitutive membrane incorporation of the cold-activated channel TRPM8. Nat Commun. 2016; 7: 10489. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWilliams KC, Coppolino MG: Phosphorylation of membrane type 1-matrix metalloproteinase (MT1-MMP) and its vesicle-associated membrane protein 7 (VAMP7)-dependent trafficking facilitate cell invasion and migration. J Biol Chem. 2011; 286(50): 43405–43416. PubMed Abstract | Publisher Full Text | Free Full Text\n\nProux-Gillardeaux V, Raposo G, Irinopoulou T, et al.: Expression of the Longin domain of TI-VAMP impairs lysosomal secretion and epithelial cell migration. Biol Cell. 2007; 99(5): 261–271. PubMed Abstract | Publisher Full Text\n\nSteffen A, Le Dez G, Poincloux R, et al.: MT1-MMP-dependent invasion is regulated by TI-VAMP/VAMP7. Curr Biol. 2008; 18(12): 926–931. PubMed Abstract | Publisher Full Text\n\nMolino D, Nola S, Lam SM, et al.: Role of tetanus neurotoxin insensitive vesicle-associated membrane protein in membrane domains transport and homeostasis. Cell Logist. 2015; 5(1): e1025182. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRan FA, Hsu PD, Wright J, et al.: Genome engineering using the CRISPR-Cas9 system. Nat Protoc. 2013; 8(11): 2281–2308. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPark J, Bae S, Kim JS: Cas-Designer: a web-based tool for choice of CRISPR-Cas9 target sites. Bioinformatics. 2015; 31(24): 4014–4016. PubMed Abstract | Publisher Full Text\n\nMartinez-Arca S, Alberts P, Zahraoui A, et al.: Role of tetanus neurotoxin insensitive vesicle-associated membrane protein (TI-VAMP) in vesicular transport mediating neurite outgrowth. J Cell Biol. 2000; 149(4): 889–900. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGalli T, Zahraoui A, Vaidyanathan VV, et al.: A novel tetanus neurotoxin-insensitive vesicle-associated membrane protein in SNARE complexes of the apical plasma membrane of epithelial cells. Mol Biol Cell. 1998; 9(6): 1437–1448. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartinez-Arca S, Proux-Gillardeaux V, Alberts P, et al.: Ectopic expression of syntaxin 1 in the ER redirects TI-VAMP- and cellubrevin-containing vesicles. J Cell Sci. 2003; 116(Pt 13): 2805–2816. PubMed Abstract | Publisher Full Text\n\nKuster A, Nola S, Dingli F, et al.: The Q-soluble N-Ethylmaleimide-sensitive Factor Attachment Protein Receptor (Q-SNARE) SNAP-47 Regulates Trafficking of Selected Vesicle-associated Membrane Proteins (VAMPs). J Biol Chem. 2015; 290(47): 28056–28069. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuzerelle A, Alberts P, Martinez-Arca S, et al.: Tetanus neurotoxin-insensitive vesicle-associated membrane protein localizes to a presynaptic membrane compartment in selected terminal subsets of the rat brain. Neuroscience. 2003; 122(1): 59–75. PubMed Abstract | Publisher Full Text\n\nAdvani RJ, Yang B, Prekeris R, et al.: VAMP-7 mediates vesicular transport from endosomes to lysosomes. J Cell Biol. 1999; 146(4): 765–776. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVerraes A, Cholley B, Galli T, et al.: Dataset 1 in: Comparative study of commercially available and homemade anti-VAMP7 antibodies using CRISPR/Cas9-depleted HeLa cells and VAMP7 knockout mice. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15707.d221360\n\nVerraes A, Cholley B, Galli T, et al.: Dataset 2 in: Comparative study of commercially available and homemade anti-VAMP7 antibodies using CRISPR/Cas9-depleted HeLa cells and VAMP7 knockout mice. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15707.d221361\n\nVerraes A, Cholley B, Galli T, et al.: Dataset 3 in: Comparative study of commercially available and homemade anti-VAMP7 antibodies using CRISPR/Cas9-depleted HeLa cells and VAMP7 knockout mice. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15707.d221362"
}
|
[
{
"id": "39566",
"date": "26 Oct 2018",
"name": "Alison H. Banham",
"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 study has compared the specificity of a panel of antibodies against VAMP7, a member of the SNARE protein family. Using CRISPR/Cas9 to create a HeLa-VAMP7 knockout cell line together with knockout mouse tissues the authors effectively demonstrate that the panel of antibodies are able to recognise the endogenous VAMP7 protein, the majority performing effectively in both WB and IF. Image analysis was performed to profile the signal-to-noise index in both techniques. While this does provide quantitative data with which to prioritise the antibodies, I reached the same conclusions without needing the profiling data. This approach may indeed have utility in other projects but perhaps consideration should be given to incorporating a measure for antibody sensitivity. Antibodies with similar specificity and different levels of sensitivity may have a similar specificity index score but may not be equally useful. Two antibodies were also validated for their ability to IP recombinant orthologous VAMP7 proteins and one for IP of the endogenous protein. This is a useful comparative study that will be of interest to other researchers sourcing antibodies to further characterise this molecule.\n\nComments:\n\n1) P3: For clarity I would suggest rephrasing ‘Invalidation’ of VAMP7 as inactivation, knockout or deletion.\n\n2) P5 and P9 bottom left column: Readers are referred to Table 3 for dilutions of primary antibodies yet this information seems to be provided in Table 1.\n\n3) While the image analysis provides quantitative information, the best antibodies identified using this approach were also easily identifiable by looking at the WB and IF images presented. With regard to the WB data in Figure 1, if just using the WB specificity index then there potentially appeared to be relatively little to choose between the two best antibodies. However, by eye, TG50 appears to have much greater sensitivity for VAMP7 detection, yet this distinction is not conveyed by a score which focuses on signal-to-noise.\n\n4) In Figure 2 – there is clearly significant heterogeneity in the distribution of VAMP7 both within and between individual HeLa cells. It is not clear from the methods whether data from individual cells or a large number were used to generate the IF specificity index and this information should be provided. The data would be more robust if multiple cells were sampled.\n\n5) Figure 2 indicates Cell Signaling 14811 performed poorly, however on the supplier’s site this antibody (D4D5J) is not recommended for IF and it is clearly stated by the authors in the results that an alternative (D8Y1R) was used. This might just be an error in labelling on the figure? However, the text describing these IF data also refers to the antibody being used as 14811 (which is D4D5J). This needs to be clarified and it would be helpful within the manuscript and figures to consistently refer to either the clone names or the catalogue number when describing an antibody.\n\n6) In the discussion the authors correctly comment that not using individually optimised conditions for each antibody in IF may disadvantage some of the reagents. I think it would have been more helpful to other researchers wanting to know which antibody works best for IF to be able to compare images obtained using an optimal staining protocol, such as that determined by each manufacturer.\n\n7) In the discussion the authors comment on the sensitivity of 158.2 being insufficient to detect low levels of the endogenous VAMP7 protein by WB. When TG50 seemed to be more sensitive in WB analysis then it would perhaps have been the more obvious choice for the WB detection antibody. This point could be added to the discussion.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nAre sufficient details of materials, methods and analysis provided to allow replication by others? Partly\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": "40504",
"date": "27 Nov 2018",
"name": "Marc G. Coppolino",
"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\nThis study assessed the performance of four monoclonal and four polyclonal antibodies to VAMP7. Six commercially available antibodies, and two antibodies generated by the researchers, were compared by Western blot, immunocytochemistry and immunoprecipitation. Lysates from HeLa and HeLa-VAMP7 knockout cells, and extracts from wildtype and VAMP7 knockout mouse brain tissue were used to examine specificity and background of the antibodies. The findings indicate some of the antibodies outperform comparators in blotting, immunocytochemistry, with reduced background and higher specificity, and perform well in immunoprecipitations. This work offers careful characterization of VAMP7 antibodies for use by other researchers in making decisions about antibody use for analytical studies of this protein.\nThe study is well conceived and executed. The approaches used are suitable, and the description of work is adequately detailed. Data are clearly presented, and for the most part conclusions are reasonable.\nThere are a few points that should be addressed to strengthen the study:\nFor the Western blot analyses, the antibody concentrations used (ug/ml) should be provided to allow more accurate comparison of antibody performance.\n\nIn the IF specificity index graph (Fig. 2C), it appears that the authors only measured the intensity profile in one cell for each antibody. These results would be more convincing if the intensity profiles of multiple cells were shown, as there can be variations in staining patterns that can arise from different cell morphologies, etc.\n\nIt is not clear why some of the antibodies, which performed reasonably well in Western blot and immunocytochemistry (Creative Diagnostics, Synaptic Systems), were not analysed in immunoprecipitations. An explanation for this, or information on their performance in this regard, would be useful.\n\nIs the work clearly and accurately presented and does it cite the current literature? Partly\n\nAre sufficient details of materials, methods and analysis provided to allow replication by others? 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/7-1649
|
https://f1000research.com/articles/8-165/v1
|
07 Feb 19
|
{
"type": "Systematic Review",
"title": "Therapy for acute basilar artery occlusion: a systematic review and meta-analysis",
"authors": [
"Kevin Sheng",
"Marcus Tong",
"Marcus Tong"
],
"abstract": "Purpose: This study aims to analyse the efficacy of different treatment methods for acute basilar artery occlusion, with an emphasis placed on evaluating the latest treatment methods. Method: A systematic review and meta-analysis was performed to analyse the current data on the therapies available for treating acute basilar artery occlusion. Results: A total of 102 articles were included. The weighted pooled rate of mortality was 43.16% (95% CI 38.35-48.03%) in the intravenous thrombolysis group, 45.56% (95% CI 39.88-51.28) in the intra-arterial thrombolysis group, and 31.40% (95% CI 28.31-34.56%) for the endovascular thrombectomy group. The weighted pooled rate of Modified Ranking Score (mRS) 0-2 at 3 months was 31.40 (95% CI 28.31-34.56%) in the IVT group, 28.29% (95% CI 23.16-33.69%) in the IAT group, and 35.22% (95% CI 32.39-38.09%) for the EVT group. Meta-analyses were also done for the secondary outcomes of recanalization and symptomatic haemorrhage. There was no difference between stent retriever and thrombo-aspiration thrombectomy on subgroup analysis in both clinical outcome and safety profile. Limitations: The included studies were observational in nature. There was significant heterogeneity in some of the outcomes. Conclusions: Superior outcomes and better recanalization rates for acute basilar occlusion were seen with patients managed with endovascular thrombectomy when compared with either intravenous and/or intraarterial thrombolysis. No superiority of stent‐retrievers over thrombo-aspiration thrombectomy was seen.",
"keywords": [
"basilar",
"occlusion",
"thrombolysis",
"thrombectomy",
"aspiration",
"stent retreiver",
"thromboaspiration",
"intraarterial"
],
"content": "Introduction\n\nStrokes caused by basilar artery occlusion are uncommon, with around 10% of large vessel strokes being basilar1. They are associated with very poor outcomes and high mortality; however, the condition can be heterogenous with variable prognosis2. There are a number of different pathophysiological mechanisms, including atherosclerosis, embolism, dissection and inflammation3. The presentation is variable, with patients presenting in different ways. In the preceding two to three weeks, many people may suffer from a prodrome that includes symptoms such as headache and vertigo4. On admission, patients may present with decreased GCS, dysarthria, focal weakness or cranial nerve dysfunction, ataxia and abnormal pupillary signs.\n\nEarly and successful recanalization has been shown to result in better outcomes in patients treated with intravenous thrombolysis and endovascular therapy5. While intravenous thrombolysis remains first line therapy for patients who present within the time-frame, previous studies show a low recanalization rate and poor clinical outcomes6,7. In strokes of the proximal anterior circulation, early mechanical thrombectomy has been shown through multiple randomised controlled trials to be superior to intravenous thrombolysis alone8–10. To date, there has only been a single small randomised clinical trial of 16 patients that compared intra-arterial urokinase versus control, however the study was prematurely stopped due to poor recruitment11. There are currently no other randomised trial data comparing different therapeutic approaches for basilar artery occlusion, with only observational data available.\n\nObservational data has suggested higher recanalization rates and better clinical outcomes for patients treated with endovascular thrombectomy by stent retrievers and thrombo-aspiration devices, and new techniques such as a direct-aspiration first-pass technique (ADAPT) and Combined Stent Retriever and Suction Thrombectomy (Solumbra technique) are promising. Hence it was decided to perform a systematic review and meta-analysis of the different therapeutic interventions for basilar artery occlusion, with a focus on comparing the safety profile and clinical performance of stent retriever vs thrombo-aspiration thrombectomy.\n\nData from the New England Medical Centre Posterior Circulation Registry shows that most of the patients with a basilar artery occlusion are between the ages of 50 and 802,3. Important risk factors for this disease include hypertension, diabetes mellitus, hypertension, hyperlipidaemia, smoking status, peripheral arterial disease and prior stroke. A large proportion of patients have been shown to have mini-strokes prior to the onset of basilar artery occlusion, with 58.6% of those with transient ischemic attack proceeding onto stroke.\n\nWhen managed conservatively, the literature presents a mixed but grim picture regarding prognosis. Most studies show poor outcomes of death and dependency in up to 95% of the study population12–14, however two observational studies show favourable outcomes in 71% (n=87) and 71% (n=61) respectively of patients managed conservatively3,15. This is likely reflective of numerous factors including study factors such as patient eligibility criteria and definition of outcomes. However, it also reflects the heterogenous nature of basilar artery occlusion, which has a variable severity and prognosis but can often be fatal2.\n\nVoetsch et al. analysed a group of 87 patients with moderate to severe basilar artery stenosis or occlusion from the New England Medical Centre Posterior Circulation Registry3. From this study, there were several factors that were statistically significantly associated with worse outcome. These included most significantly distal territory involvement and occlusion secondary to embolism, which likely reflects the lack of time for a collateral blood supply to form. Clinical predictors of worse clinical outcomes included decreased GCS on presentation, tetraplegia and pupillary signs. These predictors are supported by analysis of patients from the Lausanne Registry, which shows decreased GCS was the most important clinical predictor of poor outcome16.\n\n\nMethod\n\nPubMed was searched (through to August 2018) to identify pertinent research articles with the keyword “basilar”. The search strategy was deliberately kept simple in order to prevent the omission of large numbers of studies that may have occurred with a more restrictive search strategy.\n\nThe only limitations applied included article types (clinical study, clinical trial, comparative study, controlled clinical trial, dataset, evaluation studies, journal article, multicentre study, observational study, randomised controlled trial, twin study, validation studies), species (humans) and ages (Adult: 19+ years). Titles and abstracts were reviewed, with articles accessed in their entirety if they were potentially appropriate. Reference lists of research articles were further parsed for additional potentially appropriate studies. Research articles were read in their entirety if a decision on study inclusion could not be determined by reading the abstract. Attempts were made to communicate with the corresponding authors if further information was required.\n\nAll studies regardless of project design (including retrospective and prospective) were permitted. The limitation to this was that studies needed to have at a minimum of 10 patients, with studies reporting less than this being excluded. Studies were included if they included treatment for acute basilar artery occlusion. Observational and interventional studies covering intravenous thrombolytic therapy (IVT), intra-arterial thrombolytic therapy (IAT), plus/minus endovascular therapy (EVT) were allowed.\n\nOther exclusion criteria not already mentioned included the following: full text unavailable, duplicate studies, intervention other than IVT/IAT/EVT, lack of information regarding primary and secondary study outcomes (mortality, MRS 0-2, symptomatic intracranial haemorrhage, recanalization).\n\nData was independently parsed into a standardised table on Microsoft Excel 2013 by the two authors (KS and MT). Relevant data abstracted included method of therapy (IVT/IAT/EVT), clinical outcome at 3 months (Modified Rankin Score, Barthel Index, other), study population, baseline NIHSS, age, mortality, intracranial haemorrhage, study design, recanalization status, country, data collection window, time to first groin puncture, complications (dissection, perforation, embolization to new territory), adjunctive therapy. Data was combined into a single table once this process was complete.\n\nPrimary outcomes were mortality and good clinical outcome (mRS 0-2). Secondary outcomes were symptomatic intracranial haemorrhage (SICH) and recanalization. For the analysis comparing the stent retriever and thrombo-aspiration subgroups, the additional outcomes of dissection/perforation and embolization to new territory were parsed.\n\nMortality was assessed at 90 days, however if this was not available, then the nearest value was imputed as a surrogate. Clinical outcome was determined to be good if the mRS score was 0-2. If another definition was provided by the study and there was no other information to calculate the mRS score, then that particular definition was used. Clinical outcome was ideally determined at 3 months, but if this was unavailable or not determined, then the nearest value was imputed as a surrogate.\n\nSICH was defined as any haemorrhage associated with a worsening of the NIHSS score by ≥4 within 24 h, in accordance with the ECASS-II definition17. If another definition was provided by the study, then that particular definition was used. Recanalization was defined as TICI 2b/3, mTICI 2b/3, TIMI 2/3, or as per the study definition.\n\nDifferences were resolved through discussion and consensus of the two reviewers. Quality of studies was determined for each paper according to the reporting checklist proposed by the Meta-analysis of Observational Studies in Epidemiology (MOOSE) group.\n\nData synthesis. The DerSimonian and Laird random effects model was applied to perform the meta-analysis. If there was no significant heterogeneity, then the fixed effect model was used. In order to generate standard errors, the Freeman-Tukey double arc-sine transformation was applied to data extracted from individual studies. These are then back-transformed in order to form mean weighted probabilities with 95% confidence intervals.\n\nAssessment of heterogeneity and publication bias. Heterogeneity was evaluated through Cochran’s χ2 test (Cochran Q test), tau-squared and Higgin’s I-square statistic. A p-value of less than 0.05 and I-square >50% was regarded as significant. Publication bias was assessed through a variety of methods. These included the Begg and Mazumdar’s rank correlation test and Egger’s linear regression test (with a significance of p<0.05). In addition, funnel plots were generated, and trim and fill plot analysis was also conducted to adjust for any significant publication risk and publication bias adjusted weighted pooled rates are calculated.\n\nSensitivity analysis. Subgroup analyses and meta-regression were performed to determine potential sources of heterogeneity. The subgroups included intra-arterial thrombolysis (with angioplasty +/- stenting), intra-arterial thrombolysis (without angioplasty +/- stenting), stent retriever thrombectomy and thrombo-aspiration thrombectomy. For the stent retriever and thrombo-aspiration subgroups, studies were only included if data could be parsed specifically for outcomes relating to each. Meta-regression was done to compare outcome against publication year and time to first puncture. Comparison of subgroups was undertaken using the z-test of interaction. Exclusion sensitivity analysis was also performed.\n\nSoftware. All analyses and calculations in this meta-analysis were performed using the Mix V2.0 Pro statistical package.\n\n\nResults\n\nA total of 4994 articles were identified from PubMed. After evaluating the titles and abstracts of these articles, 148 remained eligible for assessment. The full texts of these articles were assessed, and 102 articles fulfilled the inclusion criteria. A PRISMA flow diagram has been included (Figure 1).\n\nSee Table 1 and Table 2 and Figure 2a–c and Figure 3.\n\nIVT, Intravenous thrombolysis. IAT, Intraarterial thrombolysis. EVT, Endovascular thrombolysis. SR, Stent retriever. TA, Thromboaspiration. MWP, mean weighted probability. CI, confidence interval. Q, Cochran’s Q. AP, angioplasty.\n\nIVT, Intravenous thrombolysis. IAT, Intraarterial thrombolysis. EVT, Endovascular thrombolysis. SR, Stent retriever. TA, Thromboaspiration. MWP, mean weighted probability. CI, confidence interval. Q, Cochran’s Q. AP, angioplasty.\n\nMeta-analysis for mortality in the (a) intravenous thrombolysis subgroup; (b) intra-arterial thrombolysis subgroup; (c) endovascular thrombectomy subgroup.\n\nMeta-analysis for good outcome in the (a) intravenous thrombolysis subgroup; (b) intraarterial thrombolysis subgroup; (c) endovascular thrombectomy subgroup.\n\nSee Table 3 and Table 4.\n\nIVT, Intravenous thrombolysis. IAT, Intraarterial thrombolysis. EVT, Endovascular thrombolysis. SR, Stent retriever. TA, Thromboaspiration. MWP, mean weighted probability. CI, confidence interval. Q, Cochran’s Q. AP, angioplasty.\n\nIVT, Intravenous thrombolysis. IAT, Intraarterial thrombolysis. EVT, Endovascular thrombolysis. SR, Stent retriever. TA, Thromboaspiration. MWP, mean weighted probability. CI, confidence interval. Q, Cochran’s Q. AP, angioplasty.\n\nSee Table 5 and Table 6. The z test of interaction between the weighted pool rate for mortality in the EVT group versus the IAT and IVT was statistically significant, indicating that the treatment effect likely differs between the groups.\n\nIVT, Intravenous thrombolysis. IAT, Intraarterial thrombolysis. EVT, Endovascular thrombolysis.\n\nIVT, Intravenous thrombolysis. IAT, Intraarterial thrombolysis. EVT, Endovascular thrombolysis.\n\nThe z test of interaction between the weighted pool rate for good clinical outcome in the EVT group versus IVT was not statistically significant but was statistically significant compared to the IAT group.\n\nSee Table 7. The z test of interaction between the weighted pool rates for the above clinical outcomes in the stent retriever (SR) versus thromboaspiration (TA) groups was not statistically significant, indicating that the treatment effect likely did not differ between the groups.\n\nEVT, Endovascular thrombolysis. SR, Stent retriever. TA, Thromboaspiration. MWP, mean weighted probability. CI, confidence interval. Q, Cochran’s Q.\n\nSee Table 8 and Table 9.\n\nEstimate, slope co-efficient. se, standard error. ci-/ci+, confidence interval. z, z value.\n\nEstimate, slope co-efficient. se, standard error. ci-/ci+, confidence interval. z, z value.\n\nMeta-regression demonstrated a statistically significant correlation towards decreased mortality across time with IVT, IAT and EVT therapy. There was a statistically significant correlation towards better clinical outcome at 3 months across time with EVT therapy, but not IVT or IAT. There was a statistically significant correlation towards increased recanalization across time with IAT and EVT therapy, but not IVT therapy. There was a statistically significant positive correlation towards decreased SICH across time with EVT therapy, but not IVT or IAT therapy.\n\nMeta-regression demonstrated a statistically significant correlation towards increased mortality across time to groin puncture with IAT and EVT therapy. There was a statistically significant correlation towards lower rates of good clinical outcome at 3 months across time with EVT therapy, but not IAT therapy. There was no statistically significant correlation towards decreased recanalization across time to puncture.\n\nSee Table 1–Table 4 and Table 7 for heterogeneity statistics and publication bias adjusted weighted pooled rates.\n\nThis systematic review and meta-analysis shows better mortality, good clinical outcome and recanalization rates for acute basilar occlusion patients managed with endovascular thrombectomy when compared with either intravenous and/or intraarterial thrombolysis. Further subgroup analysis revealed no significant difference between the use of stent retriever and aspiration thrombectomy.\n\nThere appears to be no benefit to IV or IA thrombolysis, either alone or in combination, as sole therapy for the management of acute basilar artery occlusion. The results of this meta-analysis show that endovascular thrombectomy is a superior approach to intra-arterial and intravenous thrombolysis, affording statistically significantly lower rates of mortality and symptomatic intracranial haemorrhage, and higher rates of recanalization. The result was also statistically significant for good clinical outcome for endovascular thrombectomy versus intra-arterial thrombolysis, however not versus intravenous thrombolysis. Intra-arterial thrombolysis was not statistically significantly superior to intravenous thrombolysis for both primary and secondary outcomes. However, there are confounding factors that may explain this apparent lack of difference in treatment effect.\n\nThese include stroke severity and time to treatment after symptom onset. The BASICS trial18 as one of the studies that held the most weight in the synthesis of outcomes for intravenous thrombolysis. When compared with intra-arterial therapy group, the initial stroke severity was less when comparison is made with baseline NIHSS (21 vs 25). Additionally, time to therapy was also better, with 81% in the IVT group treated in the first 6 hours compared with 64% in the intra-arterial therapy group. This is more marked for the 0-3-hour interval (55% vs 23%).\n\nAnother significant factor is inclusion criteria. Whilst diagnosis of acute basilar artery occlusion and assessment of recanalization is intrinsic to intra-arterial therapy, this is not the case for the IVT group. Most of the patients in the BASICS trial in the IVT group were diagnosed without angiography, and instead with non-invasive modalities such as CTA and MRA. In the study by Lindsberg et al.6, the vast majority of patients were included on the basis of MRA (TOF) as opposed to DSA, with MRA (TOF) used to assess recanalization in the days (median 1 day, IQR 1-3) following thrombolysis. This is significant because of potential false positives that may bias the outcome of the study. Other potential confounding factors include thrombus, volume, location and length, and presence of collateral circulation2.\n\nThe chief aim in the management of patients with acute basilar artery occlusion is the achievement of early recanalization. A meta-analysis by Kumar et al. demonstrates that recanalization is associated with a two-fold decrease in death rate (number needed to treat - 2.5) and a 1.5-fold decreased in futile outcome rate (NNT 3)5. However there remains a significant difference between the rate of recanalization and achievement of a good clinical outcome, which is likely due to differences in baseline admission factors.\n\nA meta-regression was performed to assess whether the outcomes of mortality and good clinical outcome systematically varied with time to first groin puncture in the EVT and IAT groups. For the outcome of mortality, there was a statistically significant association for both groups, and for good clinical outcome, there was a statistically significant association for the EVT group. This supports results from the BASICS trial18 and Eckert et al.19 which show that the rate of poor outcome was increased when time to recanalization therapy increased, with a significantly higher rate of poor outcomes if the time-period was greater than 6 hours from symptom onset. However, numerous studies in both the IAT and EVT groups find no consistent statistically significant association between time to treatment, and mortality and favourable clinical outcome20–22. This is likely due to other factors such as collateral flow from posterior communicating arteries and baseline ischemia/infarction.\n\nBaseline ischemia/infarction is another potential factor that affects mortality and clinical outcome. There are two main classification schemes presently, the posterior circulation Acute Stroke Prognosis Early CT Score (pc-ASPECTS) and the brain stem DWI score. Strbian et al. showed in a large prospective observation study that the absence of extensive brain infarction is associated with good clinical outcome as measured by the pc-ASPECTS score23. mRS 2-0 at 3 months was seen in 50% of patients with a pc-ASPECTS >= 8 and TIMI 2-3, as opposed to 5.9% in those with pc-ASPECTS <8. Recanalization of up to 48 hours after onset of symptoms was also shown to be beneficial in patients who did not have extensive baseline infarction. Cho et al. likewise showed in a cohort of 29 patients treated with intra-arterial therapy that only the brainstem DWI score was associated with futile outcome on both univariate and multivariate analysis24. The implication of this is that a holistic approach is required to determine the optimum treatment approach rather than arbitrary windows of treatment.\n\nOther factors also play a role, and these include age, baseline NIHSS, whether ventilatory support was required, atrial fibrillation, embolic origin and previous stroke25.\n\nThere was significant heterogeneity in the syntheses for the primary outcomes of both the IAT and EVT groups. Sensitivity analyses and meta regression performed helps to explain the heterogeneity found. Meta-regression for the EVT group showed a statistically significant association between the study characteristics of year of publication and time to puncture, and the primary outcome of mortality and mRS 0-2. Meta-regression for the IAT group showed a statistically significant association between the study characteristics of year of publication and time to puncture, and mortality but not mRS 0-2.\n\nIAT studies were very variable in design and ranged across a large span of time from 1986 to 2016 that encompassed significant clinical and technological improvements. Subgroup analysis based on treatment modality showed that the introduction of adjunctive angioplasty +/- stenting resulted in decreased mortality (49.79% vs. 36.55%) and increased rates of good clinical outcomes, (26.48% vs 31.50%), however this was not statistically significant on the z test of interaction (p>0.05), and significant unexplained heterogeneity remains in the synthesis for the primary outcomes.\n\nEVT studies were similarly variable in design with a very heterogenous collection of approaches and devices used. Insignificant heterogeneity was seen for the stent retriever subgroup for the outcomes of mortality, good clinical outcome and SICH and for the thrombo-aspiration subgroup for the outcomes of good clinical outcome, recanalization and SICH. Subgroup analysis however based on treatment modality (SR vs TA) was not statistically significant for any outcome.\n\nOther sources of heterogeneity have the potential to influence the above analyses and these include differences in terms of patient population and size, inclusion and exclusion criteria, definitions, use of adjunctive therapy, follow-up protocol, country of study, and management of relevant physiological parameters such as blood glucose level and blood pressure.\n\nThere have been significant advances in the mechanical devices used for endovascular thrombectomy, which have resulted in better mortality, good clinical outcome and mortality rates. Broadly, there are two main types of endovascular thrombectomy, the first being stent retriever thrombectomy and the second being aspiration thrombectomy, with combination therapy (Solumbra technique) or switching therapy (e.g. ADAPT) becoming increasingly used.\n\nWeighted pooled analysis of the stent retriever and aspiration thrombectomy subgroups showed no statistically significant difference in the primary and secondary outcomes of mortality, mRS 0-2, recanalization, SICH. There was also no significant different in the safety profile when comparing the outcomes of dissection/perforation and embolization to new territory. However, there was a large imbalance in the number of studies, with significantly fewer analysing the use of aspiration thrombectomy in acute basilar artery occlusion. Recent observational studies by Kang et al.26 and Gory et al.27 demonstrated similar treatment outcomes between patients who received either stent retriever or aspiration thrombectomy as first-line therapy. However, in the study by Gory et al., thrombo-aspiration was superior in terms of achieving complete perfusion as defined by mTICI 3and shorter length of treatment (0.543 vs 0.315 and 45 vs 56min, p<=0.05 for both)27. This was also seen in studies by Son et al.28 and Gerber et al.29, but not supported in studies by Kang et al.26 and Mokin et al.30.\n\nThe use of stent retrievers including Solitaire been shown to be superior compared to the earlier retrievers and thrombectomy systems in anterior circulation LVO stroke9. The randomised parallel-group SWIFT trial that compared the Solitaire with the Merci device showed better recanalization, mortality and good clinical outcome rates for the stent retriever device. However, there is only one study by Lutsep et al. (n=27) that provides data relating to Merci patients alone, and not in a pooled cohort with other endovascular modalities31. Mortality rate was shown to be 44%, with a recanalization and good clinical outcome rate of 78% and 41%, respectively.\n\nThere are numerous limitations to this systematic review and meta-analysis. The main weakness was that data was pooled together from prospective and retrospective observational studies, some including studies with only 10 patients. To date, there has been only one small study with randomised trial data that has looked at therapy pertaining to acute basilar artery occlusion. That particular study only had eight people in each arm (intra-arterial urokinase vs control). In addition, some papers looked more generally at vertebrobasilar or anterior circulation strokes, and in this case, data was separately extracted for outcomes relation to acute basilar artery occlusion or were excluded if this was not possible.\n\nThere were also significant differences in terms of patient population and size, inclusion and exclusion criteria, definitions, use of adjunctive therapy, follow-up protocol and time to treatment across the studies. This has resulted in significant heterogeneity in the syntheses for the primary outcomes of both the IAT and EVT group. Sensitivity analyses and meta-regression were conducted to explore this; however, for some analyses, significant unexplained heterogeneity remains. Hence these limitations should be taken into consideration when considering the results of this review.\n\nIn conclusion, the above data supports superior outcomes and better recanalization rates for acute basilar occlusion patients managed with endovascular thrombectomy when compared with either intravenous and/or intraarterial thrombolysis. Further subgroup analysis shows at this stage, there is no significant difference between the use of stent retriever and aspiration thrombectomy both in terms of their efficacy or safety profile. More systematic data is required, preferably randomised clinical trials, to determine the optimal approach to this potentially devastating disease.\n\nFull reference list for studies included in meta-analysis are available: Open Science Framework: Therapy for acute basilar artery occlusion: a systematic review and meta-analysis, http://doi.org/10.17605/OSF.IO/4A27M32\n\nSupplementary material including forest plots, funnel plots, exclusion sensitivity plots, meta-regression scatter plots and heterogeneity data for primary and secondary outcomes are available: http://doi.org/10.17605/OSF.IO/4A27M32\n\nPRISMA checklist: http://doi.org/10.17605/OSF.IO/4A27M32",
"appendix": "Grant information\n\nThe authors declare that no grants were involved in supporting this work.\n\n\nReferences\n\nSmith WS: Intra-arterial thrombolytic therapy for acute basilar occlusion: pro. Stroke. 2007; 38(2 Suppl): 701–3. PubMed Abstract | Publisher Full Text\n\nMortimer AM, Bradley M, Renowden SA: Endovascular therapy for acute basilar artery occlusion: a review of the literature. J Neurointerv Surg. 2012; 4(4): 266–73. PubMed Abstract | Publisher Full Text\n\nVoetsch B, DeWitt LD, Pessin MS, et al.: Basilar artery occlusive disease in the New England Medical Center Posterior Circulation Registry. Arch Neurol. 2004; 61(4): 496–504. PubMed Abstract | Publisher Full Text\n\nvon Campe G, Regli F, Bogousslavsky J: Heralding manifestations of basilar artery occlusion with lethal or severe stroke. J Neurol Neurosurg Psychiatry. 2003; 74(12): 1621–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKumar G, Shahripour RB, Alexandrov AV: Recanalization of acute basilar artery occlusion improves outcomes: a meta-analysis. J Neurointerv Surg. 2015; 7(12): 868–74. PubMed Abstract | Publisher Full Text\n\nLindsberg PJ, Mattle HP: Therapy of basilar artery occlusion: a systematic analysis comparing intra-arterial and intravenous thrombolysis. Stroke. 2006; 37(3): 922–8. PubMed Abstract | Publisher Full Text\n\nBhatia R, Shobha N, Menon BK, et al.: Combined full-dose IV and endovascular thrombolysis in acute ischaemic stroke. Int J Stroke. 2014; 9(8): 974–9. PubMed Abstract | Publisher Full Text\n\nBerkhemer OA, Fransen PS, Beumer D, et al.: A randomized trial of intraarterial treatment for acute ischemic stroke. N Engl J Med. 2015; 372(1): 11–20. PubMed Abstract | Publisher Full Text\n\nCampbell BC, Mitchell PJ, Kleinig TJ, et al.: Endovascular therapy for ischemic stroke with perfusion-imaging selection. N Engl J Med. 2015; 372(11): 1009–18. PubMed Abstract | Publisher Full Text\n\nGoyal M, Demchuk AM, Menon BK, et al.: Randomized assessment of rapid endovascular treatment of ischemic stroke. N Engl J Med. 2015; 372(11): 1019–30. PubMed Abstract | Publisher Full Text\n\nMacleod MR, Davis SM, Mitchell PJ, et al.: Results of a multicentre, randomised controlled trial of intra-arterial urokinase in the treatment of acute posterior circulation ischaemic stroke. Cerebrovasc Dis. 2005; 20(1): 12–7. PubMed Abstract | Publisher Full Text\n\nArcher CR, Horenstein S: Basilar artery occlusion: clinical and radiological correlation. Stroke. 1977; 8(3): 383–90. PubMed Abstract | Publisher Full Text\n\nHacke W, Zeumer H, Ferbert A, et al.: Intra-arterial thrombolytic therapy improves outcome in patients with acute vertebrobasilar occlusive disease. Stroke. 1988; 19(10): 1216–22. PubMed Abstract | Publisher Full Text\n\nSchonewille WJ, Algra A, Serena J, et al.: Outcome in patients with basilar artery occlusion treated conventionally. J Neurol Neurosurg Psychiatry. 2005; 76(9): 1238–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMehler MF: The rostral basilar artery syndrome: diagnosis, etiology, prognosis. Neurology. 1989; 39(1): 9–16. PubMed Abstract | Publisher Full Text\n\nDevuyst G, Bogousslavsky J, Meuli R, et al.: Stroke or transient ischemic attacks with basilar artery stenosis or occlusion: clinical patterns and outcome. Arch Neurol. 2002; 59(4): 567–73. PubMed Abstract | Publisher Full Text\n\nHacke W, Kaste M, Fieschi C, et al.: Randomised double-blind placebo-controlled trial of thrombolytic therapy with intravenous alteplase in acute ischaemic stroke (ECASS II). Second European-Australasian Acute Stroke Study Investigators. Lancet. 1998; 352(9136): 1245–51. PubMed Abstract | Publisher Full Text\n\nSchonewille WJ, Wijman CA, Michel P, et al.: Treatment and outcomes of acute basilar artery occlusion in the Basilar Artery International Cooperation Study (BASICS): a prospective registry study. Lancet Neurol. 2009; 8(8): 724–30. PubMed Abstract | Publisher Full Text\n\nEckert B, Kucinski T, Pfeiffer G, et al.: Endovascular therapy of acute vertebrobasilar occlusion: early treatment onset as the most important factor. Cerebrovasc Dis. 2002; 14(1): 42–50. PubMed Abstract | Publisher Full Text\n\nArnold M, Nedeltchev K, Schroth G, et al.: Clinical and radiological predictors of recanalisation and outcome of 40 patients with acute basilar artery occlusion treated with intra-arterial thrombolysis. J Neurol Neurosurg Psychiatry. 2004; 75(6): 857–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNoufal M, Schmidley JW, Erdem E, et al.: Basilar artery occlusion treated with mechanical thrombectomy beyond eight hours with successful recanalization and good functional outcomes. Cerebrovasc Dis. 2009; 27(6): 614–5. PubMed Abstract | Publisher Full Text\n\nSchulte-Altedorneburg G, Hamann GF, Mull M, et al.: Outcome of acute vertebrobasilar occlusions treated with intra-arterial fibrinolysis in 180 patients. AJNR Am J neuroradiol. 2006; 27(10): 2042–7. PubMed Abstract\n\nStrbian D, Sairanen T, Silvennoinen H, et al.: Thrombolysis of basilar artery occlusion: impact of baseline ischemia and time. Ann Neurol. 2013; 73(6): 688–94. PubMed Abstract | Publisher Full Text\n\nCho TH, Nighoghossian N, Tahon F, et al.: Brain stem diffusion-weighted imaging lesion score: a potential marker of outcome in acute basilar artery occlusion. AJNR Am J Neuroradiol. 2009; 30(1): 194–8. PubMed Abstract | Publisher Full Text\n\nLindsberg PJ, Sairanen T, Nagel S, et al.: Recanalization treatments in basilar artery occlusion—Systematic analysis. Eur Stroke J. 2016; 1(1): 41–50. Publisher Full Text\n\nKang DH, Jung C, Yoon W, et al.: Endovascular Thrombectomy for Acute Basilar Artery Occlusion: A Multicenter Retrospective Observational Study. J Am Heart Assoc. 2018; 7(14): pii: e009419. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGory B, Mazighi M, Blanc R, et al.: Mechanical thrombectomy in basilar artery occlusion: influence of reperfusion on clinical outcome and impact of the first-line strategy (ADAPT vs stent retriever). J Neurosurg. 2018; 129(6): 1482–1491. PubMed Abstract | Publisher Full Text\n\nSon S, Choi DS, Oh MK, et al.: Comparison of Solitaire thrombectomy and Penumbra suction thrombectomy in patients with acute ischemic stroke caused by basilar artery occlusion. J Neurointerv Surg. 2016; 8(1): 13–8. PubMed Abstract | Publisher Full Text\n\nGerber JC, Daubner D, Kaiser D, et al.: Efficacy and safety of direct aspiration first pass technique versus stent-retriever thrombectomy in acute basilar artery occlusion-a retrospective single center experience. Neuroradiology. 2017; 59(3): 297–304. PubMed Abstract | Publisher Full Text\n\nMokin M, Sonig A, Sivakanthan S, et al.: Clinical and Procedural Predictors of Outcomes From the Endovascular Treatment of Posterior Circulation Strokes. Stroke. 2016; 47(3): 782–8. PubMed Abstract | Publisher Full Text\n\nLutsep HL, Rymer MM, Nesbit GM: Vertebrobasilar revascularization rates and outcomes in the MERCI and multi-MERCI trials. J Stroke Cerebrovasc Dis. 2008; 17(2): 55–7. PubMed Abstract | Publisher Full Text\n\nSheng K: Therapy for acute basilar artery occlusion: a systematic review and meta-analysis. 2019. http://www.doi.org/10.17605/OSF.IO/4A27M"
}
|
[
{
"id": "44183",
"date": "14 Feb 2019",
"name": "José Biller",
"expertise": [
"Reviewer Expertise Cerebrovascular Disease (Clinical Trials",
"Stroke in the Young) JB",
"Neurocritical Care RG"
],
"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“Therapy for acute basilar artery occlusion: a systematic review and meta-analysis”\n\nBackground:\n\nBasilar artery stroke is most commonly caused by atherothrombosis and cardio embolism. The clinical manifestations of basilar artery occlusive disease vary according to the site and nature of vascular compromise. Pertinent clinical profiles pertaining to the topic and the discussion include: 1) proximal and middle basilar artery occlusive disease, frequently of atherosclerotic origin accounting for unilateral or bilateral pontine dysfunction, and less often cerebellar, midbrain, occipital, or mesial temporal lobe ischemia; 2) distal basilar artery occlusion (\"top of the basilar syndrome\"), frequently of embolic origin and accounting for signs of midbrain and thalamic ischemia, occipital and mesial temporal lobe ischemia, or both.\n\nThe authors performed a systematic review and meta-analysis comparing three methods of recanalization for acute basilar artery occlusion. The authors should be commended for tackling a difficult topic.\n\nQuestions and Comments:\n\nTitle:\nIdentified as a systematic review and meta-analysis.\n\nAbstract:\nConsider adding one-two sentences about background/significance/context of basilar artery occlusion. The purpose and objective is made clear. Would clarify what databases are used, study selection criteria, interventions compared, data extraction method (i.e. PRISMA) all in the methodology portion of the abstract. Results are satisfactory. In the limitations section, consider adding a comment about heterogeneity. This is well described in the discussion, but would add something in the abstract.\n\nIntroduction:\nDescribes the importance that the meta-analyses adds to the current literature – no comments.\n\nNatural History of Disease: No comments.\n\nMethodology:\nSearch Strategy: Why was only PubMed used? We would recommend cross referencing other data-bases (SCOPUS, Google Scholar, COCHRANE). Were non-English studies included? We would recommend making an explicit statement regarding this detail.\n\nData extraction and Statistical Analyses: No comments. Excellent detail.\n\nResults:\nWe appreciate that results are not in both tabular and text form.\n\nDiscussion:\nEndovascular thrombectomy versus other approaches: We would expand as much as possible on the last sentence – confounding by virtue of “thrombus volume, location and length, and presence of collateral circulation. This is a major source of confounding when comparing a thrombolysis responder versus non-responder.\n\nConsider the following references: Mehta et al., (20121), Alemseged et al., (20172) and Goyal et al., (20163).\n\nIn the above meta-analyses by Goyal et al. (20163), consider commenting on the Forest Plot in Figure 2. No subgroup effect of alteplase, but also no evidence of effect modification. This supports your paper’s conclusion.\n\nWould re-search the literature for updates in this arena, as it would greatly add biological plausibility to your clinical conclusion.\n\nFactors affecting outcome: No comments\n\nAssessment of heterogeneity: No comments.\n\nLimitations: No comments.\n\nOverall, this is an excellent paper. We suggest from a methodology standpoint to stay as close to PRISMA guidelines as possible. Additionally, we suggest expanding on sources of confounding when comparing recanalization therapies.\n\nJosé Biller, MD, FACP, FAAN, FANA, FAHA Ravi Garg, MD\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": "45712",
"date": "08 Apr 2019",
"name": "Fana Alemseged",
"expertise": [
"Reviewer Expertise Stroke and acute imaging. Main area of interest is posterior circulation stroke",
"basilar artery occlusion."
],
"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 systematic review and meta-analysis aims to analyse the efficacy and safety of different treatment approaches for acute basilar artery occlusion. The paper reads reasonably well and the authors acknowledged the main limitations of the study. However, the paper could be improved by the following revisions:\nAbstract: Please add a description of the main outcomes in the methods. Please include relevant definitions for “good clinical outcome” (mRS 0-2), “recanalization” (TICI 2b-3, mTICI 2b-3, TIMI 2-3) and “sICH” (ECASS II).\n\nPlease cite the results of the BEST trial in the introduction.\n\nPlease briefly describe the main results in the “Primary outcomes for the IVT, IAT and EVT groups” and “Secondary outcomes for the IVT, IAT and EVT groups”, in addition to the relevant tables and figures.\n\nCan the authors provide a descriptive figure for clinical outcome, recanalization and safety profile for ADAPT vs stent retriever thrombectomy?\n\nFigures 2b-c and 3b-c are not legible. Please improve the quality of these figures.\n\nPlease add a comment and cite the following article in the discussion: “Firstline a direct aspiration first pass technique versus firstline stent retriever for acute basilar artery occlusion: a systematic review and meta-analysis” by Ye et al. (20191).\n\nPlease rephrase the sentence \"There was a statistically significant correlation towards increased recanalization across time with IAT and EVT therapy, but not IVT therapy\" for greater clarity.\n\nConsider deleting the paragraph on the use of Solitaire vs Merci device in anterior LVO strokes as not pertinent.\n\nThere are several typos in the abstract, main body of the manuscript and figures: i.e. modified Ranking Score, incleded, etc. Please review the text for accuracy.\n\nPlease use scientific terminology rather than common language throughout the paper (i.e. people>patients, in different ways>different clinical presentation, etc.).\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? Partly\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-165
|
https://f1000research.com/articles/7-1813/v1
|
19 Nov 18
|
{
"type": "Research Article",
"title": "Real-time 3D Photoacoustic Visualization System with a Wide Field of View for Imaging Human Limbs",
"authors": [
"Kenichi Nagae",
"Yasufumi Asao",
"Yoshiaki Sudo",
"Naoyuki Murayama",
"Yuusuke Tanaka",
"Katsumi Ohira",
"Yoshihiro Ishida",
"Atsushi Otsuka",
"Yoshiaki Matsumoto",
"Susumu Saito",
"Moritoshi Furu",
"Koichi Murata",
"Hiroyuki Sekiguchi",
"Masako Kataoka",
"Aya Yoshikawa",
"Tomoko Ishii",
"Kaori Togashi",
"Tsuyoshi Shiina",
"Kenji Kabashima",
"Masakazu Toi",
"Takayuki Yagi",
"Yoshiaki Sudo",
"Naoyuki Murayama",
"Yuusuke Tanaka",
"Katsumi Ohira",
"Yoshihiro Ishida",
"Atsushi Otsuka",
"Yoshiaki Matsumoto",
"Susumu Saito",
"Moritoshi Furu",
"Koichi Murata",
"Hiroyuki Sekiguchi",
"Masako Kataoka",
"Aya Yoshikawa",
"Tomoko Ishii",
"Kaori Togashi",
"Tsuyoshi Shiina",
"Kenji Kabashima",
"Masakazu Toi",
"Takayuki Yagi"
],
"abstract": "Background: A breast-specific photoacoustic imaging (PAI) system prototype equipped with a hemispherical detector array (HDA) has been reported as a promising system configuration for providing high morphological reproducibility for vascular structures in living bodies. Methods: To image the vasculature of human limbs, a newly designed PAI system prototype (PAI-05) with an HDA with a higher density sensor arrangement was developed. The basic device configuration mimicked that of a previously reported breast-specific PAI system. A new imaging table and a holding tray for imaging a subject's limb were adopted. Results: The device’s performance was verified using a phantom. Contrast of 8.5 was obtained at a depth of 2 cm, and the viewing angle reached up to 70 degrees, showing sufficient performance for limb imaging. An arbitrary wavelength was set, and a reasonable PA signal intensity dependent on the wavelength was obtained. To prove the concept of imaging human limbs, various parts of the subject were scanned. High-quality still images of a living human with a wider size than that previously reported were obtained by scanning within the horizontal plane and averaging the images. The maximum field of view (FOV) was 270 mm × 180 mm. Even in movie mode, one-shot 3D volumetric data were obtained in an FOV range of 20 mm in diameter, which is larger than values in previous reports. By continuously acquiring these images, we were able to produce motion pictures. Conclusion: We developed a PAI prototype system equipped with an HDA suitable for imaging limbs. As a result, the subject could be scanned over a wide range while in a more comfortable position, and high-quality still images and motion pictures could be obtained.",
"keywords": [
"photoacoustic imaging",
"optoacoustic",
"hemispherical detector array",
"laser",
"limb",
"haemoglobin oxygen saturation",
"motion pictures",
"blood vessel"
],
"content": "Introduction\n\nBlood vessels are important for the delivery of oxygen and nutrients to the entire body. Vascular imaging plays an important role in the diagnosis of congenital vascular abnormalities, diseases of the blood vessels themselves, and angiogenesis associated with diseases such as cancer1. Various modalities are used in clinical practice to diagnose diseases by imaging blood vessels. These modalities have some disadvantages, such as the need for contrast medium, exposure to X-ray radiation, or expensive equipment, such as magnetic resonance imaging (MRI) systems. On the other hand, blood flow imaging is also possible with Doppler ultrasound (US), which does not require contrast medium and is not invasive; however, its spatial resolution is limited.\n\nPhotoacoustic tomography (PAT) can be used to visualize blood vessels with high resolution noninvasively2. In particular, systems3–10 using a hemispherical detector array (HDA) as a sensor can reconstruct blood vessel images with good 3D reproducibility. We produced a prototype breast-specific PAT system (PAI-03, 04)3,5–10 and used it to conduct clinical research by scanning breast cancer patients in the prone position and visualize tumor-related blood vessels caused by breast cancer. The probe arrangement of the HDA was designed to solve the \"limited view problem\"11 that causes PAT image degradation. This probe arrangement surrounds the measurement target with a much wider solid angle than that of a linear probe with the PAT apparatus12–17.\n\nThe above studies demonstrated the acquisition of high-resolution 3D still images with a hemispherical probe over a large area by wide-range scanning of the measurement area.\n\nWe also scanned the healthy blood vessels in various body parts other than the breasts with these devices and reported the potential clinical usefulness of this approach6–10. By analyzing the blood vessels in the palms of healthy volunteers, we showed that the tortuosity of blood vessels increases with age, suggesting the possibility of assessing the risk of diseases such as arteriosclerosis and other lifestyle diseases. In addition, by imaging perforators in the anterolateral thigh (ALT), this method was demonstrated to be effective for the preoperative planning of ALT free flaps. Although the possibility of this system being useful for imaging blood vessels in contexts other than breast cancer, especially in limbs, has been suggested, there were some related issues, such as the subject having to assume an awkward position and the measurement range being limited due to the shape of the imaging table, which was originally designed for studying breast cancer.\n\nWith the aim of imaging blood flow in limbs, we have developed a new PA imaging (PAI) system prototype (PAI-05) equipped with an HDA that generally follows a conventional design. Furthermore, the system is capable of real-time 3D imaging. This paper outlines the device configuration and introduces examples of biological images obtained using this device.\n\n\nMethods\n\nThe basic design of PAI-05, which was developed in cooperation with Canon, Inc., Hitachi, Ltd., and Japan Probe Co., Ltd., mimics that of the conventional prototype3,5–10 equipped with an HDA that receives the PA signal. An overall image of PAI-05 is shown in Figure 1a. The PAI-05 system consists of a bed unit, including an HDA and a scanning stage, a light source, a fiber bundle, a data acquisition system (DAS), a real-time image reconstruction unit, and an operation personal computer (PC). In this paper, the horizontal plane is the x-y plane, and the x and y axes are the short and long axes of the bed unit, respectively. The z axis is the vertical axis.\n\n(a) Top view of the PAI-05 system, which consists of a laser light source, a data acquisition unit, an image reconstruction unit, a bed unit, and an operation PC. A scanning stage with an hemispherical detector array (had) set in the bed unit. Oscillating laser light is delivered to the bottom of the HDA through a fiber bundle. A subject can be observed by an upper camera. (b) Cross-sectional view of the HDA and holding tray cut at the center during the photoacoustic (PA) measurement. The generated PA wave is received by the HDA through the holding tray, which consists of plastic mesh and a separate film., The PA signal is then sent to the data acquisition system (DAS).\n\nWhen scanning, the body part of the subject is inserted into the holding tray on the bed unit (Figure 1b). The previous prototypes adopted hemispherical holding cup shapes because they were designed for imaging breasts, but in the PAI-05, the holding tray has a flat bottom to facilitate imaging of the hands and feet. In addition, the bed size was enlarged so that a subject could be measured in various postures. To maintain the flat shape of the bottom surface, it was supported by plastic mesh, and a sheet made of polyethylene terephthalate (PET) film was used on the mesh to separate the acoustic matching water on the HDA from the tray water, which allowed the subject to be immersed (Supplementary Video 1).\n\nConventionally, the subject's breast was inserted in the hemispherical holding cup while the subject was in the prone position. On the other hand, in the PAI-05 system, to make it easier to scan limbs, the subject lies next to the holding tray filled with water for acoustic matching. This configuration allows the subject to be scanned in a more comfortable posture than that required for imaging the palm with the previous prototype6. Figures 2a–c show the posture assumed when scanning the palm, back of the hand, and sole of the foot, respectively. The size of the holding tray was 529 × 259 mm, and the depth was 29.4 mm. The bed size was 1.4 m along the x axis, 2.7 m along the y axis and 0.48 m along the z axis.\n\n(a) Example posture assumed during the capture of a PA image of a palm. The subject lies down on the side of the holding tray and inserts the hand into the tray naturally. (b) Example posture used for imaging the back of the hand. (c) Example posture used for imaging the sole of the foot.\n\nIn the conventional breast-specific PAI system, the object to be imaged was nearly round, so it was spirally scanned to form a PA image of circular scanning range3,5–10. On the other hand, in the case of limbs, since the measurement target is often noncircular with the direction of the bone as the major axis, the scanning range is rectangular. To avoid a decrease in speed when changing the scanning direction at the vertex, the corners were scanned in a manner producing a smooth locus (Supplementary Video 2).\n\nThe measurement target was immersed in a holding tray, and the scanning unit containing the HDA was placed under the bed. The maximum still image measurement area was 270 mm × 180 mm, and the measurement time was proportional to the imaging area. The measurement area size and measurement time are shown in Table 1.\n\nA giant-pulse laser beam with a pulse width of less than 20 ns was directed upward along the z axis from the emission end of the fiber bundle on the bottom surface of the HDA. One laser system per wavelength (Lotis TII, Belarus) was used. Using two sets of lasers, fiber bundles were used to guide laser light to the HDA, and alternate irradiation at the two wavelengths was performed. Part of the fiber bundles was pulled out to detect and synchronize the signal reception timing at the DAS unit. The diverging lens located at the base of the HDA conically spread the light to a diameter of 24.8 mm at the surface of the holding tray. Figures 3a–b show the laser profiles.\n\n(a) Photograph of the laser irradiation profile. This picture was taken by a digital camera while a diffusion film was on the bottom of the tray. (b) Energy density distribution of the laser light along the dotted line in Figure 3a.\n\nOne laser irradiated light at a repetition frequency of 10 Hz. Each wavelength could be selected from the wavelength range of 750 to 850 nm using an optically pumped Ti:Sa laser and a Q-switched Nd:YAG laser. The wavelength could be set at steps of 1 nm.\n\nTo form a hemoglobin oxygen saturation distribution image, wavelengths of 756 and 797 nm were selected, as in the conventional case; 756 nm is the maximum point of the absorption coefficient of deoxygenated hemoglobin, and 797 nm is the isosbestic point of oxyhemoglobin and deoxygenated hemoglobin.\n\nIn PAI-05, alternate irradiation10,18 was performed using two different laser systems every 50 ms. The exposure was set to be smaller than the maximum permissible exposure recommended by ANSI at any wavelength. When two lasers were set to the same wavelength, they could irradiate at 20 Hz.\n\nWe used an HDA manufactured by Japan Probe Co., Ltd., for the PA signal detector. An ultrasonic flexible array probe19 using a 1–3 composite piezoelectric transducer made of lead zirconate titanate (PZT) was adopted to arrange 1024 channel sensors on a 55-mm-radius hemisphere of epoxy resin. By applying this device to the hemispherical film, a piezoelectric vibrator with a three-layer structure consisting of a protective layer, a composite resonator, and a damper material was formed (Figure 4a). Next, the lead wire was mounted on the hemispherical film-like piezoelectric vibrator using a fine soldering technique, thereby forming a 1024-channel element in a spherical shape, thus completing the HDA (Figure 4b). The diameter of the single circular element was 2 mm. On the hemispherical sensor, the sensor element was arranged according to a 3D Fibonacci grid20, as shown in Figures 4c–d. With this arrangement, the receiving element density per solid angle became nearly uniform, and artifacts generated by image reconstruction could be suppressed. The center frequency and the fractional bandwidth of the device were 3.34 MHz and over 85%, respectively. The average conversion efficiency was 1.9 mV/kPa at 2.5 MHz. The noise equivalent pressure (NEP) measurement was examined using a low-pass filter with a cut-off frequency of 5 MHz, and the NEP, the effective value without signal averaging, of this system was 3.42 Pa.\n\n(a) Photograph of a film-shaped ultrasound sensor used in the HDA. Thereafter, a lead wire was connected to a predetermined position to form a sensor array of 1024 channels. (b) Photograph of the completed HDA module. The large hole in the center is a laser emission hole. The drainage outlet and lower camera are installed next to the laser emission hole at the bottom of the HDA. (c) Side view of the probe array. Sensor positions are indicated as blue circles. (d) Top view of the probe array. Sensors are placed according to the 3D Fibonacci grid, excluding areas indicated by blank or orange X marks, where the laser emission hole, drainage outlet and lower camera are located. (e)–(j) Diagrams schematically showing the photoacoustic (PA) signal intensities reaching each sensor element when a rod-shaped phantom is installed directly above the HDA and a PA signal is generated. Pure red color indicates a strong PA signal arriving at a sensor, and white color indicates no PA signal. (e) Side view of the PA signal intensity distribution on the HDA when a rod-shaped phantom is horizontally installed and a signal reaches the sensor. (f) Top view of the PA signal intensity distribution under the same conditions as in Figure 4e. (g) Side view of the PA signal intensity distribution on the HDA when a rod-shaped phantom is installed inclined at an angle of 30 degrees from the horizontal and a signal reaches the sensor. (h) Top view of the PA signal intensity distribution under the same conditions as in Figure 4g. (i) Side view of the PA signal intensity distribution on the HDA when a rod-shaped phantom is installed inclined at an angle of 30 degrees from the horizontal and a signal reaches the sensor. (j) Top view of the PA signal intensity distribution under the same conditions as in Figure 4g.\n\nThe two video images, one from the lower camera (not shown) installed on the bottom of the HDA, and one from the upper camera, as shown in Figure 1a, at the top were displayed on the screen during scanning.\n\nThe PA signal was received by the HDA, and the signal data were transferred to the DAS, which was manufactured by Hitachi, Ltd. The DAS amplified the PA signal of 1024 channels at the time of light irradiation from the laser unit, simultaneously sampled 1024 channels at 60 MHz and 12 bits, and converted the signals into digital data. The digital data were sequentially transferred to the image reconstruction PC.\n\nReal-time image reconstruction could be performed during scanning, and the reconstructed PA image was displayed on the operation PC.\n\nThe Digital Imaging and Communications in Medicine (DICOM) image data reconstructed inside the PAI-05 system based on the DICOM standard (version 3.0) could be automatically transferred to the DICOM image server of Kyoto University Hospital.\n\nThe PAI-05 system could perform two types of imaging: one in still mode, in which a large area could be imaged by scanning with the HDA; and another in movie mode, in which a specific place was irradiated with laser light to repeatedly obtain updated PA images of that area.\n\nIn still mode, images could be acquired with the scanning sizes shown in Table 1, and the PA images were generated by scanning with the HDA and taking the average of more than 24 scans within the imaging range. For PAI-05, imaging ranges of 40 mm × 40 mm, 50 mm × 100 mm, 100 mm × 50 mm, 135 mm × 180 mm, and 270 × 180 mm could be selected. The scanning stage was controlled to move in a rectangular spiral shape.\n\nTwo types of HDA scanning operations were available in movie mode. One type was denoted as simple movie (SM) mode, in which the PA image is updated by repeatedly irradiating one spot with laser light while the HDA remained stationary on the spot. The other type was denoted as fluctuation movie (FM) mode, in which imaging was performed during continuous minute rotational motion. In any movie mode, it was possible to acquire an accumulation of images over time for a reconstructed image according to the number of times of laser light was irradiated. In FM mode, the scanning operation occurred at 3 rotation/sec such that the center of the HDA would follow the trajectory of a circle with a diameter of 3 mm (Supplementary Video 3).\n\nWe used universal back-projection (UBP)21 for PA image reconstruction. An image was created for each shot of the laser and recorded in movie or still mode. The former method yielded real-time motion pictures, and the latter method yielded a high-quality still image by accumulating scanned images to construct an image with a wide field of view (FOV). The volume shape of the reconstructed image per irradiation, which was occurred every 50 ms, was a circular cylinder. The base was a circle with a diameter of 20 mm in the x-y plane, and the height was 30 mm in the z direction. These image reconstructions were realized in real time by pipeline-processing data transfer and reconstruction operations using five graphics processing units (GPUs, FirePro S9150, Advanced Micro Devices, Inc., USA). The state of reconstruction during scanning was displayed on the operation PC. The voxel size when displaying the reconstruction on the display in real time was 0.1 mm in the x, y, and z directions.\n\nThe limitation of this system is the “limited view problem”11, which is commonly found in other PA systems. Adopting the HDA in the PAI-05 system remarkably alleviates the problem compared to the use of handheld (HH) devices. Nonetheless, the “limited view problem” persists in the PAI-05 system. Figures 4e–j show the result of simulating PA waves generated from rod-like subjects placed at the center of the sensor. The angle of each rod-like subject was set to 0, 30, and 60 degrees, as illustrated in Figures 4e–f, 4g–h, and 4i–j, respectively. When the rod-like subject was set to the horizontal plane or when the inclination was approximately 30 degrees, the signal could be fully received by the HDA. If the subject was set to an inclination of 60 degrees, the PA signal was received at almost the top of the HDA. Although our simulation showed that signals could be received from subjects up to a tilt to 70 degrees, image reconstruction cannot be achieved with a rod-like absorber that is more inclined, i.e., close to vertical. It is considered that many vessels underlying the limbs targeted by the PAI-05 system are largely parallel to the surface of the skin in general, but careful attention is required for analysis.\n\nIdeally, oxygen saturation can be calculated using equation (1) if the absorption coefficients of two wavelengths can be correctly obtained.\n\nSO2=[HbO2][HbO2]+[Hb]= μaλ2(r)μaλ2(r). εHbλ1–εHbλ2εΔHbλ2–μaλ2(r)μaλ1(r).εΔHbλ1(1)\n\nwhere λ1 and λ2 represent wavelengths of 756 and 797 nm, respectively; r is the position to be calculated; εΔHb is the molar extinction coefficient of deoxyhemoglobin; and εΔHb is the difference in the molar extinction coefficients between deoxyhemoglobin (Hb) and oxyhemoglobin (HbO2).\n\nIn real situations, since ideal hemoglobin oxygen saturation (SO2) cannot be obtained due to various error factors, we refer to the parameter obtained by the two wavelengths as the S-factor15. As described previously10, the relation of the magnitude of SO2 and the S-factor is maintained within the range where the amount of irradiated light can be regarded as the same as that in a neighboring region.\n\nCalculation of the S-factor was not performed by the PAI-05 system; instead, after data acquisition, PA signal data were copied to another PC, and the calculation were performed off-line. A weighted S-factor15 was used for image display by producing a weighted image with a signal intensity of 797 nm. To calculate these S-factors and display PA images, we used a PAT-dedicated viewer named Kurumi [version 3.9122].\n\nKurumi is equipped with a body surface detection function that uses cloth simulation23, which makes it easier to analyze deeper areas by deleting unnecessary image information, such as that related to subcutaneous vein networks, as necessary. To detect the position of the body surface, an image obtained at 797 nm was mainly used.\n\nThe phantom used in this study was as follows. Surgical thread (11-0, thickness, 10–19 μm, Monosof, Medtronic plc, Ireland) was used for a line spread function (LSF) evaluation (not shown). The thread was placed parallel to the horizontal plane at a depth of a few mm closer to the tray bottom from the center of curvature of the HDA. Figures 5a–b show the phantom structure that was used to evaluate the penetration depth. A wire phantom (diameter, 0.3 mm, μa: 0.22 mm-1) mimicking the light absorption coefficient of blood at 797 nm was placed in Intralipos Injection 20％ (Otsuka Pharmaceutical Co., Ltd., Japan) diluted to 1% concentration (μa: 0.0022 mm-1, μs': 0.921 mm-1); μa is the absorption coefficient, and μs' is the equivalent scattering coefficient. Five wire phantoms were installed at positions of 0, 5, 10, 15 and 20 mm along the z axis from the bottom of the tray.\n\n(a) Photograph of a phantom used for evaluating the penetration depth of the PAI-05 system. (b) Schematic illustration of the phantom shown in Figure 5a. (c) Photograph of a phantom used for evaluating the viewing angle of the PAI-05 system. (d) Schematic illustration of the phantom shown in Figure 5c. (e) Enlarged image of schematic in Figure 5d.\n\nTo evaluate the allowance of the visualization range of the “limited view problem,” we used wire phantoms (Figures 5c–d) installed in Intralipos at an angle to the sensor scanning plane or the x-y plane.\n\nTo simply evaluate the wavelength dependence of the PA signal intensity, a line written on white office paper with four colors of oil-based ink pen (black, red, green and blue) was used. A V-680 spectrophotometer (JASCO Co., Japan) was used to evaluate the reflectance spectral characteristics of the office paper and oil-based ink written on the paper.\n\nFor living body measurements, one healthy male subject was recruited. He was 173 cm tall, his foot was 25.5 cm long, and he was in his fifties. The hair of the imaged limb was cut beforehand with hair clippers, but when the remaining hair roots interfered with the analysis, the hair root data was excluded using the body surface detection method described above. This subject was registered as a healthy volunteer in a control group of an exploratory clinical study for examining the vascular condition of patients with a skin disease using the same PAI-05 system. In the current study, biological data from the clinical study were only utilized for the presentation of imaging examples. The results of the skin disease analyses will be reported in the near future.\n\nThe present study was approved by the Ethics Committee of the Kyoto University Hospital (UMIN 000022767), and written informed consent was obtained from the subject. This study was conducted in accordance with the Declaration of Helsinki.\n\n\nResults\n\nThe performance of the system was evaluated using phantoms. The results are shown in Figure 6. The results obtained with the surgical thread phantom are shown in Figures 6a-c. As a result of evaluating the LSF, the full width at half maximum (FWHM) in the direction parallel to the HDA scanning plane was 0.21 mm, and the FWHM for the z axis was also 0.21 mm.\n\nExperimental results of phantoms for the line spread function (LSF) (a–c), penetration depth (d–e) and viewing angle (f–g). (a) Maximum intensity projection (MIP) image on the x-y plane obtained by scanning the phantom made of a thin thread. (b) Cross-sectional view of Figure 6a as viewed from the y-z plane. (c) Line profile of photoacoustic (PA) intensity at the centerline of Figure 6b. (d) MIP image of a phantom for penetration depth evaluation. (e) Profile of PA intensity at x=0 in Figure 6d. (f) MIP image of a phantom for viewing angle evaluation. (g) PA intensity diagram for each phantom with different angles.\n\nFigures 6d–e show the results of the depth performance evaluation. A wire phantom at a depth of 20 mm in the Intralipos could be visualized with a signal-to-noise ratio (SNR) of the maximum intensity projection (MIP) image of approximately 8.5, where the noise value was the average of the signal background level. The results of evaluating the “limited view” are shown in Figures 6f–g. It was experimentally shown that a 70-degree wire phantom could be visualized. A 75-degree wire was difficult to observe.\n\nThe image quality of the movie mode was evaluated using a wire phantom while changing the number of PA images averaged in both SM and FM modes.\n\nFigures 7a–b show examples of images without averaging, and Figures 7c–d show examples of five images averaged in FM mode. There was an obvious difference in the background noise of the cross-sectional view of the phantom without averaging (i.e., an average of 1) and with averaging of 5 images, as shown in Figures 7b and 7d. To quantify the amount of noise, the background noise of data obtained in FM and SM mode was evaluated using the root mean square (RMS) value. The background was defined as inside the area (x: 6 mm, y: 2.5 mm, z: 2.5 mm) where the wire phantom did not exist.\n\n(a) Maximum intensity projection (MIP) image of the x-y plane of only one shot (i.e., without averaging). (b) MIP image of the y-z plane in Figure 7a. (c) MIP image of the x-y plane with N=5 in fluctuation movie (FM) mode. (d) MIP image of the y-z plane in Figure 7c. (e) Graph showing the normalized background noise as a function of N. In SM mode, the noise level does not decrease even if N is increased, but the noise level decreases with 1/Route N in FM mode. However, the dependence on the number of averaged images disappears after N=10, and the noise remained nearly constant in FM mode. The dashed line indicates an approximate curve of N in FM mode, showing 1/Route N before the N=10 and a constant line after N=10. (f) The upper row shows tomographic images of the y-z plane with only one shot for the first to the fourth images in FM mode. To facilitate the artifact analysis, binarization was performed; values 1/20 or more of the peak intensity was considered noise, and the intensity value of that position was set to 1; the intensity value of the remaining positions, where the noise was less than 1/20 of the peak intensity, was set to zero. From left to right, the lower row shows the images obtained in FM mode with N=2, N=3, and N=4. The intensity of the true signals of the phantom is almost 1, and the intensity of the artifact decreases as N increases.\n\nFor the noise analysis, a plot of the background noise after normalization by the background noise without averaging was made as a function of the number of images averaged (N), as shown in Figure 7e. The background noise in FM mode decreased in proportion to 1/N until N reached ten; furthermore, when N was 11 or more, the background noise did not decrease but remained nearly constant. Because the rotational scanning of the sensor in FM mode was 3 rotations/sec, the first data acquisition position coincided with the eleventh position.\n\nIn SM mode, reducing the background noise only slightly reduced the system noise with respect to N. The FM mode showed a greater noise reduction effect than the SM mode. To analyze the noise reduction effect in FM mode, Figure 7f was created. Each image in the upper row represents 1 shot at a different position in FM mode; binarization processing was performed with 1 as the part exceeding the 1/20 luminance value of the wire phantom contained in the screen and displayed. Among these areas, the noise area is a portion that expresses 1 even though no wire phantom subject exists in that area. The noise extends in a streak shape from the wire portion and is considered an image artifact. The lower row shows FM mode images obtained by averaging 2 shots, 3 shots, and 4 shots, respectively, from the diagram on the left. As the images overlapped after shifting the position of the streak artifact, the intensity of the noise could be reduced by averaging the whole images. Needless to say, the image averaging caused degradation of the temporal resolution. There was an obvious trade-off between temporal resolution and image quality in FM mode, and degradation of the temporal resolution occurred almost without improving image quality in SM mode.\n\nWith the PAI-05 system, the wavelength could be arbitrarily selected from 750 to 850 nm. Figure 8a shows the reflectance spectral characteristics of the white office paper and four colors of oil-based ink colors (black, red, green, blue) written on the paper. Figure 8b shows the wavelength dependence of the PA intensity obtained using the PAI-05 system. As the light intensity reflected by the sample was small, the light absorbance was increased, which resulted in increased PA intensity. This is a valid result as a PA property. No PA signals were obtained from the white office paper or red ink at all.\n\n(a) Diagram showing the reflectance spectra for oil ink on white office paper. In the legend, white indicates the reflectance of the office paper, and black, red, green, and blue indicate the reflectance of the ink on the office paper. (b) Diagram showing the photoacoustic (PA) signal intensity of a line of oil-based ink on white office paper. For the black, green, and blue ink, the wavelength dependence of the PA signal intensity was measured. No PA signal was detected for either the unmarked office paper or the location of the red ink.\n\nNext, the living subject was imaged. This system was newly designed to facilitate the imaging of limbs. We designed the holding tray to be shallower and wider than before to make it more comfortable for a subject to assume the posture required for scanning. Figures 9a–g are examples of PA images of the extremities of a living body. It was confirmed that PA images in a wide range of 270 mm × 180 mm could be obtained in each area.\n\n(a) Palm. (b) Back of the hand. (c) Forearm. (d) Anterolateral thigh (ALT). (e) Lower thigh. (f) Top of the foot. (g) Sole of the foot. (h) Schematic diagram showing the locations of Figures 9a–g.\n\nSimilar to the results of the phantom experiments, as shown in Figure 9a and other figures, PA images of blood vessels could be reconstructed with definition equal to or higher than that of the previous prototypes6–8.\n\nIn the image of the sole of the foot shown in Figure 9f, the PA image of blood vessels in the heel region was not recognized as a network shape, unlike the surroundings. This may be because body weight was applied to the heel at the time of measurement, thus inhibiting blood flow in the subcutaneous veins.\n\nFigure 9g shows a PA image of the dorsum of the foot. The whole image of the region could not be shown because the whole dorsal foot was not immersed in water in the shallow holding tray. Likewise, the upper and lower parts of the lower leg shown in Figure 9e were not visualized because they protruded from the water.\n\nThe thigh shown in Figure 9d was shaved in advance, but black hair roots remained, and the signal intensity of these roots was stronger than that of the blood vessels. Therefore, the image produced by removing the hair volume data using cloth simulation is shown.\n\nFigures 10a–e show examples of images acquired using different scanning areas in still mode. The maximum still image area of 270 mm × 180 mm (Figure 10e) was imaged in approximately 10 minutes, and the minimum still image area of 40 mm × 40 mm (Figure 10a) was imaged within one minute, as shown in Table 1.\n\nFigure 11 shows S-factor images of the palms and thigh at 756 and 797 nm under the same conditions as in previous papers5,10,15 using alternating irradiation from two different lasers10. A PA image showing blood vessels that could be distinguished as an artery or vein was obtained. Figures 11a–b show images of blood vessels in the palm. Figure 11a shows a total MIP image including the skin surface, and 11b shows a MIP image of the deep region after deletion of the subcutaneous vein network. It is generally known that an artery is accompanied by one or two vein(s). Such accompanying blood vessels near the common palmar digital arteries could be visually recognized from these figures that by their different S-factor values. Figures 11c–d show images of the blood vessels in the ALT. Figure 11c shows a total MIP image including the skin surface, and Figure 11d shows a MIP image of the deep region after deletion of the surface image of hair roots, skin melanin, and subcutaneous vein network. As with the image of the palm, it was visually recognized that both arteries and veins were present side by side.\n\n(a) Image 40 * 40 mm in the x-y direction. (b) Image 100 mm in the x-direction * 50 mm in the y direction. (c) Image 50 mm in the x-direction * 100 mm in the y direction. (d) Image 180 mm in the x-direction * 135 mm in the y direction. (e) Image 180 mm in the x-direction * 270 mm in the y direction. The yellow bar represents 20 mm in all figures.\n\n(a) Gray scale image of the palmar. Yellow square suggests the area of calculation of S-factor shown in Figure 11b. (b) Examples of common palmar digital arteries and their accompanying veins. (c) Gray scale image of whole anterolateral thigh (ALT) image. Yellow square suggests the area of calculation of S-factor shown in Figure 11d. (d) Artery and vein in the ALT. Yellow arrows suggest blood vessels considered perforators.\n\nIn the PAI-05 system, two laser wavelengths can be arbitrarily selected within the range of 750 to 850 nm. Stable S-factor images were obtained by adopting alternating irradiation. Figures 12a–b show the area used for analyzing PA intensity for different wavelengths. Figure 12c is an example of the analysis of the signal intensity of each region, showing dependence of the wavelength on each component of the blood vessels and melanin.\n\n(a) The results of an artery and a vein. (b) The result of skin melanin. (c) Each PA intensity as a function of the wavelength.\n\nFigures 13a–h show snapshots captured in movie mode within an FOV 2 cm in diameter. Figures 13a–e are snapshots of a motion picture (Supplementary Video 4) captured while the palm was moving left and right. Figures 13f–h are snapshots of motion pictures captured while the subject was pressing a finger onto the bottom of the tray; immediate changes in blood flow can be observed when the pressure was released (Supplementary Video 5). Although motion pictures captured by PAI have previously been reported, the effective the FOV of each individual volumetric frame was smaller than that of our system4. Our system could provide more extensive motion pictures.\n\n(a) Snapshots from the motion pictures of a hand being swung left and right in the horizontal plane (Supplementary Video 4). Yellow arrows indicate the same blood vessels. (b) Snapshots from the motion pictures of blood flow changing when a fingertip is pressed against the tray and then released (Supplementary Video 5).\n\n\nDiscussion\n\nIn the PAI-05 system, the HDA consisted of 1024 high-density sensors arranged inside a hemispherical casing with a diameter of 110 mm.\n\nIn the previous designs3,5–10, approximately 500 discrete rod-shaped sensors were inserted into a hemispherical casing with a diameter of 254 mm, but designing such a high-density arrangement for this study might have been difficult with conventional methods. The HDA design developed for this study using film-type sensors has the potential to greatly simplify the sensor manufacturing process for practical use because it was able to realize a 1024-channel arrangement, which may be difficult with discrete elements. This design made it possible to obtain images with less noise even with one shot, contributing to the realization of real-time motion pictures.\n\nWhile real-time motion pictures in PAT have mainly been reported in combination with conventional B-mode ultrasound16,17, the reproducibility of the blood vessel morphology in the PA images seems to be poor. Although the resolution seems to be good in the plane of a previously reported ring sensor24, the resolution in the direction normal to the ring surface is reportedly several mm or more. 3D structural analysis using images with such large anisotropy would be difficult.\n\nAs shown in Figure 6 and in other papers3–10, PA images obtained by a PAI system with an HDA show nearly isotropic spatial resolution, which is considered optimal for reproducing the morphology of blood vessels. Information for diagnosis can be obtained by performing 3D observations from different directions using not only still image but also real-time motion pictures. There is no doubt that larger imaging areas for capturing both still images and motion pictures will greatly contribute to the diagnosis of conditions in living organisms.\n\nRegarding the phantom evaluation, sufficient contrast was obtained at a depth of 2 cm in still mode. This depth is nearly sufficient for imaging the subcutaneous vessels of limbs. Linear-type phantoms could be imaged from 0 to 70 degrees with respect to the evaluation of the “limited view problem,” so it can be expected that most actual blood vessels in the living subject can be observed using the PAI-05 system. On the other hand, there is still an issue that perforators rising from deep regions may be difficult to visualize, as noted in a previous paper7. Clinical research in biological organisms should be continued, and diagnostic capability of the system should be verified.\n\nIn the phantom experiment for evaluating the image quality in movie mode, the noise could be reduced by increasing the number of images averaged (N), as shown in Figure 7. Generally, the amount of noise is proportional to 1/N in the case of white noise. The noise improvement effect in SM mode was smaller than 1/N, suggesting that the noise was not a random system noise. As shown in the upper row of Figure 7f, the shape of the noise showed good reproducibility. Therefore, it is suggested that it was a fixed artifact pattern. In FM mode, the position of the HDA changes with each laser irradiation. It seems that both the system noise and the effect of the artifact were reduced by increasing N, as shown in the lower row of Figure 7f. Because this suppression effect was in accordance with 1/N up to N=10, for which the position of the probe varied, it can be considered that the noise suppression effect was realized by superimposed data in an almost uncorrelated spatial state.\n\nAs for the evaluation of the living subject, it was possible to appropriately scan the limbs by designing the system configuration to be suitable for limb imaging. Since it is now possible to obtain PA images in a wide range, it is expected that diagnosing conditions affecting blood vessel will be made easier by viewing more complete images. These images will also be able to provide clinically useful information, such as that necessary for preoperative planning.\n\nAdopting a rectangular spiral scan instead of the conventional circular spiral3,5–10 could not only reduce the dead space when scanning long body parts but also shorten the scanning time.\n\nNonetheless, because the scanning time varies depending on the size of the area when applied for routine clinical diagnosis, physicians may need to carefully determine the minimum required scanning range in consideration of the state of the patient.\n\nThe PAI-05 system is limited to four scanning modes, as shown in Table 1; however, if it becomes possible to freely select the region of interest (ROI) according to the shape of the subject, to the scanning time could be further shortened.\n\nThere were body parts that could not be imaged as they were protruding from the water in the holding tray. This may be a large problem to be solved, especially in the context of scanning elderly patients or areas that require an awkward posture. Another approach for acoustic matching between the subject and the holding tray, such as applying an acoustic matching gel15, may have to be considered.\n\nThe ability to arbitrarily select the wavelength may contribute to improving the quantitative analysis of SO225,26. This capability could also enable the imaging of externally added dyes, such as indocyanine green (ICG), and thus enhance the applicability of PAT27. There are still no useful clinical application based on the clinical evidence of real-time 3D PAI, but new clinical applications of our PAI system will be developed and are expected to be proven in the near future. For example, diagnostic imaging using molecular probes is currently under development for future applications. Additionally, high-definition 3D imaging modes could potentially clarify drug delivery characteristics with high temporal resolution. The potential applications of the PAI-05 system reported here could be considered great advances in PAT technology.\n\nClinical research in patients will be conducted in the future. A high-quality 3D motion picture (so-called 4D imaging) with submillimeter resolution may be difficult even with CT or MRI, and it is expected that new clinical findings will be obtained with the PAI-05 system. This system is promising for obtaining a wide range of high-definition images not only for the preoperative planning of free flap surgery but also for the PAT-based diagnosis of breast cancer, as previously reported10.\n\nIn summary, we developed a new system, named PAI-05, dedicated to limb imaging and showed its properties in phantom experiments. We also imaged the limbs of a living subject in a wide range using the PAI-05 system. As with the conventional prototype, a high-resolution arteriovenous image was obtained by label-free imaging. Real-time motion pictures in an area with a diameter of 20 mm could be obtained. We expect to develop new clinical applications for the new PAT system.\n\n\nData availability\n\nF1000Research: Dataset 1. Zip file containing the underlying data of the presented results in excel files, https://doi.org/10.5256/f1000research.16743.d22486928\n\nDescription of content\n\nFigure 3 – Laser irradiation profile data (Fig3_LaserProfile)\n\nFigure 6c - Line profile of photoacoustic (PA) intensity data (Fig6c_LSF)\n\nFigure 6e – Contrast profile data (Fig6e_ContrastProfile)\n\nFigure 6g – Photoacoustic (PA) intensity for each phantom with different angles (Fig6g_oblique)\n\nFigure 7e – Movie photoacoustic (PA) intensity data (Fig7e_MoviePAIntensity)\n\nFigure 8a - Reflectance spectra data for oil ink on white office paper (Fig8a_Spectrum_OilPen)\n\nFigure 8b – Photoacoustic (PA) signal intensity data for oil ink on white office paper (Fig8b_PAoil-ink)\n\nFigure 12 – Signal intensity data from clinical samples (Fig12_ClinicalSignalIntensity)",
"appendix": "Grant information\n\nThis work was funded by the ImPACT Program of the Council for Science, Technology, and Innovation (Cabinet Office, Government of Japan).\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 thank T. Kosaka and K. Kobayashi for coordinating the activities for this clinical research.\n\n\nSupplementary materials\n\nSupplementary Video 1: Motion picture showing the PAI-05 system configuration and the principle of data acquisition.\n\nClick here to access the data\n\nSupplementary Video 2: Motion picture showing the scan sequence of the still mode of the PAI-05 system.\n\nClick here to access the data\n\nSupplementary Video 3: Motion picture showing the scan sequence of the fluctuation movie mode of the PAI-05 system.\n\nClick here to access the data\n\nSupplementary Video 4: Motion picture of a hand being swung left and right in the horizontal plane.\n\nClick here to access the data\n\nSupplementary Video 5: Motion picture of blood flow changing when a fingertip is pressed against the tray and then released.\n\nClick here to access the data\n\n\nReferences\n\nJain RK, Carmeliet PF: Vessels of death or life. Sci Am. 2001; 285(6): 38–45. PubMed Abstract | Publisher Full Text\n\nWang LV, Hu S: Photoacoustic tomography: in vivo imaging from organelles to organs. Science. 2012; 335(6075): 1458–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKruger RA, Kuzmiak CM, Lam RB, et al.: Dedicated 3D photoacoustic breast imaging. Med Phys. 2013; 40(11): 113301. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeán-Ben XL, Razansky D: Functional optoacoustic human angiography with handheld video rate three dimensional scanner. Photoacoustics. 2013; 1(3–4): 68–73. PubMed Abstract | Publisher Full Text | Free Full Text\n\nToi M, Asao Y, Matsumoto Y, et al.: Visualization of tumor-related blood vessels in human breast by photoacoustic imaging system with a hemispherical detector array. Sci Rep. 2017; 7: 41970. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMatsumoto Y, Asao Y, Yoshikawa A, et al.: Label-free photoacoustic imaging of human palmar vessels: a structural morphological analysis. Sci Rep. 2018; 8(1): 786. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTsuge I, Saito S, Sekiguchi H, et al.: Photoacoustic Tomography Shows the Branching Pattern of Anterolateral Thigh Perforators In Vivo. Plast Reconstr Surg. 2018; 141(5): 1288–1292. PubMed Abstract | Publisher Full Text\n\nIshida Y, Otsuka A, Honda T, et al.: Photoacoustic imaging system visualizes restoration of peripheral oxygenation in psoriatic lesions. J Eur Acad Dermatol Venereol. 2018; 15032. PubMed Abstract | Publisher Full Text\n\nShiina T, Toi M, Yagi T: Development and clinical translation of photoacoustic mammography. Biomed Eng Lett. 2018; 8(2): 157–165. Publisher Full Text\n\nMatsumoto Y, Asao Y, Sekiguchi H, et al.: Visualising peripheral arterioles and venules through high-resolution and large-area photoacoustic imaging. Sci Rep. 2018; 8(1): 14930. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXu Y, Wang LV, Ambartsoumian G, et al.: Reconstructions in limited-view thermoacoustic tomography. Med Phys. 2004; 31(4): 724–33. PubMed Abstract | Publisher Full Text\n\nHeijblom M, Piras D, van den Engh FM, et al.: The state of the art in breast imaging using the Twente Photoacoustic Mammoscope: results from 31 measurements on malignancies. Eur Radiol. 2016; 26(11): 3874–3887. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKitai T, Torii M, Sugie T, et al.: Photoacoustic mammography: initial clinical results. Breast Cancer. 2014; 21(2): 146–53. PubMed Abstract | Publisher Full Text\n\nFakhrejahani E, Torii M, Kitai T, et al.: Clinical Report on the First Prototype of a Photoacoustic Tomography System with Dual Illumination for Breast Cancer Imaging. PLoS One. 2015; 10(10): e0139113. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAsao Y, Hashizume Y, Suita T, et al.: Photoacoustic mammography capable of simultaneously acquiring photoacoustic and ultrasound images. J Biomed Opt. 2016; 21(11): 116009. PubMed Abstract | Publisher Full Text\n\nHoriguchi A, Shinchi M, Nakamura A, et al.: Pilot Study of Prostate Cancer Angiogenesis Imaging Using a Photoacoustic Imaging System. Urology. 2017; 108: 212–219. PubMed Abstract | Publisher Full Text\n\nKim CH, Erpelding TN, Maslov K, et al.: Handheld array-based photoacoustic probe for guiding needle biopsy of sentinel lymph nodes. J Biomed Opt. 2010; 15(4): 046010. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeán-Ben XL, Bay E, Razansky D, et al.: Functional optoacoustic imaging of moving objects using microsecond-delay acquisition of multispectral three-dimensional tomographic data. Sci Rep. 2014; 4: 5878. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNakahata K, Tokumasua S, Sakaib A, et al.: Ultrasonic imaging using signal post-processing for a flexible array transducer. NDT E Int. 2016; 82: 13–25. Publisher Full Text\n\nSwinbank F, Purser RJ: Fibonacci grids: A novel approach to global modelling. Q J R Meteorol Soc. 2006; 132(619): 1769–1793. Publisher Full Text\n\nXu M, Wang LV: Universal back-projection algorithm for photoacoustic computed tomography. Phys Rev E Stat Nonlin Soft Matter Phys. 2005; 71(1 Pt 2): 016706. PubMed Abstract | Publisher Full Text\n\nSekiguchi HTK: Development of the Rapid MIP Viewer for PAT data -KURUMI: Kyoto University Rapid and Universal MIP Imager. IEICE Tech Report, Med Imaging. 2017; 116: 163.\n\nSekiguchi H, Yoshikawa A, Matsumoto Y, et al.: Body surface detection method for photoacoustic image data using cloth-simulation technique. Proc SPIE. 2018; 1049459. Publisher Full Text\n\nLin L, Hu P, Shi J, et al.: Single-breath-hold photoacoustic computed tomography of the breast. Nat Commun. 2018; 9(1): 2352. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeán-Ben, XL, Deliolanis NC, Ntziachristosa V, et al.: Fast unmixing of multispectral optoacoustic data with vertex component analysis. Opt Laser Eng. 2014; 58: 119–125. Publisher Full Text\n\nTzoumas S, Nunes A, Olefir I, et al.: Eigenspectra optoacoustic tomography achieves quantitative blood oxygenation imaging deep in tissues. Nat Commun. 2016; 7: 12121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartel C, Yao J, Huang CH, et al.: Photoacoustic lymphatic imaging with high spatial-temporal resolution. J Biomed Opt. 2014; 19(11): 116009. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNagae K, Asao Y, Sudo Y, et al.: Dataset 1 in: Real-time 3D Photoacoustic Visualization System with a Wide Field of View for Imaging Human Limbs. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16743.d224869"
}
|
[
{
"id": "41068",
"date": "06 Dec 2018",
"name": "Jun Xia",
"expertise": [
"Reviewer Expertise Optics",
"ultrasonics",
"and photoacoustics."
],
"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 demonstrates performance of the latest photoacoustic system from Canon Inc. The images are very impressive and definitely showcase the potential of photoacoustic imaging for clinical applications. Compared to the previous generation, major improvements in the current system include the use of a 1024-element transducer array (vs. 512 elements), implementation of dual wavelength imaging, a flat bed for limb imaging, and dynamic imaging of motion. Overall the manuscript is well written and the study is comprehensive.\nThe study could be further improved in the following aspects:\nPlease quantify the imaging depth in human. Calculation in phantom is not sufficient as the phantom composition could not precisely represent the optical attenuation of human tissue. Application of the system is unclear and needs to be clarified. The palm images could be used for biometric application as mentioned in a previous publication 1. Could the system be miniaturized for that application? Foot imaging could be used for chronic ulcer assessment. However, most of those patients are immobile and the imaging pose may not be ideal for these patients. The authors suggested ICG imaging. For that application, photoacoustic imaging will compete with fluorescence imaging, such as the SPY system from Novadaq. What would be the pros and cons of PAI in comparison to a clinical fluorescence imaging system? Please indicate the magnitude of amplification in the data acquisition system. The original system from Dr. Kruger's group used a 128-element hemispherical array 2. The system has been improved over years to 1024 elements. Is there a cut-off number, beyond which further increase in element number will not improve the image quality? After reconstruction, did the authors use any image processing techniques to enhance vascular structures?\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? Yes",
"responses": [
{
"c_id": "4402",
"date": "05 Feb 2019",
"name": "Yoshiaki Matsumoto",
"role": "Author Response",
"response": "Dear Prof. Jun Xia (Reviewer #1)The authors would like to thank Reviewer #1 for your detail comments and suggestions. It is a great honor to receive high evaluation. However, if there is concern about duplication with the results and discussion of clinical research papers planned to be submitted in the future, we cannot describe that point deeply. Therefore, there may be points that are not answered enough for comments.(1) Response: Partially agreedWe agree that the phantom's evaluation is inadequate as you pointed out, and evaluation by the human being is indispensable for showing the device performance. Nonetheless, we think that it is not good data to quantify it with just one healthy person this paper. There is no guarantee that this person is typical.The results of clinical research using the device disclosed this time will be papers in future. The depth performance of human imaging should be discussed in multiple cases among them. So, we added the following sentence to the discussion section.(Additional sentence)Clinical research in patients will be conducted in the future. The device performance for actual human subjects will be evaluated at that time.(2) Response: Agreed but not revisedThe applicability to biometric authentication has already been mentioned in reference [6], so it is omitted in this paper.In this paper, we designed the device that can measure limbs in a versatile medical usage. It is primarily targeted for medical applications and has been evaluated as useful for research on plastic surgery area using this device. Through clinical research, issues for patients have come to be seen and we are aware that the current situation is not ideal as you pointed out. Therefore, it is described as follows.“This may be a large problem to be solved, especially in the context of scanning elderly patients or areas that require an awkward posture. Another approach for acoustic matching between the subject and the holding tray, such as applying an acoustic matching gel [15], may have to be considered.”It is no doubt that it can be downsized by technical development, but we have not confirmed details, so I think that it is not appropriate to describe as a paper.Further clinical issues to be solved should be listed in individual clinical papers to be submitted in the future.We would like to maintain the text from the original paper without any additional sentence.(3) Response: Agreed but not revisedWe also started clinical studies using ICG imaging, and also presented the results of comparison between SPY (or PDE) and PAI at domestic and international conferences. It will be appeared in a clinical paper shortly. Since conflicts are concerned, we would like to avoid detailed description. (4) Response: AgreedWe added the sentence to the device configuration section.(Additional sentence)We set the amplification rate to 48 dB in this study.(5) Response: Agreed but not revisedWe calculated it and we concluded that until the densest sensor arrangement on the hemispherical array, the image quality, which corresponds the number of artifacts, monotonically improves in proportion to the number of sensors. In other words, there is no cutoff number.We did not modify the text for this comment.(6) Response: Not revisedNo special image processing has been done. We made no additional sentence.Best regards,Kenichi Nagae and Yasufumi Asao"
}
]
},
{
"id": "41248",
"date": "28 Dec 2018",
"name": "Hideyuki Hasegawa",
"expertise": [
"Reviewer Expertise Medical ultrasonics"
],
"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 describes a state-of-the-art system, which is capable of real-time photoacoustic imaging of a living human body. The authors evaluated fundamental characteristics of the system, such as spatial resolution and angle-dependent visibility of a target, and also showed the feasibility of the system in in vivo imaging of vascular trees in a living human body. The system is able to visualize small vessels with good resolution and contrast. The contents of the paper would be of interest for the readers. Upon indexing, I suggest revisions described below:\n\nThere are number of referenced papers published from authors' group. On the other hand, the papers from other research groups seem few. Are there no more related works from other research groups?\n\nAlthough the method for image reconstruction is described in [21], it would be better to describe how to do that briefly. Particularly, the diverged laser light has a significant spatial inhomogeneity as shown in Fig. 3. Can you obtain an image by one shot of laser pulse with such a spatial inhomogeneity? Using such a diverging light source, photoacoustic signals would also be spatially inhomogeneous. Do you need any correction compensating for such a spatial inhomogeneity in the light intensity?\n\nIt would be more informative if you describe the reason for the use of fluctuation of the light source in the fluctuation mode (FM). Does the fluctuation change the spatial distribution of artifacts? If so, it is understandable that image quality is improved by averaging.\n\nThe size of an ultrasonic transducer element is relatively large (2 mm). I understand that the sensitivity of each transducer element should be high, but are there any effect of the element directivity on image quality, such as artifact generation?\n\nIn p. 7 and l. 2 in the right column, epsilon_Delta Hb should be epsilon_Hb?\n\nIn p. 7 and l. 3 in the right column, I found an expression \"epsilon_Delta Hb and is\". \"and\" should be removed?\n\nAs explained in the paper, the scan time increases with widening the field of view. Do you need any motion compensation technique because the spatial resolution is very high?\n\nIn the case of alternative laser irradiation, are there any effects of movements of blood cells during the repetition period of 50 ms? In the in vivo experimental results, relatively large vessels are also visualized, and movements of blood cells in such a relatively large vessel would be significant owing to relatively high flow velocity.\n\nIn Fig. 6b, I can see some artifacts. Such a distribution of the artifacts (sidelobes?) depends on the spatial arrangement of array elements based on the Fibonacci grid? It would be better to describe why artifacts can be suppressed by the Fibonacci grid and how large the sidelobe level is when a regular grid is used.\n\nParticularly in Fig. 6e, I can see remarkable range (y-direction) sidelobes. What is the source of those sidelobes? Also, such sidelobe components do not affect the estimation of oxygen saturation?\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?\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? Yes",
"responses": [
{
"c_id": "4403",
"date": "05 Feb 2019",
"name": "Yoshiaki Matsumoto",
"role": "Author Response",
"response": "Dear Prof. Hideyuki Hasegawa (Reviewer #2)The authors would like to thank Reviewer #2 for your decision of “Approved\". In order to aim more high-quality paper, we responded to the comments of Reviewer # 2 and made corrections as necessary(1) Response: Not revisedAs well-known, there are many papers on PAT, but there are not many papers as for PAT with HDA, especially for actual clinical use. Although there are various system configurations such as linear probe method, method using Fabry Perot, ring-shaped method and so on, there are too many to describe all other methods in a well-balanced manner. For this reason, we focused on HDA-PAT system and its clinical use. Based on this policy, we did not add the other references.(2) Response: AgreedYour point is right. When calculating the absorption coefficient from the initial sound pressure, it is necessary to correct the spatial non-uniformity of the light intensity including the in-vivo depth direction.On the other hand, if only image evaluation is performed qualitatively, it is not always necessary to correct the light intensity. This is because, in the Still mode, spatial nonuniformity of light intensity is corrected by averaging images obtained by scanning the hemispherical sensor in a wide range.In the Movie mode, since the light nonuniformity cannot be compensated, the signal may become dark around the image. But we can evaluate the image qualitatively. We revised the following sentences.(Before)We used universal back-projection (UBP) [21] for PA image reconstruction.(After)We used universal back-projection (UBP) [21] for PA image reconstruction. When calculating the absorption coefficient from the initial photoacoustic pressure, it is necessary to perform the light intensity correction. In this paper, however, we did not carry out it and limited the discussion within the qualitative consideration.(3) Response: Agreed Your point is right. We revised the following sentences.(After)The spatial distribution of artifacts is changed by shifting the relative position of the HDA with respect to the absorber. Artifacts in FM can be suppressed by averaging the various spatial distributions.(4) Response: Not revisedWe are reconstructing a spatial region with high receiving sensitivity of photoacoustic waves specified by element size and element placement. Since the image reconstruction range is limited, we have confirmed the influence of artifacts is not large.On the other hand, if the element size is smaller, image reconstruction in a wider spatial region becomes possible. However, it is necessary to expand the light irradiation region. In that case, there is a possibility that the artifact increases due to increase of artifact sources. We made no additional sentence.(5) Response: Agreed Your point is right. We revised the following sentences.(Before)ε∆Hb is the molar extinction coefficient of deoxyhemoglobin;(After)εHb is the molar extinction coefficient of deoxyhemoglobin;(6) Response: Agreed Your point is right. We commented on the docx file.(7) Response: Agreed but no revisedYes. We consider that the motion compensation technique can derive a better image, but it is not utilized in this paper.We can do it as written in the previous article, so we will adopt it as necessary in future clinical research. We made no additional sentence.(8) Response: Not revised With the spatial resolution of the current device, we cannot discriminate red blood cells individually, so we believe there is no effect of blood flow. We made no additional sentence.(9) Response: Not revisedThe artifacts seen in Fig. 6 are generated not by arrangement of the elements but by the discontinuation of the detectors around edges of the HDA.The effectiveness of adopting Fibonacci grid is to suppress the artifacts generated in a specific direction since the detectable area of each element is uniform. We made no additional sentence and quantitative arguments. (10) Response: Not revised It is also an artifact due to the discontinuation of the detectors around edges of the HDA as same as the (9). Oxygen saturation of blood vessels overlaid on this artifact may be affected by artifacts. We made no additional sentence.Best regards,Kenichi Nagae and Yasufumi Asao"
}
]
}
] | 1
|
https://f1000research.com/articles/7-1813
|
https://f1000research.com/articles/8-164/v1
|
06 Feb 19
|
{
"type": "Correspondence",
"title": "Sensory-specific predictive models in the human anterior insula",
"authors": [
"Gil Sharvit",
"Patrik Vuilleumier",
"Corrado Corradi-Dell'Acqua",
"Gil Sharvit",
"Patrik Vuilleumier"
],
"abstract": "Expectations affect the subjective experience of pain by increasing sensitivity to noxious events, an effect underlain by brain regions such as the insula. However, it has been debated whether these neural processes operate on pain-specific information or on more general signals encoding expectation of unpleasant events. To dissociate these possibilities, two independent studies (Sharvit et al., 2018, Pain; Fazeli and Büchel, 2018, J. Neurosci) implemented a cross-modal expectancy paradigm, testing whether responses to pain could also be modulated by the expectation of similarly unpleasant, but painless, events. Despite their differences, the two studies report remarkably convergent (and in some cases complementary) findings. First, the middle-anterior insula response to noxious stimuli is modulated only by expectancy of pain but not of painless adverse events, suggesting coding of pain-specific information. Second, sub-portions of the middle-anterior insula mediate different aspects of pain predictive coding, related to expectancy and prediction error. Third, complementary expectancy effects are also observed for other negative experiences (i.e., disgust), suggesting that the insular cortex holds prospective models of a wide range of events concerning their sensory-specific features. Taken together, these studies have strong theoretical implications on the functional properties of the insular cortex.",
"keywords": [
"Pain",
"Expectancy",
"Nocebo",
"Bayesian Coding",
"Unpleasantness"
],
"content": "\n\nOne of the most striking breakthroughs in pain research has been the discovery of expectancy modulations, according to which subjective experiences do not only reflect nociceptive input but also individuals’ previous knowledge and beliefs1. Expectancy modulations are noteworthy for their clinical implications, as convincing individuals of the effectiveness of an analgesic might induce a strong pain relief (placebo effect), sometimes comparable to the effects of active agents2. Furthermore, expectancy effects have sparkled a major theoretical debate, with influential models suggesting that pain symptoms might be better explained through a Bayesian framework, where the brain estimates the (posterior) probability of body damage, based on the integration of sensory inputs and prior representations3–6.\n\nMany studies investigated the neural structures underlying expectancy modulations of pain, pointing to an extensive network including, among other regions, the insular cortex1,7–10. In particular, whereas the posterior portion of the insula is known to receive thalamic nociceptive projections11–13 and thought to process bottom-up components of the painful experience8, the middle-anterior portions may integrate such bottom-up signals with prior expectations7,8, and generate prediction-error signals, serving to update the representation of future events8. However, the insular cortex (like other interconnected regions such as the cingulate cortex) does not respond to pain specifically, but also to a wide range of aversive events14, including disgust15,16, negatively-valenced pictures17,18, or even unfairness15,19,20. Accordingly, a part of pain-evoked activity in this region might reflect supramodal dimensions of affect or motivation, such as unpleasantness15, arousal or even salience21,22. This raises the question about the nature of the predictive information encoded on the middle-anterior insula, and whether it relates to pain-specifically (“this will hurt”), or rather to an undistinctive negative event (“this will be bad”).\n\nAddressing this issue is not a trivial matter, as it would require testing whether pain-evoked activity in the middle-anterior insula is also sensitive to the expectation of a painless event of same unpleasantness or salience. Interestingly, two recent independent studies (each unbeknownst to the other) did precisely this, reaching remarkably similar results23,24. The first study from Sharvit and others23 compared the expectancy of pain with that of a disgusting odorant of similar unpleasantness (see also Sharvit and others25 for an earlier behavioral implementation of the task), whereas the second from Fazeli and Büchel24 used as control pictures of aversive content. By expanding on well-known paradigms of pain expectancy7,8, both studies were able to replicate evidence that the middle-anterior insula integrated bottom-up nociceptive information with signals from predictive cues, but this did not occur when cues were incongruent with the subsequent event (e.g., disgust/image cues followed by painful stimulus)23,24. Such convergence between researches with important differences in sensory stimuli, task structures, and data analyses23,24, provides a compelling case that expectancy modulations of pain in the insular cortex are sensory-specific, and do not generalize to a broad code of unpleasantness. This also accords with other work showing for shared and segregated portions in insula for representations of pain, disgust, and unfairness15.\n\nAlthough sharing a similar take-home message, the two cross-modal experiments by Sharvit and Fazeli differ (and in some case complement each other) concerning the information coded by the middle-anterior insula. By employing rigorous Bayesian modelling, Fazeli and Büchel24 dissociated a portion in the middle and dorsal-anterior portion of the insula, responsible for integrating bottom-up signals with prior expectancies, from a portion in ventral-anterior insula, responsible for generating error signals whenever the painful stimulus greatly diverged from what was predicted by the cues (see Figure 1). This was not the case in Sharvit and others23 who adopted a paradigm where divergences between cues and subsequent stimuli were purposefully subtle to pass unnoticed7,25. It is interestingly to notice, however, that Sharvit and others23 reported a dissociation between the middle insula, exerting a mediatory role in the way in which predictive cues influenced subjective reports (as previously found7), and the most anterior insula, exerting instead an opposite role of suppression. Hence, in Sharvit and others23 activity in the anterior insula seemed to prevent individuals from being influenced by their expectations, an effect that is consistent with the notion of prediction-error modeled by Fazeli and Büchel24. The two studies also differ regarding the insular sub-sections involved: Sharvit and others23 mapped mediation and suppression effects along the middle-to-anterior axis, whereas Fazeli and Büchel24 described expectancy and prediction-error effects also along the dorsal-to-ventral axis of the anterior insula (Figure 1). Future studies will need to further clarify how different components of expectancy relate to the various insula portions.\n\nBlue shades over the Posterior Insula (PI) refer to processing of pain based solely on nociceptive inputs. Orange shades over the Middle (MI) and Anterior Insula (AI), refer to processing of pain (and disgust23) based also on prior expectations. Green shades in AI refer to regions coding prediction errors24, and suppressing the effect of previous expectations23.\n\nA further, and critical, point of divergence relates to whether the insular cortex is also susceptible to sensory-specific expectancy for other events than pain. This question was addressed only by Sharvit and others23 who described complementary effects to those observed in pain, also for the case of olfactory disgust. These results suggested that the middle-anterior insula may hold multiple predictive representations of upcoming events, which are then updated by bottom-up sensory input. Hence, although the middle-anterior insula appears sensitive to a wide range of stimuli14, it may retain sensory-specific information about each of them. Anatomical studies on primates subfields in this region26, with a level of detail that exceeds that derived from neuroimaging research in humans27,28. It is therefore foreseeable that different kinds of sensory events might be represented in the anterior insula through neighbouring, but distinct, neuronal populations, which could be difficult to distinguish through radiological imaging, but nonetheless selectively dissociated through well-crafted expectancy manipulations.\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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: CCD is supported by the Swiss National Science Foundation [PP00O1_157424/1].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nAtlas LY, Wager TD: How expectations shape pain. Neurosci Lett. 2012; 520(2): 140–8. PubMed Abstract | Publisher Full Text\n\nVase L, Amanzio M, Price DD: Nocebo vs. placebo: the challenges of trial design in analgesia research. Clin Pharmacol Ther. 2015; 97(2): 143–50. PubMed Abstract | Publisher Full Text\n\nMorrison I, Perini I, Dunham J: Facets and mechanisms of adaptive pain behavior: predictive regulation and action. Front Hum Neurosci. 2013; 7: 755. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTabor A, Thacker MA, Moseley GL, et al.: Pain: A Statistical Account. PLoS Comput Biol. 2017; 13(1): e1005142. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOngaro G, Kaptchuk TJ: Symptom perception, placebo effects, and the Bayesian brain. Pain. 2019; 160(1): 1–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWiech K: Deconstructing the sensation of pain: The influence of cognitive processes on pain perception. Science. 2016; 354(6312): 584–7. PubMed Abstract | Publisher Full Text\n\nAtlas LY, Bolger N, Lindquist MA, et al.: Brain mediators of predictive cue effects on perceived pain. J Neurosci. 2010; 30(39): 12964–77. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGeuter S, Boll S, Eippert F, et al.: Functional dissociation of stimulus intensity encoding and predictive coding of pain in the insula. eLife. 2017; 6: pii: e24770. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKoyama T, McHaffie JG, Laurienti PJ, et al.: The subjective experience of pain: where expectations become reality. Proc Natl Acad Sci U S A. 2005; 102(36): 12950–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWiech K, Lin CS, Brodersen KH, et al.: Anterior insula integrates information about salience into perceptual decisions about pain. J Neurosci. 2010; 30(48): 16324–31. PubMed Abstract | Publisher Full Text\n\nCraig AD: Pain mechanisms: labeled lines versus convergence in central processing. Annu Rev Neurosci. 2003; 26(1): 1–30. PubMed Abstract | Publisher Full Text\n\nCraig AD, Zhang ET: Retrograde analyses of spinothalamic projections in the macaque monkey: input to posterolateral thalamus. J Comp Neurol. 2006; 499(6): 953–64. PubMed Abstract | Publisher Full Text\n\nCraig AD, Chen K, Bandy D, et al.: Thermosensory activation of insular cortex. Nat Neurosci. 2000; 3(2): 184–90. PubMed Abstract | Publisher Full Text\n\nKurth F, Zilles K, Fox PT, et al.: A link between the systems: functional differentiation and integration within the human insula revealed by meta-analysis. Brain Struct Funct. 2010; 214(5–6): 519–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCorradi-Dell’Acqua C, Tusche A, Vuilleumier P, et al.: Cross-modal representations of first-hand and vicarious pain, disgust and fairness in insular and cingulate cortex. Nat Commun. 2016; 7: 10904. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWicker B, Keysers C, Plailly J, et al.: Both of us disgusted in My insula: the common neural basis of seeing and feeling disgust. Neuron. 2003; 40(3): 655–64. PubMed Abstract | Publisher Full Text\n\nChang LJ, Gianaros PJ, Manuck SB, et al.: A Sensitive and Specific Neural Signature for Picture-Induced Negative Affect. PLoS Biol. 2015; 13(6): e1002180. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCorradi-Dell’Acqua C, Hofstetter C, Vuilleumier P: Felt and seen pain evoke the same local patterns of cortical activity in insular and cingulate cortex. J Neurosci. 2011; 31(49): 17996–8006. PubMed Abstract | Publisher Full Text\n\nSanfey AG, Rilling JK, Aronson JA, et al.: The neural basis of economic decision-making in the Ultimatum Game. Science. 2003; 300(5626): 1755–8. PubMed Abstract | Publisher Full Text\n\nCorradi-Dell’Acqua C, Civai C, Rumiati RI, et al.: Disentangling self- and fairness-related neural mechanisms involved in the ultimatum game: an fMRI study. Soc Cogn Affect Neurosci. 2013; 8(4): 424–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLegrain V, Iannetti GD, Plaghki L, et al.: The pain matrix reloaded: a salience detection system for the body. Prog Neurobiol. 2011; 93(1): 111–24. PubMed Abstract | Publisher Full Text\n\nMouraux A, Diukova A, Lee MC, et al.: A multisensory investigation of the functional significance of the “pain matrix”. NeuroImage. 2011; 54(3): 2237–49. PubMed Abstract | Publisher Full Text\n\nSharvit G, Corradi-Dell’Acqua C, Vuilleumier P: Modality-specific effects of aversive expectancy in the anterior insula and medial prefrontal cortex. Pain. 2018; 159(8): 1529–1542. PubMed Abstract | Publisher Full Text\n\nFazeli S, Büchel C: Pain-Related Expectation and Prediction Error Signals in the Anterior Insula Are Not Related to Aversiveness. J Neurosci. 2018; 38(29): 6461–74. PubMed Abstract | Publisher Full Text\n\nSharvit G, Vuilleumier P, Delplanque S, et al.: Cross-modal and modality-specific expectancy effects between pain and disgust. Sci Rep. 2015; 5: 17487. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEvrard HC, Logothetis NK, Craig AD: Modular architectonic organization of the insula in the macaque monkey. J Comp Neurol. 2014; 522(1): 64–97. PubMed Abstract | Publisher Full Text\n\nFan L, Li H, Zhuo J, et al.: The Human Brainnetome Atlas: A New Brain Atlas Based on Connectional Architecture. Cereb Cortex. 2016; 26(8): 3508–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKelly C, Toro R, Di Martino A, et al.: A convergent functional architecture of the insula emerges across imaging modalities. NeuroImage. 2012; 61(4): 1129–42. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "44134",
"date": "01 Mar 2019",
"name": "Lauren Y. Atlas",
"expertise": [
"Reviewer Expertise Pain",
"Expectancy",
"FMRI",
"Placebo",
"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\nIn this paper, Sharvit, Vuilleumier and Corradi-Dell’Acqua draw on evidence from two recent studies (Sharvit et al. (20181) and Fazeli and Büchel (20182)) to suggest that the middle-anterior part of the insula cortex encodes sensory-specific expectancy effects of pain while the posterior insula process nociceptive inputs. Both investigations employ within-subjects designs to examine whether the brain shows sensory-specific expectancy effects. Cues predict the intensity or unpleasantness of pain or a different modality (olfactory or aversive images), and cues about one modality are occasionally followed by the other outcome. The authors of each paper make conclusions about cross-modal versus sensory-specific effects, which are summarized in the current review with a focus on effects within the insula. The paper is well-written and covers important issues regarding whether the insula processes expectation of general unpleasantness or pain-specific expectation. This question is timely and relevant for a large body of work asking whether brain responses are specific to pain or whether responses in regions like the insula reflect domain-general processes, such as salience or unpleasantness (Legrain et al. (20113). The review is a useful addition to the literature on pain and sensory specificity. However, certain issues that should be addressed in this review to benefit future investigations on the question of expectancy effects and modality specificity.\nIn our opinion, the authors need to be more cautious in making conclusions regarding sensory specificity since neither study included comprehensive tests for specificity. First, they did not evaluate potentially confounding factors between pain and non-pain stimuli such as stimulus-related salience, arousal, uncertainty (the degree of learned association between cue and outcome), and threat/aversiveness. For example, a modality-specific expectancy effect in the anterior insula might be influenced by the lower level of uncertainty in the congruent conditions compared to incongruent conditions. The insula is also a core region of the salience network and affected by the salience of the stimulus (Menon and Uddin (20104)) as well as threat value (Wiech et al. (20105)) which might differ between pain and the other modality (i.e. pain has the potential to cause harm, whereas odors do not). Since these factors were not measured or controlled in either study, it should be noted that those factors might contribute to the effect of expectancy in the insula as well.\nSecond, Fazeli and Büchel (20182) did not evaluate sensory-specific or sensory-general processing of expectancy in response to visual stimuli paired with either pain or visual cues. If the effect of expectancy on insula activity is truly sensory-specific and does not affect the expectancy effect on a stimulus of another modality, the same phenomenon should be found in the non-pain modality. Finally, Sharvit et al. (20181) trained with outcome-specific cues during conditioning, and tested responses to crossed cues only in the test phase. This may have led to differences in the cross-modal cue pairs, only because those were novel in the test phase, which is also related to the issue of uncertainty mentioned above.\n\nFinally, we think the authors should also consider that within-subjects tests of sensory specificity may lead to very different conclusions from between-subjects designs, wherein subjects would be randomly assigned to experience a single outcome modality and modalities would vary across groups. When individuals experience multiple outcomes, this engenders both value-based learning (expectations about intensity/unpleasantness) and sensory learning (expectations about outcome identity). Some regions respond similarly to both types of learning (e.g. dopaminergic neurons involved in value-based prediction error respond when outcome identity changes and value is held constant (Takahashi et al. (20176) and Chang et al. (20177)), whereas other brain regions respond to value irrespective of outcome identity (OFC; Padoa-Schioppa et al. (20088)) or sensory outcome irrespective of value (lateral OFC; Boorman et al. (20169)). In the two papers reviewed here, cues denote both outcome type (i.e. pain vs. odor/image) and outcome value (i.e. unpleasantness/intensity), which requires both types of learning. While this is theoretically interesting, the question Sharvit and co-authors focus on here is whether value-based learning (i.e. expectancy-based modulation) is sensory specific. A purer test of this question would be to use a between-subjects design to compare value-based processing across modalities, i.e. whether brain responses are similar when pain is preceded by pain expectancy cues and when olfactory/visual outcomes are paired with expectancy cues in those domains. When both forms of learning are combined, differences might emerge that reflect sensory learning, rather than purely testing whether value-learning is sensory-specific. The authors should acknowledge this alternative in the present review.\n\nIn addition to the two conceptual issues above, we have several minor suggestions:\nSeveral terms should be defined or explained more precisely (either in a box or early in the manuscript). This includes: “sensory specificity”; the distinction between ‘unpleasant events’, ‘negative experiences’, and ‘threat’; and the relationship between Sharvit’s manipulation of ‘unpleasantness’ (Sharvit et al. (20181)) and Fazeli’s manipulation of ‘intensity’ (Fazeli and Büchel (20182)).\n\nThe manuscript would benefit from more description of the two original study paradigms and how they differ (e.g. conditioning, instructions, test phases).\n\nThe review focuses on findings in the insula, yet both papers also found important effects outside of this region (Sharvit et al. (20181) found interesting and similar cue predictive effects in vmPFC and TPJ, and Fazeli and Büchel (Fazeli and Büchel (20182)) showed intensity and expectancy effects in ACC in addition to the insula). Since insula has a strong relationship (functional and anatomical) with those regions in pain processing and modulation, the authors should acknowledge that other regions show similar effects and also discuss differences.\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? Partly\n\nIs the conclusion balanced and justified on the basis of the presented arguments? Partly",
"responses": []
},
{
"id": "44441",
"date": "01 Mar 2019",
"name": "Daniele Romano",
"expertise": [
"Reviewer Expertise cognitive neuroscience",
"experimental psychology and neuropsychology. One focus of my research is how body representation and multisensory integration influence the pain perception."
],
"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 by Sharvit and colleagues is an interesting commentary paper mostly based on two studies aiming at exploring the role of the insular cortex in integrating sensory input and internal predicting models to formulate the perception of an aversive stimulus (Fazeli and Büchel (20181) and Sharvit et al. (20182)). Sharvit and co-workers also wrote one of the two papers on which this commentary is based.\nThe two original studies focused on comparing the response to painful stimuli to another negative stimulation. In one case a disgusting olfactory stimulation was used, while in the other, a set of negative-valence pictures from the IAPS database were used.\nThe commentary paper nicely points out the convergences of the two studies and their potential limitations. It also importantly noticed the differences in the two findings discussing their convergences and the complementary nature of the two. I have no significant concerns about the commentary paper, and I believe that this offers an interesting supplement to the original discussion of Sharvit and colleagues’ work.\nI only wish to point out a few comments to further develop and hopefully stimulate the discussion about the potential role of the insula in predictive coding.\nIt is known that it can be difficult to match the saliency of negative stimuli with the saliency of positive stimuli. The former are typically more arousing than the latter (Ferri et al. (20153)). However, when this aspect is taken into consideration carefully, results (at least at the behavioural level) show that saliency matters more than valence (Spaccasassi et al. (20194)). Although it is clearly the aim of the two original studies, is it possible that the insula contributes to the predictive coding of every salient stimulus? In other words, is it possible that the role of the insula is not limited to the two hypotheses “this will hurt” or “this will be bad”, but may extend to “this will be arousing”?\n\nRelatedly, an even more extreme hypothesis could be made. The insula might be a sort of bottle-neck for the integration of internal models with sensory stimuli. Returning to the previous example, it is possible that the insula processing is the following: “are the sensory information and the internal models on the same page? Can I integrate the two sources of information?” The middle and anterior insula might be involved in the processing of this very general aspect. Sharvit and Fazeli papers mostly work in the prediction of stimuli of different sensory modalities with a similar intensity and the same negative valence, revealing a crucial role of the middle and anterior insula in coding prediction errors across sensory modalities. Potentially, it is possible to create a prediction incompatible with the incoming stimulus, so that it won’t be integrated, within a sensory modality. For example, the prediction of a very low painful stimulus coupled with a very high level of painful stimuli may generate the same prediction error signal that leads to the non-integration of the internal/external signals when coming from different sensory modalities. Fazeli and Büchel (20181) already proposed this point as a limitation of their study in their discussion.\n\nThere is a debate that social exclusion may generate what is called “social pain”, which would have a brain footprint largely overlapping with physical pain (Eisenberger and Lieberman (20045)). I am not personally a great sponsor of this “overlap theory”, the difference between the two experiences is immense perceptually so that the overlap between the two functions should be taken cautiously. Besides my personal concerns, one hypothesis is that with current experimental procedures we capture effects that are shared by all the arousing stimuli and not limited to pain (Legrain et al. (20116)). Do the authors predict that the role of the insula in integrating expectations and sensory input would somehow work also for social pain? Alternatively, can the sensory-specific pattern of response for expectations, discussed in this commentary, help to distinguish the processing of physical pain from social pain?\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": []
}
] | 1
|
https://f1000research.com/articles/8-164
|
https://f1000research.com/articles/8-162/v1
|
06 Feb 19
|
{
"type": "Research Article",
"title": "MUC1 and E-cadherin immunohistochemistry of endometrium cannot predict the outcome of in vitro fertilization: A case-control study",
"authors": [
"Saumitra Chakravarty",
"Mohammed Kamal",
"Mohammed Kamal"
],
"abstract": "Background: Although in vitro fertilization (IVF) has played a major role in the management of infertility, its failure rate is still 60-80% and most of the causes failure are unknown. Therefore, a histomorphology-based predictive tool to forecast IVF outcome that utilizes expression data of certain cellular adhesion molecules in endometrium pertaining to successful implantation might provide the theoretical basis to develop a low-cost laboratory investigation suited for low to middle income countries as opposed to the expensive gene expression based tools like endometrial receptivity array. In this study, mucin 1 (MUC1) and E-cadherin immunohistochemistry of endometrium from aspiring IVF women were analyzed to see if there is any correlation between signal intensities and endometrial receptivity in terms of IVF outcome. Methods: This was a case-control study conducted among women of reproductive age with infertility who underwent IVF at the Centre for Assisted Reproduction (CARe), Dhaka between March and December 2017. Endometrial biopsy samples were collected and routine histological as well as immunohistochemical analysis was performed on those samples at the Department of Pathology, Bangabandhu Sheikh Mujib Medical University (BSMMU), Dhaka. A total of 21 patients, 17 cases (IVF failure) and four controls (IVF success), were included in the study by consecutive convenient sampling. Relevant history and medical records of each of the patients were also obtained accordingly. Results: No statistically significant correlation was found between IVF outcomes and the signal intensities in endometrium produced by MUC1 and E-cadherin immunohistochemistry. Conclusions: Despite the fact that this study did not find any statistically significant correlation between endometrial immunohistochemistry of MUC1 and E-cadherin and IVF outcome, further studies may incorporate gene expression arrays to supplement or revise those findings.",
"keywords": [
"MUC1",
"E-cadherin",
"IVF",
"IHC"
],
"content": "Introduction\n\nAlthough birth control measures are highlighted as one of the major reproductive health concerns in third world countries, the population suffering from infertility often goes unnoticed. Globally, 12.4% of women who are trying to conceive in the age range of 20–44 years are infertile, accounting for both primary and secondary infertility; south Asia is one of regions with the highest prevalence of infertility1. Despite the fact that about 16,700 infertility cases are enrolled annually in the handful number of facilities available in Bangladesh and many more of the cases remain unreported, the medical, social, financial and psychological burden associated with the condition is almost always under-appreciated in health statistics2. There is no account of how many infertile couples seek help from fraudulent and pseudoscientific agencies in the country, which may add significantly to the burden3. The high cost of infertility treatment happens to be a major barrier for aspiring couples to seek medical help in Bangladesh2. Failure rates as high as 60%, even in the most developed countries, make the application of in vitro fertilization (IVF) and other assisted reproductive technologies an expensive gamble4. Moreover, the cause of infertility remains unexplained in 10.95% of cases2.\n\nTo find a solution to this problem, a research group in Spain analyzed mRNA from endometrial biopsy samples of expecting women and identified 238 genes that play a crucial role in implantation5. Using the results, they developed a customized mRNA array and bioinformatic tool for testing expression of those 238 genes to predict not only the possibility of implantation failure but also the personalized window of implantation (WOI) during which the chance of a successful implantation is maximum. That is called endometrial receptivity array (ERA). As good as it may sound, ERA is not without its own shortcomings. Firstly, it is quite expensive6. Secondly, availability is limited, only at a handful of institutes in Spain, Japan, the USA and a few other countries. Thirdly and most importantly, even if we can improve the odds for success by analyzing all 238 genes’ expression, the results are not translatable in terms of targeted therapy of infertility because there are simply too many targets. Despite these limitations, there has been no other tool as reliable and precise as ERA to date to predict and improve IVF outcome (i.e., success/failure)7. Therefore, efforts are being made to discover clever work-arounds to cut both the cost and number of targets with little compromise in terms of predictive values. ERA-directed proteomic studies of uterine biopsies from both groups of women (with implantation success and failure) indicate that secretory endometrium expresses certain cellular adhesion molecules that firmly attach the blastocyst at the site of implantation8. This cell-cell interaction is pivotal not only for successful implantation but also for maintenance of pregnancy. Proteomics provide a list of few pathways that have a handful of protein molecules9. These are the key players of implantation. So far, five to ten proteins, including mucin 1 (MUC1) and E-cadherin, have shown potential to constitute a panel for predictive screening of implantation failure10.\n\nThe peculiarity of MUC1 is that it acts as a ‘guiding molecule’ for the blastocyst during implantation in order to direct it to the implantation site of the decidua8. In receptive endometrium, MUC1 expression is generally high, especially in the endometrium during secretory phase, compared to non-receptive endometrium11,12. Interestingly, there is always a small patch of decidua where the receptive endometrium almost completely lacks MUC1 expression during the WOI and that marks the site of implantation. Hence the connotation of a ‘guiding molecule’, as if the blastocyst wanders about the decidua until it finds the designated spot to implant marked by the absence of MUC18,13. In case of non-receptive endometrium as in IVF failures or recurrent spontaneous abortions, global down-regulation of MUC1 throughout the whole of the endometrium fails to elicit the appropriate signal for implantation. However, it should be noted that if implantation is successful then MUC1 expression is quickly depleted and therefore it is one of the negative stains in normal placenta14.\n\nCadherins constitute a group of glycoproteins responsible for the calcium-dependent cell-cell adhesion mechanism. It has been postulated that E-cadherin exhibits dual attributes consisting of an increased expression during secretory phase and early implantation: epithelial cell adhesiveness by E-cadherin is controlled by intracellular calcium and rising progesterone levels induce calcitonin expression and thus increase the concentration of intracellular calcium, which then suppresses E-cadherin expression at cellular contact sites. It corresponds to the adhesiveness of decidua-blastocyst interface while a subsequent spell of its down-regulation corresponds to epithelial dissociation and trophoblastic invasion8. Although cyclical variation in mRNA levels of E-cadherin is not reflected at protein level as detected by IHC, there are indeed immunohistochemically detectable differences of E-cadherin expression between receptive and non-receptive endometrium15. Interestingly, progesterone level has somewhat inverse relationship with E-cadherin expression16,17. Moreover, it has been demonstrated that targeted mutations in E-cadherin gene are associated with implantation failure18.\n\nWe sought to evaluate in this study if there is any correlation between IVF outcome and semi-quantitative immunohistochemistry scores (H-score) of MUC1 and E-cadherin in endometrial epithelium and stroma.\n\n\nMethods\n\nThe present study was a case-control study. The study subjects, i.e., all female patients attending Centre for Assisted Reproduction (CARe), Dhaka, Bangladesh who fulfilled the selection criteria, including giving informed written consent to participate in the study, between 1st March and 31st December 2017 were sorted into either case or control groups according to the selection criteria. The key outcome variable that distinguished between the two groups was post-IVF pregnancy status. If the status was non-pregnant then the subject was included in the case group (n=17) and if the participant was pregnant then she was included in the control group (n=4). The sample size was estimated using Kesley’s formula for case-control study19. All other criteria being adequately fulfilled as well as normalizing for possible confounders, e.g. age, body mass index (BMI), obstetric history, menstrual history, contraceptive history, previous IVF outcome, follicular stimulating hormone (FSH) level and luteinizing hormone (LH) level.\n\nEndometrial biopsy samples collected from both groups prior to IVF procedure, preserved as formalin-fixed paraffin embedded (FFPE) blocks in the archive, were assessed both by routine histology and immunohistochemically to determine the values of key exposure variables, i.e., endometrial dating and immunohistochemical H-scores for MUC1 and E-cadherin stains. Those values were used to determine if the exposure variables correlated to the outcome variable, which can predict the outcome of IVF within statistical limits.\n\nInclusion criteria for cases\n\n1. Selected as a candidate for IVF by competent physician.\n\n2. Provided informed written consent to participate in the study.\n\n3. Underwent endometrial biopsy according to the protocol mentioned in data collection procedure.\n\n4. Post-IVF pregnancy status indicates IVF failure.\n\nInclusion criteria for controls\n\n1. Selected as a candidate for IVF by competent physician.\n\n2. Provided informed written consent to participate in the study.\n\n3. Underwent endometrial biopsy according to the protocol mentioned in data collection procedure.\n\n4. Post-IVF pregnancy status indicates IVF success.\n\nExclusion criteria for both cases and controls\n\n1. Refusal to participate in the study or withdrawal from the study at any point.\n\n2. If it is not possible or contra-indicated to obtain endometrial biopsy at all or at the designated time as mentioned in data collection procedure.\n\n3. If the performance of IVF is cancelled or postponed beyond the specified duration of study.\n\n4. If post-IVF pregnancy status could not be ascertained within the specified duration of study or the patient is dropped out from follow-up.\n\n5. Co-morbid conditions like tuberculosis, endometrial hyperplasia, HIV/AIDS, thyroid diseases, diabetes mellitus, hypertension, immunological diseases etc.\n\n6. Incomplete medical record or inadequate sample.\n\nProcedures performed at CARe. Endometrial scratching was the method used to collect endometrial biopsy sample. The procedure was performed by a competent physician on the 21st day of the natural cycle or on a day accordingly adjusted for irregular, longer or hormone-assisted cycle immediately before the commencement of IVF cycle. The samples were transferred immediately to 10% buffered formalin in a properly labeled container. The container was then kept at room temperature for 24 hours before it was transported to Bangabandhu Sheikh Mujib Medical University (BSMMU) where the laboratory procedures commenced. It was imperative not to freeze the sample because that would cause freezing artifacts rendering it unsuitable for subsequent histological procedures as well as microscopy.\n\nInformed written consent was taken prior to the biopsy procedure. Serum levels of FSH and LH were measured on day 3 of the cycle immediately prior to the IVF cycle. Relevant history, clinical information and most of the investigation findings featuring the key variables were collected by the attending physician or the primary investigator via a pre-formed questionnaire at the time of taking the biopsy. Embryo was transferred on the next suitable WOI as determined by the conventional IVF protocol, usually after 72 hours following the oocyte retrieval during which the fertilization took place in vitro.\n\nPost-IVF pregnancy check was performed by measuring serum human chorionic gonadotropin (hCG) level after two weeks of embryo transfer, taking into account the iatrogenic effect of hCG which was being administered during that period of time. The pregnancy status was recorded accordingly. The rest of the relevant investigation findings unavailable during collection of endometrial biopsy were then recorded. In the case of irregular menstrual cycle, the consultant physician determined the adjusted dates of IVF protocol according to the patient’s previous cycles over one year along with the findings from transvaginal sonography. Table 1 summarizes the IVF protocol that was followed along with the various points in data or sample collection for this study.\n\nET, embryo transfer; FSH, follicle stimulating hormone; GnRH, gonadotropin releasing hormone; hCG, human chorionic gonadotropin; IVF, in vitro fertilization; LH, luteinizing hormone.\n\nProcedures performed at BSMMU. Routine tissue processing was done followed by preparation of FFPE blocks for every sample, one for each patient. Hematoxylin-Eosin (HE) stained permanent slides were prepared from those FFPE blocks. Routine microscopy of the HE slides was performed to look for the possibility of tuberculosis, hyperplasia and malignancy as well as for the estimation of endometrial date. After receiving information regarding post-IVF pregnancy status from CARe Hospital, and also combining the information gathered from routine HE histology as well as the data collection sheet, each sample was either included in the study as a case or a control according to inclusion criteria, or excluded from the study according to exclusion criteria. After the sorting of the patients into case and control groups according to the selection criteria, a computer-aided randomized patient coding technique was applied to ensure double blind design so that neither the investigator, nor the supervisor or the personnel directly involved in the study can know which patient or slide belonged to which group until the master datasheet is prepared.\n\nFFPE blocks of cases and controls both were then used for preparation of immunohistochemistry slides, by staining each section with antibodies (Table 2). Tissue microarray technique was used for the purpose to reduce the expenditure.\n\nRTU, ready to use.\n\nAntigen retrieval was performed according to the manufacture’s manual that comes with the designated IHC markers. Endogenous peroxidase activity was blocked by 3% hydrogen peroxide for 20 minutes. Slides were then incubated with appropriate clone of designated antibodies at suitable dilutions. Then all the slides underwent the same protocol: slide reagents A (Polymer Helper) and B (polyperoxidase anti-mouse IgG) were used for the reaction and the slides were counterstained with Mayer’s hematoxylin followed by staining with 3,3’-diaminobenzidine.\n\nThe slides were then evaluated in terms of H-score by two Pathology consultants independently to rule out inter-observer variation. H-score takes into account two aspects of cellular staining by immunohistochemical markers: relative intensity (RI) and percentage of cells (%C) stained. RI is ranked 0–3 as follows: 0, no/ negative/ background staining; 1, weak positive; 2, strong positive; 3, very strong positive. In a given section, the percentage of cell (%C) that fits each of the four RI ranks are also estimated. Then the RI is multiplied by the corresponding %C value. Lastly, four of such products, one for each RI, are added up to get the H-score of that sample for a given marker9,20. Figure 1 and Table 3 show an example of the scoring system.\n\nNumbered labels are placed adjacent to the epithelia corresponding to their respective relative intensities (Table 3). Stromal cells show lack of staining.\n\nAll of the slides were scanned by Hamamatsu nanoZoomer® slide scanner to create digitally accessible virtual slides for archival purpose. Microsoft Office Excel 2010 was used to perform statistical analysis19–22. For all statistical tests, p-value <0.05 was considered as statistically significant.\n\n\nResults and discussion\n\nResults are prepared in four steps. Firstly, sorting and matching of case and control groups are done for probable confounding variables. Secondly, routine histomorphological evaluation of endometrial biopsy samples was done which includes endometrial dating. Thirdly, immunohistochemical markers in endometrial biopsy samples were evaluated with H-scores and optimum cutoffs were estimated. Lastly, correlations of immunohistochemical markers with IVF outcomes were calculated and statistical inferences were made. In summary, correlations between IVF outcome and H-scores of immunohistochemical markers MUC1 and E-cadherin are weakly positive and statistically insignificant.\n\nOut of 21 patients, 17 patients (81%) had IVF failure (not pregnant) and four patients (19%) had IVF success (pregnant). After case-control categorization, the former group was considered case and the latter group was control. All the participants were from middle class to upper middle class according to socioeconomic strata. The women of the case group had a mean age of 34.00 ± 2.31 years while the mean age of the control group was 33.75 ± 4.79 years; the difference between the two groups were not statistically significant. There were no significant difference between them in terms of BMI, FSH, LH, FSH:LH ratio, menstrual and obstetric history which were considered to be the potential confounding factors. Their husbands were all within normal range of relevant reproductive parameters. None of the patients used any contraceptive method ever, except barrier method, and actively trying to conceive for at least two years. Degree of expression of MUC1 and E-cadherin as detected by immunohistochemistry was considered to be the exposure (H-scores ranging from 0-300) and post-IVF pregnancy status was considered as the outcome (pregnant or non-pregnant) in the present study.\n\nIt is interesting to note that a peculiar staining pattern was observed with E-cadherin where the surface epithelia stained with markedly greater intensity than the glandular epithelia (Figure 2). Most of the samples exhibit similar staining pattern with E-cadherin.\n\nComparison between case and control groups by unpaired t-test show statistically non-significant difference for mean H-scores of MUC1 as well as E-cadherin although MUC1 has a larger effect size (0.91) compared to that of E-cadherin (0.69). Calculation of inter-observer variation between two independent observers (Pathology Consultants) was measured on 10 randomly chosen tissue sections out of 21 samples by Chronbach’s alpha correlation technique21. Inter-observer variation was found to be within acceptable limits for each of the stains22.\n\nDue to the variability of data obtained from two subsets of samples, namely, collected before 20th day and collected on or after 20th day, the statistical analysis for derivation of the cutoff scores were calculated separately for the two subsets on each of the markers. At every attempt to find the optimum cutoff, a receiver-operator characteristics (ROC) curve was drawn first and then its corresponding Youden’s indices (specificity + sensitivity – 1) were plotted if the ROC curve had an area under curve (AUC) greater than zero. The point with the maximum value of the Youden’s index was chosen as the optimum cutoff because it maximized the specificity and sensitivity of the marker in question23. Instances where ROC curve had an AUC of zero unit, Youden’s index could not be meaningful, thus yielding the corresponding cutoff value undefined. AUCs of ROC curves were calculated using trapezoid rule and maximum Youden’s indices were obtained by direct measurement.\n\nROC curve for MUC1 suggested the existence of optimum cutoff values for samples collected before, on or after the 20th chronological day since AUC>0 held for all the instances (Figure 3 and Figure 4). The cutoff for samples collected on or after 20th day was 120 (Figure 5) and the rest of the samples yielded a cutoff of 190 (Figure 6). MUC1 status is to be considered negative if the H-score is less than or equal to the cutoff; if it is above the cutoff then the status is positive.\n\nFurther calculations concluded that MUC1 status thus had a weak correlation with the IVF outcome, which was statistically insignificant (Figure 7; Table 4).\n\nROC curve for E-cadherin suggested the existence of optimum cutoff values for samples collected on or after the 20th chronological day but not before, since AUC>0 and AUC=0 held for those two instances, respectively (Figure 8 and Figure 9). The cutoff for samples collected on or after 20th day was 75 (Figure 10). E-Cadherin status is to be considered negative if the H-score is less than or equal to the cutoff; if it is above the cutoff then the status is positive. Subsequent calculations determined that E-cadherin status had a weak correlation with the IVF outcome, which was statistically insignificant (Figure 11; Table 4).\n\nVarious confounding variables were matched between case and control groups in the present study24,25. Age and BMI are two such anthropometric variables, which were matched in similar studies where comparison was being done between receptive and non-receptive endometrium9. However, ratio of sample size between case and control groups in the present study is about 5:1 which is not shown with most of the studies with similar design where the ratio is almost 1:1. This deviation is in part due to relatively lower IVF success rates that ultimately resulted in fewer number of successes accomplished within the study period. Nevertheless, the imbalance between the two groups in the present study was taken into account statistically by applying adjustments for unequal variance wherever appropriate.\n\nAlthough endometrial dating by histomorphological features are indeed helpful, it must be noted that the method has its pitfalls. Basal-only sampling, tangential cuts resulting in false crowding of glands, difficulties in detecting spiral arteries and predecidual change, epithelial-stromal discordance, polyps, inflammation etc. often hinder the accuracy of endometrial dating. Also, some of the samples in the present study lack surface epithelium due to the small quantity of the tissue collected, which was somewhat unavoidable (discussed later). All those factors, compounded by subjective nature of the methodology, makes histological dating of endometrium a challenging feat for the pathologist26. Despite the difficulties, best efforts were made to ensure the dating in the present study was as accurate as possible, which included consultation with expert pathologists, checking and double-checking the findings and cross-referencing with multiple authentic sources.\n\nH-score, acronym for ‘histological’ score, was initially developed to quantify immunohistochemical staining of certain tumors but later its use diversified to serve many other arenas of immunohistochemistry including the assessment of endometrial receptivity9. The current study utilized this method to provide a uniform platform for the immunohistochemical markers to be assigned a number that directly reflects its staining characteristics and also makes it more amenable to the robust statistical tools applied subsequently. Due to lack of consensus regarding the roles of immunohistochemical markers and their optimum cutoff values for useful categorization as far as the investigator could search for in English scientific literature, the present study attempted to develop such a scheme tailored for this study from scratch according to the fundamental statistical principles underlying some of the most widely used scoring and categorization systems in diagnostic pathology27.\n\nHow should the endometrial expression pattern of MUC1/EMA be in order to facilitate implantation remains an open question since non-congruent results regarding this are frequently found in English scientific literature8,10. The present study attempted to address the question by incorporating MUC1 into the immunohistochemistry panel but found its correlation with IVF outcome to be weak and statistically insignificant.\n\nAlthough correlation of E-cadherin H-score with IVF outcome was not statistically significant in the present study, higher H-scores of E-cadherin were found to be more associated with IVF failure which resonates with the findings of other similar studies8,16. An interesting finding of the present study about E-cadherin was the marked difference of its relative intensity between surface epithelia and glandular epithelia where the surface epithelial staining was almost consistently higher in intensity than glands. A similar finding is noted in a study where such differential staining is appreciated in non-receptive endometrium, except the marker was PGRMC1 and not E-cadherin9. Perhaps further inquiry into such peculiarities would unravel more mysteries in the ever-expanding field of reproductive biology.\n\nAlthough this study was confined to endometrial samples at a single point of the menstrual cycle which may contribute to non-representative sampling, further research may be designed incorporating endometrial biopsy collection at more than one day of the patient’s menstrual cycle. This might be achieved by iatrogenic thickening of endometrium by progesterone and/or luteinizing hormone administration, preferably six months to one year prior to the IVF embryo transfer cycle to avoid possible complications in implantation process as practiced in case of ERA9. Besides the prediction of IVF outcome, this modified technique may provide necessary information to calculate the WOI for individual patient that might be used to transfer embryo during IVF, thus improving the odds of IVF success. However, hormone-induced changes in immunohistochemical expression of the markers would have to be taken into account if that method is followed26.\n\nDue to ethical and methodological constraints, the amount of endometrial tissue obtained by scratch procedure was generally less than that of routine endometrial curettage. This added to the difficulty in endometrial dating and in appreciating the potential heterogeneous nature of endometrial tissue. One sample had to be kept out of the study pool simply due to sample inadequacy. Moreover, endometrial sample had to be collected just once per patient, thus eliminating the possibility to evaluate samples at more than one point in a menstrual cycle for a patient. Hormonal influence on endometrial tissue may alter the morphological and immunohistochemical findings9,26. Therefore in the present study, samples were collected prior to commencement of hormone treatment. Since the endometrial ‘scratch’ biopsy procedure was performed on each of the patients of both case and control groups, any possible effect of the procedure itself would have been cancelled out between the two groups. Since this study was partly motivated by an attempt to complement the genomic approach for IVF outcome prediction known as ERA, it would be ideal to be able to cross-validate the findings by ERA of each of the samples. However, financial and time constraints prevented the investigators to pursue this aspect.\n\nIt is anticipated that the study will be used as a guide for further research and will provide essential information to design a low-cost alternative diagnostic tool to ERA for predicting IVF failure as well as individualized WOI. Thus, it may help reduce the financial and associated burdens of women undergoing IVF or other infertility treatment.\n\n\nEthical statement\n\nEthical approval was obtained from the Institutional Review Board of Bangabandhu Sheikh Mujib Medical University (approval no. BSMMU/2017/1576). Written informed consent for publication of the patients’ details was obtained from the patients.\n\n\nData availability\n\nThe underlying and extended data is publicly available on Open Science Framework: MUC1 and E-cadherin immunohistochemistry of endometrium cannot predict the outcome of in vitro fertilization: A case-control study, https://doi.org/10.17605/OSF.IO/UAMV428.\n\nSupplementary_material_1: Master datasheet [BMI = Body Mass Index; Menstrual history: 0 = Regular, 1 = Infrequent; IVF = In vitro fertilization; Endometrial phase: 0 = Secretory; 1 = Proliferative; Ep. = Epithelium; St. = Stroma; Pregnancy outcome: 0 = Success/conceived by IVF, 1 = Failure/did not conceive by IVF; Previous IVF outcome: 0 = No H/O IVF, 1 = Previous one failure; FSH = Follicle stimulating hormone; LH = Luteinizing hormone.]\n\nRaw data for Figure 3–Figure 11\n\nDigital microscopic slide files are available on request from the corresponding author (saumitra1880@yahoo.com). As these files are large in size (500MB–1GB each) they are not available on OSF.\n\nSupplementary_material_2: Proof of internal consistency (Measurement of interobserver variability by Chronbach's Alpha)\n\nSupplementary_material_3: Data analysis regarding MUC1 to calculate its correlation with IVF outcome and odds ratio along with P value (Cohen's kappa correlation or degree of agreement)\n\nSupplementary_material_4: Data analysis regarding E-cadherin to calculate its correlation with IVF outcome and odds ratio along with P value (Cohen's kappa correlation or degree of agreement)\n\nSupplementary_material_5: Patient consent form used in the study in both Bangla and English\n\nSupplementary_material_6: Routine histology lab protocol for Hematoxylin and Eosin stain used in the study\n\nSupplementary_material_7: Immunohistochemistry lab protocol for MUC1/EMA and E-cadherin antibody markers used in the study\n\nSupplementary_material_8: Ethical approval from Institutional Review Board (IRB) of Bangabandhu Sheikh Mujib Medical University (BSMMU) allowing the authors to conduct the study\n\nSupplementary_material_9: Memorandum of understanding (MOU) between CARe and Department of Pathology, BSMMU\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\nThis study was partially funded by Bangabandhu Sheikh Mujib Medical University.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nMascarenhas MN, Flaxman SR, Boerma T, et al.: National, regional, and global trends in infertility prevalence since 1990: a systematic analysis of 277 health surveys. PLoS Med. 2012; 9(12): e1001356. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFatima P, Ishrat S, Rahman D, et al.: Quality and quantity of infertility care in Bangladesh. Mymensingh Med J. 2015; 24(1): 70–73. PubMed Abstract\n\nNahar P: Invisible women in Bangladesh: Stakeholders’ views on infertility services. Facts Views Vis Obgyn. 2012; 4(3): 149–156. PubMed Abstract | Free Full Text\n\nSociety of Assisted Reproductive Technology: National Summary Report. [Online].; 2014. [cited 2016 October 7]. Reference Source\n\nDiaz-Gimeno P, Horcajadas JA, Martínez-Conejero JA, et al.: A genomic diagnostic tool for human endometrial receptivity based on the transcriptomic signature. Fertil Steril. 2011; 95(1): 50–60.e15. PubMed Abstract | Publisher Full Text\n\nOak Clinic: Endometrial Receptivity Array (ERA) Guidance. [Online].; 2016. [cited 2016 October 7]. Reference Source\n\nDíaz-Gimeno P, Ruiz-Alonso M, Blesa D, et al.: The accuracy and reproducibility of the endometrial receptivity array is superior to histology as a diagnostic method for endometrial receptivity. Fertil Steril. 2013; 99(2): 508–517. PubMed Abstract | Publisher Full Text\n\nAchache H, Revel A: Endometrial receptivity markers, the journey to successful embryo implantation. Hum Reprod Update. 2006; 12(6): 731–746. PubMed Abstract | Publisher Full Text\n\nGarrido-Gómez T, Quiñonero A, Antúnez O, et al.: Deciphering the proteomic signature of human endometrial receptivity. Hum Reprod. 2014; 29(9): 1957–1967. PubMed Abstract | Publisher Full Text\n\nBhusane K, Bhutada S, Chaudhari U, et al.: Secrets of Endometrial Receptivity: Some Are Hidden in Uterine Secretome. Am J Reprod Immunol. 2016; 75(3): 226–236. PubMed Abstract | Publisher Full Text\n\nBouquet de Jolinière J, Ayoubi JM, Lesec G, et al.: Identification of displaced endometrial glands and embryonic duct remnants in female fetal reproductive tract: possible pathogenetic role in endometriotic and pelvic neoplastic processes. Front Physiol. 2012; 3: 444. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAplin JD, Hey NA, Li TC: MUC1 as a cell surface and secretory component of endometrial epithelium: reduced levels in recurrent miscarriage. Am J Reprod Immunol. 1996; 35(3): 261–266. PubMed Abstract | Publisher Full Text\n\nAplin JD, Hey NA, Graham RA: Human endometrial MUC1 carries keratan sulfate: characteristic glycoforms in the luminal epithelium at receptivity. Glycobiology. 1998; 8(3): 269–276. PubMed Abstract | Publisher Full Text\n\nSinger G, Kurman RJ, McMaster MT, et al.: HLA-G immunoreactivity is specific for intermediate trophoblast in gestational trophoblastic disease and can serve as a useful marker in differential diagnosis. Am J Surg Pathol. 2002; 26(7): 914–920. PubMed Abstract | Publisher Full Text\n\nPoncelet C, Leblanc M, Walker-Combrouze F, et al.: Expression of cadherins and CD44 isoforms in human endometrium and peritoneal endometriosis. Acta Obstet Gynecol Scand. 2002; 81(3): 195–203. PubMed Abstract | Publisher Full Text\n\nHantak AM, Bagchi IC, Bagchi MK: Role of uterine stromal-epithelial crosstalk in embryo implantation. Int J Dev Biol. 2014; 58(2–4): 139–46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMovaghar B, Askarian S: Expression of e-cadherin, leukemia inhibitory factor and progesterone receptor in mouse blastocysts after ovarian stimulation. Cell J. 2012; 14(3): 225–30. PubMed Abstract | Free Full Text\n\nRiethmacher D, Brinkmann V, Birchmeier C: A targeted mutation in the mouse E-cadherin gene results in defective preimplantation development. Proc Natl Acad Sci U S A. 1995; 92(3): 855–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKelsey JL, Whittemore AS, Evans AS, et al.: Methods of sampling and estimation of sample size. In: Methods in Observational Epidemiology. 2nd ed. New York: Oxford University Press; 1996. Reference Source\n\nStenger M: Calculating H-Score. [Online].; 2015; [cited 2018 January 9]. Reference Source\n\nSnoper D: Analysis of Variance (ANOVA) Calculator - One-Way ANOVA from Summary Data. [Online].; 2018; [cited 2018 January 8]. Reference Source\n\nZaiontz C: Real Statistics Using Excel - Cronbach’s Alpha. [Online].; 2018; [cited 2018 January 8]. Reference Source\n\nTavakol M, Dennick R: Making sense of Cronbach’s alpha. Int J Med Educ. 2011; 2: 53–55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFluss R, Faraggi D, Reiser B: Estimation of the Youden Index and its associated cutoff point. Biom J. 2005; 47(4): 458–72. PubMed Abstract | Publisher Full Text\n\nSilva IdS: Dealing with confounding in the analysis. In: Cancer Epidemiology: Principles and Methods. Lyon: International Agency for Research on Cancer, World Health Organization; 1999; 305–331. Reference Source\n\nBhattacharya S, Maheshwari A, Mollison J: Factors associated with failed treatment: an analysis of 121,744 women embarking on their first IVF cycles. PLoS One. 2013; 8(12): e82249. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMazur MT, Kurman RJ: Diagnosis of Endometrial Biopsies and Curettings: A Practical Approach. 2nd ed. New York: Springer; 2005. Publisher Full Text\n\nChakravarty S: MUC1 and E-Cadherin Immunohistochemistry of Endometrium Cannot Predict the Outcome of in Vitro Fertilization: A Case-Control Study. OSF. 2019. http://www.doi.org/10.17605/OSF.IO/UAMV4"
}
|
[
{
"id": "44393",
"date": "27 Feb 2019",
"name": "Antonio Simone Laganà",
"expertise": [
"Reviewer Expertise Reproductive 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 read with great interest the Manuscript titled “MUC1 and E-cadherin immunohistochemistry of endometrium cannot predict the outcome of in vitro fertilization: A case-control study” (F1R-VER19607-R), which falls within the aim of F1000Research.\n\nIn my honest opinion, the topic is interesting enough to attract the readers’ attention. Methodology is accurate and conclusions are supported by the data analysis. Nevertheless, authors should clarify some points and improve the discussion citing relevant and novel key articles about the topic.\nAuthors should consider the following recommendations:\nManuscript should be further revised by a native English speaker. Was this study registered? I could not find any information about this point. Embryo transfer is a key stage in IVF, in which the quality of performance determines the outcome. According to recent evidence, transvaginal ultrasound guidance of the transfer significantly increases the percentage of pregnancies per transfer, both in the general population and in the reference population, compared with transfers performed under transabdominal ultrasound guidance. Authors should add few lines about the topic, referring to: Cozzolino et al.(2018)1; Larue et al.(2017)2. Recent and robust data suggested that endometrial scratch injury performed once, may improve clinical pregnancy and ongoing pregnancy rates in IUI/IVF cycles. I invite authors to discuss these data (refer to: Vitagliano et al. (2018)3; Vitagliano et al.(2018)4; Vitagliano et al. (2019)5), in the light of available evidence.\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": "44200",
"date": "15 Mar 2019",
"name": "Indira Hinduja",
"expertise": [
"Reviewer Expertise IVF",
"reproductive biology",
"endometrial receptivity"
],
"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 study, the authors evaluated the expression of MUC1 and E-cadherin by immunohistochemistry to determine their role in endometrial receptivity and correlated with the IVF outcome. The study was performed to develop cost effective laboratory investigation as opposed to expensive ERA. However, they did not find statistically significant correlation in case of both the markers. The manuscript is well written in different sections of introduction, methods, results and discussion. However, following are the comments related to the study design and conclusion of the study:\nIn the “procedure performed at CARe’’ section, it has been stated that ‘The procedure was performed by a competent physician on the 21st day of natural cycle or on a day accordingly adjusted for irregular, longer or hormone assisted cycle immediately before the commencement of IVF cycle’.\nPlease explain following points:\nIn case of natural cycle, whether the ovulation was detected by follicular monitoring or LH to determine the window of implantation e.g. LH+6/7? In case of women with irregular menstrual cycle, or hormone assisted cycle, how the adjustment was done? which day was considered for endometrial receptivity? In the case group (n=17), how many of the cases were included from natural cycle and how many were adjusted hormonally? Whether the authors have compared these two groups separately? In the study, the patients were selected as cases or controls based on their IVF outcome. Please mention whether all the cases and controls are given similar regimen of hormones or same the hormonal protocol? Whether the patients with PCOS, endometriosis and other hormonal abnormalities were excluded?\n\nIn table 1, it has been stated that endometrial biopsy was taken on Day 12 of the menstrual cycle for four patients (which are from the control group). It means that in controls, the endometrial biopsy was taken during proliferative phase. To have comparison of the women with IVF success and women with IVF failure, the endometrial biopsy should be on same day i.e during the window of implantation especially when authors aim to study the endometrial receptivity based on IVF outcome. Please explain this important point in detail. As mentioned in the last paragraph of discussion, please explain how the present study results can be used to predict individualized WOI? Whether the tissue samples stored in trizol or RNA later? Can the markers be correlated with mRNA levels (gene expression levels)? Based on the previous reports, hundreds of genes are involved in the receptivity of endometrium. For instance ERA predicts the day of personalized ET based on m RNA levels of 238 genes. Therefore, it would be difficult to predict the IVF outcome based on only 2 genes. The panel of some of the genes needs to be performed by IHC to achieve the scientific conclusion of designing the low cost alternative diagnostic tool like ERA.\n\nWith respect to the study outcomes are concerned i.e. development of cost-effective immunohistochemistry based test to be a predictor of IVF outcome:\nThe sample size of the study is low The ratio of cases Vs Controls is low (17Vs4) The study should have a third group of healthy fertile women as a reference group which would be the actual control group comparing fertile (natural) Vs infertile (IVF)Vs Fertile (IVF) women\nI recommend the manuscript for indexing only after the satisfactory explanation of above queries.\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?\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? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-162
|
https://f1000research.com/articles/8-156/v1
|
05 Feb 19
|
{
"type": "Research Article",
"title": "Antimicrobial resistance surveillance among gram-negative bacterial isolates from patients in hospitals in Khartoum State, Sudan",
"authors": [
"Hana Salaheldin Elbadawi",
"Kamal Mustafa Elhag",
"Elsheikh Mahgoub",
"Hisham N. Altayb",
"Muzamil Mahdi Abdel Hamid",
"Kamal Mustafa Elhag",
"Elsheikh Mahgoub",
"Hisham N. Altayb",
"Muzamil Mahdi Abdel Hamid"
],
"abstract": "Background: Antimicrobial resistance (AMR) among gram-negative bacilli is a global health problem. Surveillance of AMR is required to advise on empirical antimicrobial therapy. This study aimed at evaluating the frequency and the AMR patterns of gram-negative isolates from patients treated in eight hospitals in Khartoum State, Sudan. Methods: A cross-sectional laboratory-based study was conducted over a 6 months period at the Microbiology Department, Soba University Hospital- Khartoum State, Sudan. All gram-negative isolates from blood, urine, wound, and sputum during the period of study were included. Identification and antimicrobial susceptibility testing were carried out for all isolates. Results: A total of 734 Gram-negative bacilli were isolated. Klebsiella pneumoniae (249 isolates, 34%) was the most frequently encountered one, followed by Pseudomonas aeruginosa (153 isolates, 21%), E.coli (123 isolates, 17%), Acinetobacter baumannii (75 isolates, 10%), Burkholderia cepacia (42 isolates, 6%), Proteus mirabilis and Proteus vulgaris (28 isolates, each, (4%) Enterobacter colecaes (28 isolates, 4%), Stenotrophomonas maltophilia (21 isolates, 2.8%), and other gram-negative bacilli (15 isolates, 2.2%) The analysis of the antimicrobial susceptibility patterns showed that 134 (22.3%) isolates were resistant to three or more classes of antibiotics, including cephalosporins, β-lactam–β-lactamase inhibitor, quinolones, aminoglycosides and carbapenems. Conclusion: This high level of resistance among gram-negative bacilli in Khartoum state hospitals is alarming. The local health authorities should be prompted to step up infection control programs and introduce the concept of antimicrobial stewardship in Khartoum State hospitals.",
"keywords": [
"Gram-negative bacilli",
"Multidrug resistant bacteria",
"laboratory-based study",
"Surveillance."
],
"content": "Introduction\n\nAntimicrobial resistance (AMR) constitutes a continuously growing threat to the effective treatment of microbial infections1. However, the direct impact of AMR on the health of hospitalized patients or the people in the community, as well as the financial burden experienced by health care systems in managing the infections and complications due to AMR, are still mostly uncertain2. Antibacterial drugs are widely used worldwide both in human health and food industry. Overuse of these medications can favor the selection and the spread of multidrug resistant (MDR) bacteria1. Multi drug resistance is defined as resistance to at least three different antibiotic groups, as reported by Masgala and Kostaki3. Antibiotic resistance among a variety of bacterial species is increasing in healthcare and community setting. Extended-spectrum β-lactamase and carbapenemase production are the most frequently emerging resistance mechanisms among gram-negative bacilli4. Gram-negative bacilli including Enterobacteriaceae and non-lactose fermenting bacteria such as Pseudomonas spp. and Acinetobacter spp. are the main causes of hospital-acquired infection in critical care units2,5. The antibiotic resistance rates among these organisms have amplified uncontrollably in a matter of a few years to become worldwide2. According to the Centre for Disease Control and Prevention, gram-negative bacilli are able to develop antibiotic resistance through multiple methods and are particularly competent at spreading the resistance among species via horizontal gene transfer6. The enigma of AMR is particularly plaguing low- and middle-income countries, where infectious diseases are the commonest cause of hospitalization and death; and the newer antibiotics cannot be afforded7.\n\nAMR surveillance is key for determining the prevalence patterns of AMR, which is fundamental for the development of national and international treatment strategies. Most of the available surveillance data are derived from developed countries; the studies conducted in developing countries are unfortunately not adequate.\n\nThis surveillance study was undertaken in order to find out the different types of the AMR patterns of bacterial pathogens isolated from patients in Khartoum State, Sudan. This study may help in formulating antibiotic policies tailored to our hospitals. These data can be used as “information for action” antibiotic stewardship and interventions to optimize antibiotic prescribing practice, therefore prolongs the usefulness of existing antibiotics.\n\n\nMethods\n\nThis is a cross-sectional laboratory based study carried out in the department of medical microbiology Soba University Hospital (SUH) and Institute of Endemic Diseases, University of Khartoum, Sudan. A total of 734 Gram-negative bacteria were isolated from patients treated in eight hospitals in Khartoum state including: two university hospitals (Soba university hospital and Bashaier); three teaching hospitals (Ibrahim Malik, Bahri and Sadabulalla) , a specialized hospital (Elfouad), and two private hospitals (Imperial and Elswedy). This study included all clinical specimens received in Soba University hospital microbiology laboratory in a period from October 2016 to February 2017 from various wards in the aforementioned hospitals including: intensive care units (ICUs), neonatal ICUs, medicine units, surgery units, pediatric units, and renal units. The isolates were collected from different clinical specimens including: blood (243 isolates, 33.1%), urine (230 isolates, 31.3%), wounds (183 isolates, 25%), sputum (22 isolates, 3%), catheter tips (26 isolates, 3.4%) and body fluids (including cerebrospinal, peritoneal, pleural, acetic and synovial fluid; 30 isolates 4.2%). Microorganisms were grown on Blood, Chocolate and MacConkey agar. Then, they were identified according to standard microbiological procedures (based on colony morphology, microscopy, and biochemical tests)8. Quality control strains were used in biochemical tests and antimicrobial susceptibility testing [E. coli (ATCC #25922) and P. aeruginosa (ATCC #27853)].\n\nMost gram-negative Bacilli isolates were further identified and confirmed by PCR. Guanidine chloride method, as described by Alsadig et al.9, was used for DNA extraction followed by PCR which it was carried out using thermal cycler (analytikjena® Biometra TADVANCED, Germany), by using the following primers (Macrogen, Korea), using species-specific primers for Klebsiella pneumoniae, Escherichia coli, Pseudomonas aeruginosa, Acinetobacter baumannii and Universal 16S rRNA primers (Table 1) the reaction was carried out in a total reaction volume of 25 μl (5μl Master mix of Maxime RT premix kit (iNtRON Biotechnology, Seongnam, Korea), 0.6 μl of forward primer, 0.6 μl of reverse primer, 2μl of DNA and 16.8 μl deionized sterile water). The cycle condition as the following: initial denaturation step at 94°C for 5-min, followed by 30 cycles of denaturation at 94°C for 45 seconds, primer annealing temperature according to the primers Table 1 for 45 seconds, followed by step of elongation at 72°C for 60 seconds and the final elongation at 72°C for 5 min10. The purity and integrity of each PCR product was evaluated electrophoresis in a 2% agarose gel in TBE 1X, that contain 2.5 μl of (20mg/ml) ethidium bromide at 100V for 40 min. The specific amplified product were detected by comparing with 100 base-pairs standard DNA ladder (iNtRON BIOTECHNOLOGY, Seongnam, Korea) Bands were visualized under U.V transilluminater (analytikjena® Biometra BDAcompact, Germany). The PCR product of 16SrRNA were purified and Sanger sequencing was performed by Macrogen Company (Seoul, Korea). Then nucleotides sequences of the genes 16SrRNA achieved were searched for sequence similarity using nucleotide BLAST for species identification11.\n\nSusceptibility testing was performed using the Kirby-Bauer disc-diffusion method; each isolate was swabbed on the Muller-Hinton agar and the antibiotic discs were placed on top and incubated at 37°C for 18–24 hours12. all isolates were tested against the following antibiotic disc (Mast Diagnostic): amoxycillin clavulanate (AMC) (30 μg), cefuroxime (CXM) (30 μg), cephalexin (CL) (30 μg), ceftriaxone (CRO) (30 μg), ceftazidime (CAZ) (30 μg), meropenem (MEM) (10 μg), imipenem (IPM) (10 μg), amikacin (AK) (30 μg), gentamicin (Gen) (10 μg), ciprofloxacin (CIP) (5 μg), trimethoprim-sulfamethoxazole (SXT) (25 μg), temocillin (TEM) (30 μg), azteroname (AZT) (30 μg) and nitrofrantoine (NIT) (300 μg). Results were interpreted according to the Clinical Laboratory Standards Institute (CLSI) guidelines12.\n\nMDR has been considered for clinically significant gram-negative Bacilli such as E. coli, K. pneumoniae, P. aeruginosa, and A. baumannii based on the aforementioned antimicrobial resistance definition. Classes of antibiotics used for MDR-GNB analysis were aminoglycoside (AMG), cephalosporins (CEPH), carbapenems (CARB), and fluroquinolones (FQ) as follows: bacteria that were MDR for four classes of antibiotics (AMG+ CEPH+ CARB+ FQ) and bacteria that were MDR for three classes of antibiotics (either AMG+CEPH +FQ, CARB+CEPH+FQ, AMG+CEPH+CARB, or AMG+FQ+CARB)3.\n\nCephalosporin resistance was defined as resistance to ceftriaxone and ceftazidime, except for P. aeruginosa, where only ceftazidime was used. Carbapenem resistance was defined as resistance to both meropenem and imipenem. Aminoglycoside resistance was defined as resistance to both gentamicin and amikacin, Ciprofloxacin resistance was considered an indication to fluoroquinolones resistance.\n\nFormal permission was obtained from the managers of Soba University Hospital and the Institutional Research Ethics Committee of the Institute of Endemic Diseases, University of Khartoum, approved this study under reference number IEND_REC 12/2017. Patient consent was waived by the Research Ethics Committee.\n\nData were analysed using Microsoft Excel and SPSS version 20.0. Cross-tabulation was used to present the different relations between data, qualitative data were performed using a χ2 test (significance was set at p≤0.05), which was performed to find the differences between bacterial isolates with resistance to at least one class of antibiotics by specimens (blood, urine, wound and other samples) p-values were determined for primary and secondary outcomes.\n\n\nResults\n\nIsolated Gram-negative bacilli showed different strains, including E. coli (123 isolates, 17%), K. pneumoniae (249 isolates,34%), P. aeruginosa (153 isolates, 21%), A. baumannii (75 isolates, 10%), Burkholderia cepacia (42 isolates, 6%), Proteus mirabilis and Proteus vulgaris (28 isolates, 4%), Enterobacter colecaes (28 isolates, 4%), Stenotrophomonas maltophilia (21 isolates, 2.8%) and other gram-negative bacilli (15 isolates, 2.2%).\n\nWhile isolates were distributed among the different hospital units, most of the pathogenic strains were isolated from neonatal intensive care unit (ICU) (182 isolates, 24.8%) mainly K. pneumoniae (77 isolate, 42.3%) and pediatric units (175 isolates 23.8%) the most prevalent strains were K. pneumoniae (50 isolate, 29%) and P. aeruginosa (44 isolates, 25.1), while medicine (147 isolates, 20%), K. pneumoniae and E.coli (45 isolates 30.6%) for both. For the surgery unit (103 isolates, 14%) renal unit (78 isolates, 10.7%) and ICU (49 isolates, 6.7%) mainly K. pneumoniae (19 isolates, 39%), P. aeruginosa and A. baumannii were (9 isolates, 20.4) for both. Klebsiella pneumoniae was the most isolated organism from all hospital units. The distribution of different gram-negative isolates is shown in Table 2.\n\n*Other Gram-negative bacilli include Citrobacter species, Serratia species, Vebrio vurneficus and Morganella morganii. †Body fluids include cerebrospinal fluid, peritoneal fluid, pleural fluid, acetic fluid and synovial fluid.\n\nWith regard to the distribution of the isolates among different clinical specimens, Klebsiella pneumoniae and Pseudomonas aeruginosa were isolated mainly in blood specimens 39% and 25% respectively, while K. pneumoniae and E.coli were 36% and 30% of urine samples. A. baumannii was isolated from 14% of wound specimens and P. aeruginosa was isolated from 23% of wound specimens. For more details about gram-negative bacilli among different specimen see Table 3.\n\n*Other Gram-negative bacilli include Citrobacter species, Serratia species, Vebrio vurneficus and Morganella morganii.\n\nAntibiotic resistance pattern are shown in Figure 1. Out of 734 isolates tested using the disk diffusion method, the highest percentage of resistance, in 97% and 93.5% of isolates, were found against ampicillin and cephalexin, respectively, followed by amoxicillin/clavulanic acid(90%), cefotaxime (89.7%), ceftriaxone (88.4%) and ceftazidime (79.2%). In addition, co-trimoxazole and nitrofurantoin resistance were detected in 74.4% and 75.2% of isolates, respectively. Resistance rates also were high in ciprofloxacin (45.2%), gentamicin (52.5%) and amikacin (18.3%). Meropenem and imipenem were the most effective antibiotic tested, with resistance observed in 21.6% and 16.2% of isolates, respectively.\n\nThe antimicrobial resistance patterns of most commonly isolated organisms are shown in Figure 2. K. pneumoniae resistant pattern as the following (22%) of them were resistant to meropenem while (11%) were resistant to imipenem, ceftazidime (80.6%), Gentamicin (52%), ciprofloxacin (42%) and amikacin (16.7%). With regard to the E. coli antimicrobial resistant pattern, meropenem (9%), imipenem (8%), ceftazidime (84.2%), ciprofloxacin (66.4%), Gentamicin (53.1%) and amikacin (12.0%). In Pseudomonas aeruginosa the resistance rate was meropenem (20%), imipenem (22%) ceftazidime (81%) followed by gentamicin (57.5%), ciprofloxacin (22.5%) and amikacin (9.5%). The rate of antimicrobial resistant among A. baumannii was as the following; meropenem (73.7%), imipenem (66.7%), amikacin (63.2%), gentamicin (79%) and ciprofloxacin (79%).\n\nMER, meropenem; IMP, imipenem; AK, amikacin; GEN, gentamicin; CIP, ciprofloxacin; CAZ, ceftazidime; CRO, ceftriaxone; CXM, cefuroxime; CN, cephalexin; AM, ampicillin; AMC, amoxycillin-clavulanate; SXT, trimethoprim-sulfamethoxazole; NIT, nitrofurantoin; TEM, temocillin; ATZ, azetronam.\n\nThe gram-negative bacilli that were resistant to several antibiotic groups are shown in Table 4. Of 600 GNB isolates, 134 (22.3%) isolates were MDR. Of those MDR organisms, 48 isolates (8%) were resistant to four classes of antimicrobial drugs: A. baumannii (38 isolates, 50.6%), K. pneumoniae (9 isolates, 3.6%), and E. coli (1 isolate, 0.8%). A further 86 isolates (14.2%) were resistant to three classes of antimicrobial drugs: A. baumannii (38 isolates, 50.6%), K. pneumoniae (47 isolates, 18.8%), P. aeruginosa (21 isolates, 13.7%) and E. coli (6 isolate, 4.8%).\n\nNA, not applicable; AMG, aminoglycoside; CEPH, cephalosporins; FQ, florquinolones; CARB, carbapenem.\n\n\nDiscussion\n\nInfection with MDR gram-negative Bacilli is a major problem worldwide, associated with increased patients morbidity and mortality17. In Sudan, the increasing number of MDR bacteria is a real clinical challenge18,19. This study was undertaken; to identify the different patterns of AMR in bacterial pathogens isolated from patients treated in various wards of hospitals.\n\nIn this study, K. pneumoniae and P. aeruginosa strains were more prevalent in blood specimens while K. pneumoniae and E.coli strains were more frequently isolated in urine specimens.\n\nK. pneumoniae is a prominent pathogen causing both hospital-acquired and community-associated infections, including bacteremia, wound infection, pneumonia, urinary tract infections and other infections20,21. In this study, K. pneumoniae was the most commonly isolated organism from the blood specimens mainly in neonatal sepsis in a high rate (39.8%). Most of these strains resistant to cephalosporins and other class of antibiotics including carbapenem as reported worldwide7,21,22.\n\nE. coli is regarded as the commonest pathogen of the urinary tract, causing complicated and uncomplicated urinary tract infections23. In this study, the most frequently observed pathogens in urine specimens were E. coli (30%) and K. pneumoniae (36%). This finding is in harmony with the results of several other studies, including: de Francesco et al., who found that E.coli was the commonest GNB (42.4%) isolate from urine specimens of patients with urinary tract infection24; a study in Tanzania reported 38% of E.coli isolated from urinary specimens25 and other studies from Pakistan and India25,26.\n\nIn this study, non-lactose-fermenting gram-negative bacilli such as P. aeruginosa and A. baumannii were mostly encountered in ICU patients, these organisms were isolated in 20.4% of samples, and were found in different clinical samples such as: blood, wound and sputum. This finding agrees with a study by Vincent et al. that reported P. aeruginosa and A. baumannii were frequently isolated from ICU patients by Vincent et al.27; Jitendra et al., reported that A. baumannii was the second most common pathogen in an ICU of a tertiary care center28.\n\nConcerning the infection type, we found that P. aeruginosa were associated with 25% of blood stream infections and 23% of wound infection, while A. baumannii mainly associated with wound infection in 14% in agree with Gales et al. 201029.\n\nResistance of gram-negative bacilli is widespread and multidrug resistance has been reported in many studies2,4,5, causing challenges in the treatment of nosocomial infections. The resistance pattern was commonly reported in classes such as cephalosporins, carbapenem, aminoglycosides and quinolones30–32. In this study, we observed high rates of resistance to extended-spectrum β-lactamases (ESBL), resistance to cephalexin, cefuroxime, ceftazidime and ceftriaxone, in addition to resistance to ampicillin and amoxicillin/clavulanic acid.\n\nThe analysis of the antimicrobial susceptibility patterns of the study isolates showed high rate of MDR organisms that were resistant to three or more classes of antibiotics, including carbapenem and aminoglycosides. This pattern was observed mainly among Acinetobacter baumannii, Pseudomonas aeruginosa and Klebsiella pneumoniae.\n\nThe most resistant strain was Acinetobacter baumannii, being resistant to all four classes of antibiotics used in 50.6% of isolates. A total of 73.7% of A. baumannii were found to be resistant to meropenem, and 66.7% to imipenem, while in the cephalosporin class, more than 91% of the isolates were resistant. A. baumannii also have high resistance rate to aminoglycosides and quinolones (63.2% for amikacin and 79% for both gentamicin and ciprofloxacin). This increasing resistance among A. baumannii has become a public-health issue, because this bacterium frequently causes nosocomial infections31.\n\nIn this study the most common clinical isolates of the Enterobacteriaceae family were K. pneumonia and E. coli. In both species, there was a high prevalence of resistance against quinolones, aminoglycosides and beta-lactams. Rates of resistance to carbapenems were alarmingly high: 22% in K. pneumoniae and 9% in E. coli. These bacteria have high resistance rate to ceftazidime and cephalexin, (80.6% and 92%, respectively). These rates are much higher than those observed in 2013 by Ali in Soba University hospital, who found that ceftriaxone and ceftazidime resistance rates ranged from 56% to 79%21,33. Aminoglycoside (specifically amikacin) resistance rates among K. pneumoniae and E. coli were 16.7% and 12.1%, respectively, which is higher than those observed by Lee in 2013, who found the resistance rate to amikacin was 6.2% in K. pneumoniae and 1.3% in E. coli30. In this study Gentamicin resistance among both K. pneumoniae and E. coli was 53%, which is a high resistance rate. E. coli was highly resistant to quinolones like ciprofloxacin (in 66.4%), whereas K. pneumoniae was resistant in 42% of isolates. This finding is much lower than that observed by Moolchandani et al. in 201732.\n\nPseudomonas aeruginosa was resistant to carbapenem in 22% of isolates and was highly resistant to ceftazidime (in 81% of isolates) followed by gentamicin (57.5%), ciprofloxacin (22.5%) and amikacin (9.5%). resistance to many antibiotic among P. aeruginosa was reported in many studies4,34.\n\nThe high level of resistance in the current study can be attributed to the unrestricted use of antibiotics in Sudanese hospitals; this injudicious use has been shown to have, an important role in increasing carbapenem resistance32. During this study, 134 gram negative bacilli resistant to three or four classes of antibiotics were isolated over a period of six months, which is relatively higher rate than the rate reported in a previous study also conducted in SUH over 30 months, from January 2011 to June 201333. This concerning finding indicates the rapid acceleration in the rate of emergence of MDR organisms in our local sittings. Moreover, 80 bacterial strains were resistant to all available antibiotics, including meropenem.\n\nIn addition, microbiology laboratories play a crucial role against the spread of antimicrobial resistance by accurately identifying causative pathogens and detecting their antimicrobial susceptibility profile, so as to guide the proper use of antibiotics by the care providers. Unfortunately, in Sudan the lack of reliable microbiology laboratory services adds yet another layer to the multifactorial problem of MDR organisms’ epidemic in our country. Most of the laboratories in our country use the disk diffusion method for antimicrobial susceptibility testing; however, there are no policies regarding phenotypic screening for antibiotic resistance among certain organisms; resulting in that the laboratory report may not be accurate and misguide the doctors. For instance, in the disk diffusion method, if the inhibitory zone around ceftazidime was ≤22 mm, ideally this should trigger testing for the presence of ESBLs in the isolated pathogen, because the presence of these enzymes renders all penicillins, cephalosporins and monobactams12 useless against that certain isolate, even if they shows susceptible results in the routine test. Yet, this extra step is not taken and only the routine susceptibility results will be reported, which may inaccurately show some antibiotics as susceptible while in reality the organism is resistant to them. Lastly, there is a limited choice of available antibiotics in Sudan; the most widely available are cephalosporins, which makes them the backbone of treating infectious diseases regardless of the isolate or its susceptibility profile. Carbapenems are believed to treat all evils, once they were reserved for the treatment of MDR organisms causing systemic infections4, but the recent trend led to over use them for treating infections that can be managed by less potent antibiotics, contributing to this vicious cycle that aggravate the AMR issue even more.\n\n\nConclusion\n\nIn conclusion, there was a high prevalence of gram negative bacterial pathogen associated hospital and community acquired infections, with increasing rates of resistance to available antibiotics. Strict infection control measures should be implemented, and antimicrobial stewardship should be initiated and policed to decrease the spread of MDR pathogens in Sudanese hospitals.\n\n\nData availability\n\nRaw data for the present study, including the genotypes, isolation location and resistance status of each bacterial isolate, is available of figshare. DOI: https://doi.org/10.6084/m9.figshare.758444935.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Grant information\n\nThis research received partial research fund from the Ministry of Higher Education and Scientific Research, Sudan (grant number SUD/MOH/MMAH/08/2016).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe would like to thank the technical staff of Medical Microbiology Department in Soba University Hospital, University of Khartoum for their help in strain and data collection.\n\nAn earlier version of this article can be found on bioRxiv (DOI: https://doi.org/10.1101/486274).\n\n\nReferences\n\nHarris P, Paterson D, Rogers B: Facing the challenge of multidrug-resistant gram-negative bacilli in Australia. Med J Aust. 2015; 202(5): 243–7. PubMed Abstract | Publisher Full Text\n\nMehrad B, Clark NM, Zhanel GG, et al.: Antimicrobial resistance in hospital-acquired gram-negative bacterial infections. Chest. 2015; 147(5): 1413–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMasgala A, Kostaki K II: Multi Drug Resistant Gram Negative Pathogens in Long Term Care Facilities: A Steadily Arising Problem. J Infect Dis Diagn. 2015; 1(1): 101. Publisher Full Text\n\nKaraiskos I, Giamarellou H: Multidrug-resistant and extensively drug-resistant Gram-negative pathogens: current and emerging therapeutic approaches. Expert Opin Pharmacother. 2014; 15(10): 1351–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRosenthal VD, Maki DG, Mehta Y, et al.: International Nosocomial Infection Control Consortium (INICC) report, data summary of 43 countries for 2007-2012. Device-associated module. Am J Infect Control. 2014; 42(9): 942–56. PubMed Abstract | Publisher Full Text\n\nHuang TD, Bogaerts P, Berhin C, et al.: Increasing proportion of carbapenemase-producing Enterobacteriaceae and emergence of a MCR-1 producer through a multicentric study among hospital-based and private laboratories in Belgium from September to November 2015. Euro Surveill. 2017; 22(19): pii: 30530. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReport G: Antimicrobial resistance. 2014.\n\nWauters G, Vaneechoutte M: Approaches to the Identification of Aerobic Gram-Negative Bacteria. In: Manual of Clinical Microbiology. 11th ed. Washington, DC; 2015; 613–34. Publisher Full Text\n\nAlsadig G, Arbab MA, Aldeaf SAH, et al.: Allele Frequency Of P53 Gene Arg72Pro In Sudanese Meningioma Patients And Controls. InJ Sci Technol RES. 2014; 3(6): 2277–8616. Reference Source\n\nSrinivasan R, Karaoz U, Volegova M, et al.: Use of 16S rRNA gene for identification of a broad range of clinically relevant bacterial pathogens. PLoS One. 2015; 10(2): e0117617. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAtschul SF, Madden TL, Schäffer AA, et al.: Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acid Res. 1997; 25(17): 3389–3402. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTesting S: M100 Performance Standards for Antimicrobial. 27th ed. 2017; 106-143. Reference Source\n\nLiu Y, Liu C, Zheng W, et al.: PCR detection of Klebsiella pneumoniae in infant formula based on 16S-23S internal transcribed spacer. Int J Food Microbiol. 2008; 125(3): 230–5. PubMed Abstract | Publisher Full Text\n\nSabat G, Rose P, Hickey WJ, et al.: Selective and sensitive method for PCR amplification of Escherichia coli 16S rRNA genes in soil. Appl Environ Microbiol. 2000; 66(2): 844–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSpilker T, Coenye T, Vandamme P, et al.: PCR-based assay for differentiation of Pseudomonas aeruginosa from other Pseudomonas species recovered from cystic fibrosis patients. J Clin Microbiol. 2004; 42(5): 2074–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHiggins PG, Wisplinghoff H, Krut O, et al.: A PCR-based method to differentiate between Acinetobacter baumannii and Acinetobacter genomic species 13TU. Clin Microbiol Infect. 2007; 13(12): 1199–1201. PubMed Abstract | Publisher Full Text\n\nZilberberg MD, Nathanson BH, Sulham K, et al.: Multidrug resistance, inappropriate empiric therapy, and hospital mortality in Acinetobacter baumannii pneumonia and sepsis Crit Care. 2016; 20(1): 221. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYousif M: The prevalence of Extended Spectrum β-lactamase and AmpC- Producing Bacteria in a Sudanese Tertiary Hospital. Sudan Med J. 2015; 5(3). Reference Source\n\nIbrahim ME, Bilal NE, Hamid ME: Comparison of phenotypic characteristics and antimicrobial resistance patterns of clinical Escherichia coli collected from two unrelated geographical areas. Glob J Heal Sci. 2014; 6(6): 126–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCao X, Xu X, Zhang Z, et al.: Molecular characterization of clinical multidrug-resistant Klebsiella pneumoniae isolates. Ann Clin Microbiol Antimicrob. 2014; 13: 16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeatherall BL, Gregson D, Ross T, et al.: Incidence, risk factors, and outcomes of Klebsiella pneumoniae bacteremia. Am J Med. 2009; 122(9): 866–873. PubMed Abstract | Publisher Full Text\n\nAli MA: the prevalence and characterization of antibiotic resistance among Gram-negative bacilli. University of Khartoum; 2013.\n\nKarlowsky JA, Jones ME, Thornsberry C, et al.: Prevalence of antimicrobial resistance among urinary tract pathogens isolated from female outpatients across the US in 1999. Int J Antimicrob Agents. 2001; 18(2): 121–7. PubMed Abstract | Publisher Full Text\n\nDe Francesco MA, Ravizzola G, Peroni L, et al.: Urinary tract infections in Brescia, Italy: etiology of uropathogens and antimicrobial resistance of common uropathogens. Med Sci Monit. 2007; 13(6): BR136–44. PubMed Abstract\n\nBlomberg B, Olsen BE, Hinderaker SG, et al.: Antimicrobial resistance in urinary bacterial isolates from pregnant women in rural Tanzania: implications for public health. Scand J Infect Dis. 2005; 37(4): 262–8. PubMed Abstract | Publisher Full Text\n\nHaider G, Zehra N, Munir AA, et al.: Risk factors of urinary tract infection in pregnancy. J Pak Med Assoc. 2010; 60(3): 213–6. PubMed Abstract\n\nVincent JL, Rello J, Marshall J, et al.: International study of the prevalence and outcomes of infection in intensive care units. JAMA. 2009; 302(21): 2323–9. PubMed Abstract | Publisher Full Text\n\nJaveri JR, Patel SM, Nayak SN, et al.: A study on bacteriological profile and drug sensitivity & resistance pattern of isolates of the patients admitted in intensive care units of a tertiary care hospital in Ahmadabad. Natl J Med Res. 2016; 2(3): 330–4. Reference Source\n\nZavascki AP, Carvalhaes CG, Picã RC, et al.: Multidrug-resistant Pseudomonas aeruginosa and Acinetobacter baumannii: resistance mechanisms and implications for therapy. Expart Rev Anti infet ther. 2010; 8(1): 71–93. PubMed Abstract | Publisher Full Text\n\nLee S, Han SW, Kim KW, et al.: Third-generation cephalosporin resistance of community-onset Escherichia coli and Klebsiella pneumoniae bacteremia in a secondary hospital. Korean J Intern Med. 2014; 29(1): 49–56. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDiene SM, Rolain JM: Carbapenemase genes and genetic platforms in Gram-negative bacilli: Enterobacteriaceae, Pseudomonas and Acinetobacter species. Clin Microbiol Infect. 2014; 20(9): 831–8. PubMed Abstract | Publisher Full Text\n\nMoolchandani K, Sastry AS, Deepashree R, et al.: Antimicrobial Resistance Surveillance among Intensive Care Units of a Tertiary Care Hospital in Southern India. J Clin Diagn Res. 2017; 11(2): DC01–DC07. PubMed Abstract | Publisher Full Text | Free Full Text\n\nElhag KM: Review Article Diversification of antibiotics as a means to control antimicrobial resistance and improve treatment options in Sudan. Sudan Med J. 2013; 49(3): 128–35. Reference Source\n\nMagiorakos AP, Srinivasan A, Carey RB, et al.: Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance. Clin Microbiol Infect. 2012; 18(3): 268–81. PubMed Abstract | Publisher Full Text\n\nElbadawi H, Hamid MMA: Antimicrobial resistance surveillance among gram negative bacterial isolates from patients in hospitals in Khartoum State, Sudan .xlsx. figshare. Dataset. 2019. http://www.doi.org/10.6084/m9.figshare.7584449.v1"
}
|
[
{
"id": "84294",
"date": "04 May 2021",
"name": "Robby Markwart",
"expertise": [
"Reviewer Expertise infectious disease 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 Antimicrobial resistance surveillance among gram-negative bacterial isolates from patients in hospitals in Khartoum State, Sudan by Hana Salaheldin Elbadawi et al. investigates the spectrum of gram-negative bacterial pathogens and antimicrobial resistance patterns in Khartoum State, Sudan.\nThe manuscript is generally well written and the data are presented in a clear way. However, I have some suggestions, which may help to improve the manuscript.\nSpecific comments:\nThe authors state that isolates from eight hospitals in Khartoum state were included in the study. Include these eight hospitals all hospitals in Khartoum State or are they are a sample of more hospitals?\n\nThe study was carried out between October 2016 and February 2017, i.e. encompassing the winter season. Is it possible that the spectrum of gram-negative pathogens could be different in the summer season?\n\nIn table 2 and 3, what exactly do the p-values indicate? What is the purpose of the p values? If the authors want analyze whether there are statistical differences in the proportion of isolates between in hospital wards (Table 2) or clinical sample materials (Table 3), I also suggest adding 95% confidence intervals of the proportions.\n\nI personally do not like the visual style of the Figures (Figures 1 and 2). From my knowledge, 3-dimensional bar charts are unusual in scientific publications. I suggest designing the figures in a more scientific format. Moreover, I suggest adding figure legends that include the total number of isolates tested per antibiotic. Furthermore, the authors could consider adding 95% confidence intervals to the bar charts since the study represents a more or less random sample of isolates in Sudan. The authors should discuss the representativeness of their study for Sudan or even North-East Africa.\n\nIn the discussion, the authors conclude that “K. pneumonia was the most commonly isolated organism from the blood specimens mainly in neonatal sepsis in a high rate (39.8%).” This might be correct, but the study only analyzes the distribution of gram-negative pathogens. That means from the data presented in the manuscript, the authors only can conclude that K. pneumonia is the most frequent pathogen among gram-negative pathogens. Similarly, the authors state that the most frequently observed pathogens in urine specimens were E. coli and K. pneumonia. Again, this only true among gram-negative pathogens. At least in Germany, gram-positive Enterococcus spp. are the second most frequent found pathogens in urine samples (Klingeberg 2018).\n\nIn the discussion, it would be helpful to compare the resistance patterns with studies from the same region in Sudan or surrounding regions / countries. Are the resistance proportions higher or lower in Khartoum State compared to surrounding regions / countries? Similarly, the authors could compare their resistance patterns with other regions of the world. For Europe, I suggest data from EARS-Net (European Antimicrobial Resistance Surveillance Network), for the United States CDC data are available. For example, in Germany, a nationwide study showed that only 3.5% of all A. baumannii (complex) isolates were carbapenem resistant (Said et al. Antimicrob Resist Infect Control 2021 Mar 1;10(1):451). For China, there are multiple systematic reviews and meta-analyses for antimicrobial resistance proportions for different pathogens available. A systematic comparison of the resistance rates described in this study with resistance patterns in other regions would help to put the data in perspective.\n\nIn the introduction, the authors state “This study may help in formulating antibiotic policies tailored to our hospitals. These data can be used as information for action” antibiotic stewardship and interventions to optimize antibiotic prescribing practice, therefore prolongs the usefulness of existing antibiotics.”. However, in the discussion these aspects are not addressed properly. What clinical conclusions do the authors draw from their data? How do these data help to taylor ABS interventions? Etc.\nMinor comments:\nI suggest a profound proof reading in order to correct spelling mistakes and typos in the manuscript.\n\nI suggest adding a column “Total” in Tables 2 and 3 to indicate the total numbers of the isolated pathogens (e.g. Table 2: 249 for K. pneumonia)\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? No source data required\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "86818",
"date": "14 Jun 2021",
"name": "Rayane Rafei",
"expertise": [
"Reviewer Expertise Molecular epidemiology of 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\nThis manuscript presents the surveillance data of antimicrobial resistance patterns of Gram-negative bacteria from patients in Khartoum State, Sudan. Such data are highly valuable from underrepresented regions. Nevertheless, this manuscript contains several flaws that must be addressed before acceptance.\nComments:\n1. Is there a systematic collection of all Gram-negative bacteria isolated in the 8 hospitals mentioned in the materials and methods section? The authors should describe and clarify more their study design. The patients from whom bacteria were isolated are only inpatients admitted to hospitals or also including outpatients? These 8 hospitals are sole hospitals in the Khartoum state, are they representative of such state? It would be interesting to include more data on these hospitals.\n2. The MDR definition adopted in the manuscript to categorize the strains as MDR is not well-referenced. Usually, an isolate could be considered resistant to a group of antibiotics if it is resistant to at least one (≥1) antibiotic agent in the group, not necessarily to all agents in the group (as mentioned in the second paragraph of the Classification of MDR Gram negative Bacilli). I recommend the authors to revaluate their data based on a good definition as that described in the reference Magiorakos et al. (2012)1.\n3. Reference 3 used for the definition of MDR in the introduction (and perhaps in the materials and methods) does not define such terms. Please verify that cited references matched well the text throughout the manuscript (as for reference 29 wherein the reference did not match the text).\n4. Table 1, describing the bacterial gene primers used for PCR, must include a column for gene targets. In addition, it also contains many ambiguities to clarify:\nThe primers used for Acinetobacter identification are wrongly mentioned. Indeed, the sp4F and sp4R can identify both Acinetobacter genomic species 13TU (which is recently known as Acinetobacter nosocomialis) and A. baumannii. The primers sp2F and sp4R can identify A. baumannii. A PCR with the 3 primers (sp2F, 4F, 4R) yields 2 bands when the isolate is A. baumannii (one 294 bp with primer sp4F and sp4R and one 490 bp with sp2F and sp4R), and only one band (294 bp with sp4F and sp4R) when the isolate is A. nosocomialis.\n\nFor the universal 16S rRNA, the reverse primer is not present in the mentioned reference (table 1).\n\nPlease, also use the bona fide name for primers as they are named in their original article (for example instead of F use the 27F (5’-AGAGTTTGATCCTGGCTCAG-3’)).\n5. The statistical analysis paragraph should be reformulated. Moreover, what statistical relationships do the authors use to reveal by the mentioned P-values in tables 2 and 3?\n6. Was the 16S rRNA gene sequencing carried out to all Gram-negative bacteria isolates, even those identified using PCR with species-specific primers?\n7. Which cut-off has been used for the identification of species when blasting the 16S rDNA sequence, what is the NCBI database that has been used (nucleotide collection nr/nt or rRNA/ITS database)?\n8. Please verify the charge of antibiotic disks as the charge of trimethoprim-sulfamethoxazole (SXT) should be 1.25/23.75 μg.\n9. Why are the antibiotics ticarcillin, ticarcillin-clavulanate, piperacillin, piperacillin-tazobactam and cefepime not tested?\n10. Why include only 600 GNB for studying the MDR, however, the authors start with 743 isolates. The reasons should be clearly stated in the text.\n11. Why are the data on drug susceptibility about Stenotrophomonas maltophilia and Burkholderia cepacia and other bacteria (as Enterobacter) not mentioned in the results section and analyzed in the discussion. What are the definitions used to define MDR in these bacterial isolates?\n12. The authors define the distribution of species according to the hospital wards and specimens. However, data concerning the distribution of species according to studied hospitals is also interesting.\n13. Amikacin and imipenem have a low percentage of resistance, why was amikacin not listed as the most effective antibiotic along with imipenem (paragraph one – antimicrobial resistance pattern of clinical isolates).\n14. In the discussion, the authors refer the increase of resistance in their hospitals to the unrestricted use of antibiotics in Sudanese hospitals. However, the misuse of antibiotics in the community and extra-hospital settings can also affect this upward trend. Addressing the one health concept is a way to cope with the rising threat of antimicrobial resistance. In addition, this increase in antimicrobial resistance in Sudanese hospitals may due to the occurrence of (silent) outbreaks that need investigation and control.\n15. Could the author explain what did they mean in the discussion by: “In this study, we observed high rates of resistance to extended-spectrum B-lactamases (ESBL)”.\n16. In the discussion, the authors concluded that the resistance rate (134 gram-negative bacilli resistant to three or four classes of antibiotics isolated over a period of 6 months) is relatively higher than the rate reported in a previous study also conducted in SUH over 30 months from January 2011 to June 2013 (reference 33). The authors of reference 33 isolated 80 bacterial strains resistant to all available antibiotics including meropenem (Microbiology records SUH). Did the authors of the manuscript isolate the same number of isolates that are resistant to all available antibiotics including meropenem as stated in the last sentence, “This concerning finding...Including meropenem\")?\n17. The authors must compare their findings in the discussion to those obtained in the MENA and the surrounding region. Following are some reviews addressing resistance in such regions: Al-Orphaly et al. (2021)2, Moghnieh et al. (2018)3, and Dandachi et al. 20194.\n18. The authors should be aware when formulating their findings: is actually K. pneumoniae the most commonly isolated organism from the blood specimens among all organisms? Or within Gram-negative bacteria? The same applies to other findings? And if among all organisms, these data are not represented in the text. Please revise the text accordingly.\n19. The language is generally good in the text, however, many paragraphs and sentences are written poorly. For example:\n\"by using the following primers (Macrogen, Korea), using species-specific primers\";\n\n\"The purity and integrity of each PCR product was evaluated electrophoresis in a 2% agarose gel\";\n\nThere are also many typos in the text as (use Aztreonam instead of aztreoname, settings instead of sittings, Enterobacter cloacae instead of Enterobacter colecaes) and in Figure 1 (cotrimoxazole instead of cotramoxazole, even I prefer using the same nomenclature for antibiotics in the text as well as in the figure).\nThese aforementioned examples are not exhaustive; therefore, the text should be revised accordingly, and I suggest a language edit.\n20. Verify numbers in the text, table, and figure throughout the text. For instance, how did the authors get the number 86 isolates that are resistant to three classes of antimicrobial drugs, as well as for the 38 A. baumannii isolates that are also resistant to three classes of antimicrobial drugs? These numbers did not match table 4. Another example is the percentage of ceftazidime-resistant E. coli: in the text 84.2% vs in the figure 2 80.60%.\n21. Table 4 should be reorganized to become more visible and clearer. Simply, the authors can drop the last row (total row) because it confuses the readers about what the total row is standing for.\n22. In the text, when mentioning authors of a reference in the discussion, please write the last name of the first author (not the first name) of the citation as mentioned in the reference list. Instead of Jitendra et al., (discussion) please mention Javeri et al., as reference 28 in the reference list. Gales et al., 2010 is used to refer to reference 29, while it should be mentioned as Zavascki et al., 2010.\n23. Please capitalize the first letter in the word (gram) throughout the text.\n24. Once the full name of bacteria species is defined for the first time (ex: Klebsiella pneumoniae), use the abbreviated form (K. pneumoniae) throughout the text. The same applies to other abbreviated forms as Intensive care units, one ICU is spelled out for the first time mentioning intensive care units in the text, use the abbreviated form throughout the 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?\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? Yes",
"responses": []
},
{
"id": "86820",
"date": "17 Jun 2021",
"name": "Roberto G Melano",
"expertise": [
"Reviewer Expertise Antimicrobial resistance"
],
"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 Methods, I suggest this modification for the very long (and unclear) sentence: \"Guanidine chloride method, as described by Alsadig et al.9, was used for DNA extraction followed by PCR. Species-specific primers (Macrogen, Korea) were used for identification of Klebsiella pneumoniae, Escherichia coli, Pseudomonas aeruginosa, Acinetobacter baumannii; universal 16S rRNA primers were also used for identification of other species (Table 1). Each amplification was carried out in a total reaction volume of 25 μl: 5μl Master mix of Maxime RT premix kit (iNtRON Biotechnology, Seongnam, Korea), 0.6 μl of forward primer, 0.6 μl of reverse primer, 2μl of DNA and 16.8 μl deionized sterile water). The cycle condition were the following…”.\nIn “Antimicrobial susceptibility testing” and Fig. 1, please check the spelling of some antibiotics (e.g. amoxicillin instead of amoxicillin, etc). In Fig. 1, ‘augmentin’ is a brand name. Please, use the generic name of the drug (amoxicillin clavulanate).\nIn “Classification of MDR gram-negative Bacilli”, the definitions of MDR are not in reference 3, as mentioned by the authors. Please, add the correct reference, Magiorakos et al., (2012)1.\nIn the sentence ‘Cephalosporin resistance was defined as resistance to ceftriaxone and ceftazidime, …’, it is interesting to see that the authors did not use the optional \"or\" instead of \"and\", since resistance to these antibiotics (cephalosporins and carbapenems) can be consequence of the presence of different mechanisms. For example, Enterobacteriaceae can be resistant to ceftazidime due to the presence of ESBLs such as PER-2 or from the SHV-family, but susceptible to ceftriaxone. Same for carbapenems (e.g. deficiency of the outer membrane protein OprD confers P. aeruginosa resistance to imipenem) and aminoglycosides (different aminoglycoside modifying enzymes can confer resistance to only gentamicin or amikacin). I suggest changing these criteria, and as consequence revise the number of isolates included in this study.\nIn Results, “Bacterial identification”, I suggest changing ‘strain’ by ‘species’ in the sentence, “Isolated Gram-negative bacilli showed different strains, including…”.\nThe description in the second paragraph (‘While isolates were distributed…’) is not clear. I suggest deleting all this paragraph since Table 2 describes the same and in a very clear way. I just would keep the last 2 sentences (\"K. pneumoniae was the most isolated organism from all hospital units. The distribution of different gram-negative isolates is shown in Table 2\"). Please, use the abbreviated form of the bacterial names after these were mentioned (i.e. first mention: Klebsiella pneumoniae, after that, K. pneumoniae; same for all the species included in this manuscript).\nIt is not clear the meaning of the p-values in the Tables 2 and 3. Please, clarify.\nIn “Antimicrobial resistance pattern of clinical isolates”, Fig. 1, I don’t think the authors can present accumulative results for bacterial groups that were not tested by the same antibiotics (as showed in Figure 2). I'd delete Figure 1 and the first paragraph of this section.\nIn “Multidrug resistance patterns among gram negative Bacilli”, why were only 600 isolates were analyzed? Why were the other 134 isolates not included in this analysis? I suggest adding a sentence explaining this issue. The % of MDR (22.3%) would be an underestimation considering my comment #4 about the criteria for defining MDR.\nIn Discussion, in the paragraph starting in “Resistance of gram-negative bacilli is widespread and…”, the authors didn't study the mechanisms of resistance but the susceptibility of the isolates. Then they can't talk about high rates of ESBLs. Maybe they wanted to say \"resistance to extended-spectrum cephalosporins\" (\"resistance to ESBLs” is wrong). If that is the case, the authors have to change the sentence. I suggest: \"In this study, we observed high rates of resistance to extended-spectrum cephalosporins (ceftazidime and ceftriaxone), in addition to resistance to cephalexin, cefuroxime, ampicillin and amoxicillin/clavulanic acid.\" If there are previous studies from Sudan, and/or close countries, about AMR, it would be interesting a paragraph comparing those results with the ones obtained in this manuscript. That would show the evolution of resistance in the country/region. Please, discuss some of these data in the manuscript.\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? No source data required\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-156
|
https://f1000research.com/articles/8-151/v1
|
05 Feb 19
|
{
"type": "Method Article",
"title": "Projection layers improve deep learning models of regulatory DNA function",
"authors": [
"Alex Hawkins-Hooker",
"Henry Kenlay",
"John E. Reid",
"Henry Kenlay"
],
"abstract": "With the increasing application of deep learning methods to the modelling of regulatory DNA sequences has come an interest in exploring what types of architecture are best suited to the domain. Networks designed to predict many functional characteristics of noncoding DNA in a multitask framework have to recognise a large number of motifs and as a result benefit from large numbers of convolutional filters in the first layer. The use of large first layers in turn motivates an exploration of strategies for addressing the sparsity of output and possibility for overfitting that result. To this end we propose the use of a dimensionality-reducing linear projection layer after the initial motif-recognising convolutions. In experiments with a reduced version of the DeepSEA dataset we find that inserting this layer in combination with dropout into convolutional and convolutional-recurrent architectures can improve predictive performance across a range of first layer sizes. We further validate our approach by incorporating the projection layer into a new convolutional-recurrent architecture which achieves state of the art performance on the full DeepSEA dataset. Analysis of the learned projection weights shows that the inclusion of this layer simplifies the network’s internal representation of the occurrence of motifs, notably by projecting features representing forward and reverse-complement motifs to similar positions in the lower dimensional feature space output by the layer.",
"keywords": [
"sequence analysis",
"deep learning",
"gene regulation"
],
"content": "Introduction\n\nThe abundance of data characterising the function of noncoding DNA at high resolution facilitates the use of complex data-driven methods to learn the sequence features known as ‘motifs’ that encode this function1. A number of works have used neural networks to model human regulatory DNA, taking as input fixed-length regions of DNA sequence and predicting properties such as transcription factor binding, chromatin accessibility and histone marks using data collected by ENCODE and other consortia2–7. Several of these networks are intended to simultaneously model a wide variety of the functional characteristics of the input region, by predicting hundreds or even thousands of such measurements across multiple cell types in a multi-task learning framework. With hundreds of known regulatory motifs recorded in databases such as JASPAR8, machine learning models capable of fully characterising a significant variety of the measurable functional properties of human noncoding DNA must be able to recognise a large number of distinct patterns in the input sequence. Indeed while existing approaches have varied in the details of their neural network architectures, they have tended to share the use of relatively large numbers of convolutions as motif scanners in the first layer, and differed mainly in the subsequent layers where standard convolutions, dilated convolutions and recurrent layers have all been used to model interactions between features2–6. The best reported performance on the DeepSEA benchmark was achieved by a network having 1024 convolutional kernels in its first layer4; indeed even when experimenting with single-output networks designed to predict binding for a single transcription factor9, observed that the performance of convolutional networks continued increasing with the number of filters in the first layer up to over 100 filters1.\n\nThe use of a sufficient number of first layer filters to capture the variety of motifs relevant to the task at hand thus appears to be an important consideration in the design of neural networks for processing noncoding DNA sequences. At the same time, it raises questions. For one thing, the use of a large number of parameters in the first layer raises the possibility of overfitting. Moreover, first layers designed to recognise large numbers of specific motifs are bound to produce outputs which are relatively sparse and high-dimensional, which may hamper learning in subsequent layers10. Finally, these layers are computationally expensive, particularly when applied to long sequences, both due to the cost of computing the activations by convolving the input at each point in the sequence, and the cost in the next layer of processing sequences of high-dimensional activation vectors.\n\nStandard regularisation techniques such as dropout11 may be expected to help alleviate the problem of overfitting, and have been applied to the first convolutional layer in previous works. But there is room for further work both in terms of characterising the extent of the problem and investigating alternative solutions. Projection layers, which can be used to reduce the dimensionality of a representation without reducing its resolution, are a popular component of deep networks in computer vision where they are often referred to as 1 × 1 convolutions12,14,15. Reducing the dimensionality of a layer’s activations reduces the number of parameters required in the subsequent layer, as well as the cost of computing that layer’s activations. At the same time, depending on the nature of the features learned in the first layer, the denser representation resulting from the projection may well preserve much of the information contained therein. Even random projections are well known to preserve distances in dense representations16,17.\n\nThe common practice of including amongst the training inputs both forward and reverse-complement versions of each target sequence in particular motivates the exploration of a more compressed representation. Models are forced by this form of data augmentation to recognise distinct instances (forward and reverse-complement) of functionally equivalent motifs. Methods capable of identifying these two instantiations therefore offer the same capacity at potentially lower representational cost. Recognition of this issue has motivated the development of layers specially adapted to ensure the identity of forward and reverse-complementary sequences18. The use of projections offers an alternative approach to this problem. Here we focus on the design choices related to the capacity of multi-task networks to recognise a sufficient variety of motifs in input sequences, by jointly exploring both the effect of the number of first layer filters and the use of projection and dropout as approaches designed to mitigate the disadvantages of a large first layer. We choose to address these questions using the DeepSEA dataset2, since this has previously been used to benchmark different network architectures4,6. Initially using a reduced version of the dataset with shortened input regions, we vary the number of first layer filters for standard convolutional and convolutional-recurrent architectures with and without a projection layer and dropout, with our results indicating the importance of regularisation and the performance benefits of projection. We incorporate the projection layer into a convolutional recurrent neural network architecture with a number of modifications from the DanQ architecture proposed by 4. This new architecture achieves state of the art performance on the full DeepSEA dataset.\n\n\nMethods\n\nWe experiment with modifications to two classes of architecture which have been successfully applied for multitask prediction in regulatory genomics. Details of the hyperparameters we used when training versions of these models are provided in the sections describing the relevant experiments.\n\n1. Convolutional neural network (CNN): Both DeepSEA2 and Basset3 use 3 layer CNNs, consisting of a stack of 3 convolution and max-pooling operations followed by one or more fully connected layers. DeepSEA’s convolutional layers are regularized using dropout and a global L2 penalty, whereas Basset applies batch normalization after each convolutional layer.\n\n2. DanQ: The DanQ convolutional-recurrent architecture consists of a single convolutional layer followed by a pooling layer and a bidirectional long short-term memory (LSTM)19. The full sequence of LSTM outputs are passed through two fully connected layers in order to generate predictions4. reported results for two versions of this architecture, DanQ and DanQ-JASPAR, differing in the sizes of the layers and in the initialization used for the first layer, with half of the better-performing DanQ-JASPAR’s 1024 first-layer filters being initialized using known motifs from the JASPAR database. Like DeepSEA, both DanQ architectures use dropout after their single convolutional layer.\n\nWe investigate the use of a linear projection applied to the pooled activations of the first layer of architectures of both types. In detail, suppose that the first layer has m 1D convolutional filters and that after pooling the length of the sequence representation is l. Then the pooled activations form a sequence (a1, a2 … al) of m-dimensional vectors. The output of the projection layer is a sequence (v1, v2 … vl ) of k-dimensional vectors (k < m):\n\n\n\nwhere P is a weight matrix of size k × m. The projection layer’s output is a sequence of the same length as the sequence of the first layer’s pooled filter activations, but whose members are vectors of a lower dimension, with the same projection matrix P being used to reduce the dimension at each point in the sequence. All the results reported below were obtained using a value of k = 64, which seemed to represent a good trade-off between dimensionality reduction and preservation of information.\n\nThe best previously reported performance on the DeepSEA dataset was achieved by the DanQ-JASPAR architecture which uses a single large convolutional layer followed by a max-pooling layer with stride and pool size of 15. This layer summarises the presence of the motifs identified by the convolutional layer across relatively large 15bp stretches of input. Pooling so aggressively has the advantage of controlling the length of sequence to be fed into the LSTM, preventing computation in the recurrent layer from becoming prohibitively time consuming.\n\nWe hypothesise that this pooling involves throwing out useful positional information, which could be better preserved by splitting the downsampling across two sets of convolution and pooling layers rather than a single one. Therefore we propose an alternative convolutional recurrent (CRNN) architecture, which adds a projection layer, a second convolutional layer and a second pooling operation between the pooled outputs of the first convolutional layer and the bidirectional LSTM. To ensure fair comparison, the overall level of downsampling in the convolution and pooling layers is the same as in the DanQ-JASPAR networks, such that the length of the sequence of inputs to the bidirectional LSTM is the same (64) in both cases. In common with the DanQ networks we use a single fully-connected hidden layer before the output layer, but in order to control overfitting we use as input to this layer not the full sequence of LSTM outputs but their global mean. The proposed network, full details of which are given below, has far fewer parameters than DanQ-JASPAR and trains faster.\n\nThe DeepSEA dataset. The DeepSEA dataset consists of sequences of 1000 bp from the human noncoding genome, labelled for the presence of a peak in the central 200 bp in the signal for each of 919 chromatin features taken from ENCODE and Roadmap20,21. These features represent a range of transcription factor binding, chromatin accessibility and histone modification measurements across a variety of cell types. Both forward and reverse-complement versions of the sequence corresponding to each set of targets are included in the dataset, meaning that models must be capable of learning both forward and reverse-complement motifs. We use the original training, validation and test splits and follow4 in using as our primary evaluation metric test set area under the precision recall curve (AUPRC), which is calculated after averaging predictions across forward and reverse complement versions of each sequence.\n\nDesign choices related to first layer on reduced DeepSEA dataset. In our first set of experiments we seek to rigorously explore the optimal configuration of the early layers of instantiations of both CNN and DanQ network designs. We vary the number of first layer filters, the use of dropout immediately after the first pooling layer, and the use of a projection layer (we fix the output dimension of this layer at each point in the sequence to 64) while keeping other hyperparameters fixed for a version of each class of architecture. When dropout and projection are used together, the dropout is applied after the projection layer. A dropout rate of 0.2 is used in all cases, which is the same as that applied to the activations of the first convolutional layer in both the DeepSEA and DanQ architectures. Other modifications to the original architectures were made in the interests of retaining comparable performance while reducing computational cost and are described below.\n\nThe CNN model that we choose to explore here takes from Basset the use of 3 convolutional layers, with kernel sizes of 19, 11 and 7 respectively, but varying in several other details. We use max pooling operations of sizes 6, 2, and 2 after the convolutional layers. The number of filters in the second and third convolutional layers is held fixed at 128 and 256 respectively. The outputs of the final pooling operation are fed into a single hidden layer of 2048 neurons to which dropout with dropout factor of 0.5 is applied. Leaky ReLUs22 are used for all activations. Our DanQ architectures follow the original in most details other than those under investigation, except for the use of Leaky ReLU rather than ReLU activations, and the use of a reduced number of LSTM cells (100) in each direction. To mitigate the cost of these experiments, we run them on a reduced version of the DeepSEA dataset, using only the central 500 bp of each 1000 bp sequence. For all networks we use the Adam optimizer23 with an initial learning rate of 3 × 10−4 to minimize the multitask binary cross entropy loss via mini-batch gradient descent with a batch size of 256. The learning rate was reduced by a factor of 5 if the validation loss did not decrease for two epochs. Training was terminated if the validation loss did not improve for five epochs. All models were implemented in Keras24 using the Theano backend25.\n\nEvaluation of CRNN architecture on full DeepSEA dataset. For the second set of experiments we use the full 1000 bp for each sequence and seek to compare the performance of our improved CRNN architecture to that of DeepSEA and the two DanQ architectures. For comparison with the two variants of DanQ, DanQ and DanQ-JASPAR, which have, respectively, 320 and 1024 filters in the first layer, we explore two variants of our CRNN architecture with 320 and 700 filters of length 30 in the first layer. To evaluate the contribution of the projection layer, for each CRNN variant we train one network with projection after the first pooling operation, and one network without projection but otherwise identical to the first. All networks use a second convolutional layer with 128 filters of length 11 whose activations are pooled and fed into a bidirectional LSTM with 300 units in each direction. Max-pooling with stride and pool size of 7 after the first convolutional layer and 2 after the second convolutional layer together with unpadded convolutions ensure that the sequence of inputs to the LSTM is of the same length as in the DanQ-JASPAR model. Dropout with a rate of 0.15 is applied to the projected first layer activations if projection is used, and to the pooled first layer activations if not. Recurrent dropout26 with a rate of 0.2 is applied to the LSTM. Leaky ReLUs are used for all layer activations. Networks are trained using the same learning rate schedules as in the previous set of experiments. We compare the average test set AUPRCs of our models with those of the publicly available trained DeepSEA and DanQ networks.\n\nThe source code for all models and experiments is available on GitHub and Zenodo27.\n\n\nResults\n\nIn both fully convolutional and convolutional-recurrent architectures consistent benefits were achieved by increasing the number of first layer filters, with gradual saturation of performance (as measured by test set AUPRC averaged across the tasks) at around 1000 filters in both cases (Figure 1). In the fully convolutional networks the benefit of the projection layer was very clear, with all networks which used projection outperforming those that didn’t, often by considerable margins. A combination of dropout and projection achieved the best performance in every case. There is less evidence of benefit in the case of the networks using DanQ-style architectures, with networks with regularisation sometimes outperforming those without, but a lack of a clear pattern in the results, at least under the test set AUPRC metric. This is despite models incorporating dropout and the projection layer consistently achieving lower cross-entropy loss on the validation set. One factor in the difference between the two types of architectures is the degree of overfitting that the standard, unregularised architecture suffers. We observed that fully convolutional architectures showed a much greater tendency to overfit than convolutional-recurrent architectures (Figure 2). We note that unlike a convolutional layer, an LSTM already learns its own projection in the form of the weight matrix which transforms the inputs into the internal state space within the input and forget gates. These internal projections may help reduce both the tendency to overfit and the potential performance improvement associated with incorporating an additional projection layer. In contrast, inserting a projection layer into a CNN architecture substantially reduces the degree of overfitting (Figure 2), which allows CNN networks including projection layers continue to benefit from adding additional filters in the first layer, whereas without projection, CNN performance hardly improves beyond 500 first layer filters, as the benefit of extra feature detectors is offset by the increased likelihood of overfitting.\n\nJitter was added to the number of first layer filters for DanQ architectures to enable the points to be distinguished.\n\nThe CNN network shows much more evidence of overfitting.\n\nTable 1 shows the cross entropy losses on the validation and test sets for our best-performing convolutional recurrent (CRNN) models as well as published baselines. CRNN-700 achieves the best average test set AUPRC of the compared models while being significantly less costly to train than DanQ-JASPAR, and without requiring the use of any known motifs to initialize first layer filters, as DanQ-JASPAR does. For both CRNN models we also compare the performance of models with and without the projection layer. In both cases, the projection layer leads to a clear increase in performance and a reduction in the cost per epoch of training the network.\n\nCRNN-n is a model with 2 convolutional layers with n and 128 filters respectively, with kernel sizes of 30 and 11, followed by a bidirectional LSTM with 300 units in each direction, whose outputs are averaged and fed through a hidden layer with 919 units which in turn feeds into the output layer. CRNN-n-projection is identical to CRNN-n except for the inclusion of a projection layer between the first and second convolution layers, which effectively reduces the dimension of the first layer’s activations from n to 64. Losses and AUPRCs for DanQ and DeepSEA networks are calculated using the publicly available model weights files. AUPRCs for all models are calculated after averaging predictions for forward and reverse complement versions of each test sequence, whereas forward and reverse complement versions of each sequence contribute independently to the reported losses.\n\nTo understand the nature of the performance benefits brought by the use of the projection layer, we can investigate the relationship between the projection weights learned and the motifs learned by the first convolutional layer. To associate a motif with each filter in the first layer we follow a procedure similar to that introduced by 1: several thousand sequences from the training set are passed through the trained model, and for each first layer convolutional filter we record the identities of the nucleotides at each position in the maximally-activating stretch of input in each sequence in which that filter is activated. From this we construct a PFM which can be converted into a motif representing the typical input pattern recognised by the filter. Using TOMTOM28 to search the JASPAR 2018 database8 we find that 257 of the 700 learned motifs of the best-performing CRNN-700-projection model have at least one significant match (q < 0.01). Each learned motif is also associated with one of the columns in the 64 × 700 weight matrix of the projection layer. Suppose for example that at a certain point in an input sequence, the motif recognised by the ith convolutional filter occurs. Assuming none of the other filters are activated by this motif or its neighbouring region, the network’s representation of this region of the input will then just be the vector obtained by multiplying each weight in the ith column of the projection matrix by the filter’s activation. Thus the ith column of the projection matrix can be interpreted as representing an embedding of the motif learned by the ith convolutional filter. To visualise these embeddings, we choose to focus on a subset of the learned motifs which have the best matches to known motifs, selecting only the 44 learned motifs with q-values less than 10−8. The result of performing a PCA on the 44 columns of the projection weight matrix associated with these motifs is shown in Figure 3. Most strikingly, different versions of the same motif tend to cluster together, with the embeddings for filters which learn to recognise the forward version of a particular motif very often close to those for filters which recognise the reverse complement of the same motif. This suggests that the projection layer allows for a more efficient internal representation of motifs, recognising that forward and reverse complement patterns are functionally equivalent although completely different and therefore requiring different feature extractors at the sequence level. This representation of functional equivalence allows networks with a projection layer to harness the benefits of reverse-complement data augmentation without paying a price in terms of representational complexity.\n\nEach point represents one of the 64 dimensional column vectors of the projection weight matrix. Only columns corresponding to learned motifs with a match with q -value less than 10−8 are included in the PCA to aid visualisation. Points are labelled by name of matched motif and whether it is the forward (f) or the reverse complement (rc) version of the known motif that is matched. Points are coloured by transcription factor family (cyan: C2H2 zinc finger, green: basic leucine zipper, red: homeodomain, purple: basic helix-loop-helix, blue: all other).\n\n\nDiscussion\n\nDespite the recent progress in the application of deep learning methods to model genomic data there remains work to be done in understanding the types of architecture and design choices best suited to the domain. We provide further evidence here that the performance of networks whose goal is to predict hundreds of functional properties from the DNA sequence is strongly dependent on the number of convolutional filters in the first layer. In networks where the subsequent layer is also convolutional, performance can be further improved by inserting a dimensionality-reducing projection layer between the two sets of convolutions. A similar use of projection layers in networks designed to predict enhancers was independently proposed by29 while we were finalising this manuscript. Their network takes as input both DNA sequences and chromatin accessibility information, and intersperses projections and convolutions on each of the two data modalities. While their work shows that projections can be used in highly performing architectures for regulatory genomics problems, they did not explore the role of projections in achieving this performance. Here our aim is to draw particular attention to the performance benefits and mode of functioning of a single projection layer, inserted directly after a first DNA motif-recognising convolutional layer, since we believe these point to its potential utility beyond any single application. In particular, we show that the projection layer is capable of learning the identity between forward and reverse-complement versions of functionally equivalent motifs and thereby simplifying the representation of the functional content of the sequence. It also reduces the number of parameters required in the subsequent layer, leading to less overfitting (particularly in combination with dropout) and reducing the computational cost. Incorporating the projection layer into a convolutional-recurrent network architecture similar to the DanQ architecture leads to improved performance on the DeepSEA dataset with fewer parameters and shorter per-epoch training times. Although we have only tested the use of the projection layer on the DeepSEA dataset, we believe that its use could be of important benefit in other situations in which accurate prediction of the targets requires recognition of a large variety of motifs in the input sequence.\n\n\nData availability\n\nDeepSEA dataset: the full dataset, including train, test and validations splits, may be downloaded from http://deepsea.princeton.edu/help/.\n\n\nSoftware availability\n\nSource code for models and experiments available at: https://github.com/alex-hh/motif_projection.\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.254379629.\n\nLicense: MIT license.",
"appendix": "Author contributions\n\n\n\nAHH proposed the use of the projection layer, implemented and trained the models, ran the evaluations and wrote the paper. HK prototyped, implemented and trained early versions of the models and designed the evaluations. JR initiated the project, helped with the interpretation of the results and helped write the paper.\n\n\nGrant information\n\nThis work was funded by the UK Medical Research Council (Grant Ref MC_U105260799).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nFootnotes\n\n1 For comparison, top performing networks on ImageNet12,13 make do with only 64 filters in the first layer despite the output dimension being comparable to that of DeepSEA. This discrepancy may perhaps be explained by the fact that while the features learned by the first layers of image processing networks are small, generic and only gradually composed into more specific features by subsequent layers, the features learned by the first layers of networks in regulatory genomics are by design highly specific motifs, typically 10–20bp in length.\n\n\nReferences\n\nAlipanahi B, Delong A, Weirauch MT, et al.: Predicting the sequence specificities of DNA- and RNA-binding proteins by deep learning. Nat Biotechnol. 2015; 33(8): 831–8. PubMed Abstract | Publisher Full Text\n\nZhou J, Troyanskaya OG: Predicting effects of noncoding variants with deep learning-based sequence model. Nat Methods. 2015; 12(10): 931–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKelley DR, Snoek J, Rinn JL: Basset: Learning the regulatory code of the accessible genome with deep convolutional neural networks. Genome Res. 2016; 26(7): 990–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQuang D, Xie X: DanQ: a hybrid convolutional and recurrent deep neural network for quantifying the function of DNA sequences. Nucleic Acids Res. 2016; 44(11): e107. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKelley DR, Reshef YA: Sequential regulatory activity prediction across chromosomes with convolutional neural networks. Genome Res. 2018; 28(5): 739–750. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGupta A, Rush AM: Dilated convolutions for modeling long-distance genomic dependencies. ArXiv e-prints. 2017. Reference Source\n\nZhou J, Theesfeld CL, Yao K, et al.: Deep learning sequence-based ab initio prediction of variant effects on expression and disease risk. Nat Genet. 2018; 50(8): 1171–1179. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKhan A, Fornes O, Stigliani A, et al.: JASPAR 2018: update of the open-access database of transcription factor binding profiles and its web framework. Nucleic Acids Res. 2018; 46(D1): D1284. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZeng H, Edwards MD, Liu G, et al.: Convolutional neural network architectures for predicting DNA-protein binding. Bioinformatics. 2016; 32(12): i121–i127. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBengio Y, Ducharme R, Vincent P, et al.: A neural probabilistic language model. J Mach Learn Res. 2003; 3: 1137–1155. Reference Source\n\nSrivastava N, Hinton G, Krizhevsky A, et al.: Dropout: A simple way to prevent neural networks from overfitting. J Mach Learn Res. 2014; 15(1): 1929–1958. Reference Source\n\nHe K, Zhang X, Ren S, et al.: Deep residual learning for image recognition. In 2016 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016, Las Vegas, NV USA, June 27-30, 2016. 2016; 770–778. Publisher Full Text\n\nSimonyan K, Zisserman A: Very deep convolutional networks for large-scale image recognition. CoRR. 2014. Reference Source\n\nSzegedy C, Liu W, Jia Y, et al.: Going deeper with convolutions. CoRR. abs/1409.4842, 2014. Reference Source\n\nLin M, Chen Q, Yan S: Network in network. CoRR. 2013. Reference Source\n\nJohnson W, Lindenstrauss J: Extensions of Lipschitz mappings into a Hilbert space. In Conference in modern analysis and probability (New Haven, Conn., 1982), volume 26 of Contemporary Mathematics. American Mathematical Society, 1984; 189–206. Publisher Full Text\n\nBingham E, Mannila H: Random projection in dimensionality reduction: Applications to image and text data. In Proceedings of the Seventh ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, KDD ’01, New York, NY, USA, 2001; 245–250. Publisher Full Text\n\nShrikumar A, Greenside P, Kundaje A: Reverse-complement parameter sharing improves deep learning models for genomics. bioRxiv. 2017. Reference Source\n\nGraves A, Schmidhuber J: Framewise phoneme classification with bidirectional LSTM and other neural network architectures. Neural Netw. 2005; 18(5–6): 602–610. Publisher Full Text\n\nThe ENCODE Project Consortium: An integrated encyclopedia of DNA elements in the human genome.Nature. 2012; 489(7414): 57–74. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRoadmap Epigenomics Consortium, Kundaje A, Meuleman W, et al.: Integrative analysis of 111 reference human epigenomes. Nature. 2015; 518(7539): 317–330. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaas AL, Hannun AY, Ng AY: Rectifier nonlinearities improve neural network acoustic models. In: in ICML Workshop on Deep Learning for Audio, Speech and Language Processing. 2013. Reference Source\n\nKingma DP, Ba J: Adam: A method for stochastic optimization. CoRR. 2014; abs/1412.6980. Reference Source\n\nChollet F, et al.: Keras. 2015. Reference Source\n\nTheano Development Team: Theano: A Python framework for fast computation of mathematical expressions. arXiv e-prints. 2016; abs/1605.02688. Reference Source\n\nGal Y, Ghahramani Z: A theoretically grounded application of dropout in recurrent neural networks. In: Lee DD, Sugiyama M, Luxburg UV, et al. editors, Advances in Neural Information Processing Systems 29. Curran Associates, Inc., 2016; 1019–1027. Reference Source\n\nAlex: alex-hh/motif projection preprint. (Versionpreprint). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.2543797\n\nGupta S, Stamatoyannopoulos JA, Bailey TL, et al.: Quantifying similarity between motifs. Genome Biol. 2006; 8(2): R24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen S, Gan M, Lv H, et al.: DeepCAPE: a deep convolutional neural network for the accurate prediction of enhancers. bioRxiv. 2018. Publisher Full Text"
}
|
[
{
"id": "44128",
"date": "18 Feb 2019",
"name": "David R. Kelley",
"expertise": [
"Reviewer Expertise Regulatory sequence 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\nSummary The authors suggest the use of “projection” layers for modeling the regulatory activity of DNA sequences. “Projection” layers perform a linear transformation from a large representation size to a smaller one, by applying a width one convolution. They present experiments with two previously studied architectures in which the addition of these layers improves performance. They suggest that inserting this layer after the initial convolution layer, which most closely resembles position weight matrices, enables the model to more easily capture the similarity of forward and reverse complement motifs for some tasks.\nThis technique has been studied in the larger neural network literature, often by the name of bottleneck layers. The authors should consider citing Lin et al. Network in network (https://arxiv.org/abs/1312.4400) for readers who would like to learn more about the technique.\nThe experiments appear solid and align with my own experiences using similar layers. One difference with my previous experiments is that I've always applied a nonlinearity after the \"projection\", effectively making it a more standard neural network layer. The authors might consider benchmarking this version.\nIn addition to the projection layers, the authors observation that the final prediction layers for these tasks can use drastically fewer parameters by applying a global average pool is insightful.\nOverall, I expect this report will be useful to future practitioners of neural networks for DNA sequence 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": "55081",
"date": "21 Oct 2019",
"name": "Qin Ma",
"expertise": [
"Reviewer Expertise Bioinformatics",
"Computational Systems 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 proposed a novel approach to model the regulatory activity of DNA sequences and, by incorporating the projection layer into a new convolutional-recurrent architecture. The proposed method obtains state-of-the-art performance on the full DeepSEA dataset. In the experiment process, the authors explore what effect the projection layer and dropout rate have on DanQ, CNN and CRNN models, and explore the role of projections in achieving this performance. Meanwhile, applying the global average pool to reduce the number of parameters, which is a method to save training time and memory. The key contribution of this article is that the projection layer and dropout rate are utilized to model regulatory DNA, which provides us a new idea about how to mitigate the disadvantage s of a large first layer. The paper is well-organized and provides new insight into modeling the regulatory activity of DNA sequences, however, there are still some improvements needed.\nMajor comment 1: It is not clear to me how many times the models in Table 1 are tested. Meanwhile, regarding the results shown as figure 1 and figure 2, we usually test more times and give the boxplot of the results, to avoid the randomness.\n\nMajor comment 2: The authors should descript how to optimize their model, in detail.\n\nMajor comment 3: It is awesome that the 257 of the 700 learned motifs are obtained by the CRNN-700-projection model. If some of the learned motifs are shown in the main text, this article will be more attracting. Meanwhile, the DanQ model can also be used to extract motifs, this model should be compared to the CRNN-700-projection model.\n\nMinor comment 1: Please, ensure what the ‘4’ means, in the penult line of the Introduction.\n\nMinor comment 2: Please, provide the expansion of PAC and PFM.\n\nMinor comment 3: The flowchart can clarify the authors’ method, why not consider adding the flowchart of your approach.\n\nIs the rationale for developing the new method (or application) clearly explained? Yes\n\nIs the description of the method technically sound? Partly\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": []
}
] | 1
|
https://f1000research.com/articles/8-151
|
https://f1000research.com/articles/8-147/v1
|
04 Feb 19
|
{
"type": "Research Article",
"title": "Effects of exercise on heart rate variability by time-domain, frequency-domain and non-linear analyses in equine athletes",
"authors": [
"Ka Hou Christien Li",
"Rachel Wing Chuen Lai",
"Yimei Du",
"Vivian Ly",
"David Chun Yin Li",
"Michael Huen Sum Lam",
"Leonardo Roever",
"Sophia Fitzgerald-Smith",
"Guangping Li",
"Tong Liu",
"Gary Tse",
"Mei Dong",
"International Health Informatics Study (IHIS) Network",
"Ka Hou Christien Li",
"Rachel Wing Chuen Lai",
"Yimei Du",
"Vivian Ly",
"David Chun Yin Li",
"Michael Huen Sum Lam",
"Leonardo Roever",
"Sophia Fitzgerald-Smith",
"Guangping Li",
"Tong Liu"
],
"abstract": "Background: Heart rate variability (HRV) is an intrinsic property that reflects autonomic balance and has been shown to be predictive of all-cause and cardiovascular mortality. It can be altered by physiological states such as exercise or pathological conditions. However, there are only a handful of studies on HRV in horses. The aim of this study is to compare HRV parameters before and during exercise in horses. Methods: Time-domain, frequency-domain and non-linear analyses were applied to quantify time series data on RR intervals before and during exercise in horses (n=7). Results: Exercise increased heart rate from 44±8 to 113±17 bpm (ANOVA, P<0.05) and decreased standard deviation (SD) from 7±2 to 4±2 bpm, coefficient of variation (CoV) from 16±4% to 3±2% and root mean square of successive RR interval differences (RMSSD) from 89.4±91.5 to 6.5±3.7 ms. Contrastingly, no difference in low-frequency (0.10±0.03 vs. 0.09±0.03 Hz) and high-frequency (0.19±0.03 vs. 0.18±0.03 Hz) peaks, nor in their percentage powers (2±1 vs. 4±5%; 59±9 vs. 64±20%; 39±10 vs. 32±19%) were observed but very low-frequency, low-frequency, and high-frequency powers (ms2) were reduced from 29±17 to 2±5, 1138±372 to 22±22 and 860±564 to 9±6, respectively, as was total power (in logarithms) (7.52±0.52 to 3.25±0.73). Poincaré plots of RRn+1 against RRn revealed similar ellipsoid shapes before and after exercise. The SD along the line-of-identity (SD2) and SD perpendicular to the line-of-identity (SD1) were decreased by exercise (62±17 vs. 9±5 and 63±65 vs. 5±3), corresponding to increased SD2/SD1 ratio from 1.33±0.45 to 2.19±0.72. No change in approximate and sample entropy was detected (0.97±0.23 vs. 0.82±0.22 and 1.14±0.43 vs. 1.37±0.49). Detrended fluctuation analysis revealed unaltered short-term fluctuation slopes (0.76±0.27 vs. 1.18±0.55) but increased long-term fluctuation slopes (0.16±0.11 vs. 0.50±0.16) after exercise. Conclusion: Exercise leads to a decrease in HRV but did not affect signal entropy in horses.",
"keywords": [
"heart rate variability",
"time",
"frequency",
"non-linear",
"entropy",
"horses",
"equine"
],
"content": "Introduction\n\nBeat-to-beat alterations in heart rate, termed heart rate variability (HRV), is an intrinsic property that may be affected by distinct physiological states, such as exercise. Exercise-related sudden cardiac arrest remains a significant problem, affecting young athletes who are otherwise healthy with structurally normal hearts. Previously, our group has employed mouse models for studying cardiac electrophysiological properties1. However, these bear the disadvantages of having important differences compared to humans, such as a ten-fold higher heart rates and the lack of a plateau phase during repolarization.\n\nTherefore, the use of alternative animal models offers an important and complementary approach. For example, horses show similar heart rate patterns compared to humans, with capacity of increasing from 40 bpm to 200 bpm during exercise. This suggests that horses could be useful for exploring the consequences of exercise-induced electrophysiological changes2. However, few studies in the literature have quantified HRV in horses. We quantified beat-to-beat variability in heart rate by applying time-domain, frequency-domain and non-linear techniques for the first time to electrocardiograms obtained from horses at rest and during exercise.\n\n\nMethods\n\nHeart rate data of horses were obtained from a publicly available dataset published online with the study protocol previously described3. Briefly, healthy Thoroughbred horses in race training presented for workups at an equine hospital were screened. This yielded a total of seven horses with electrocardiographic recordings for 10 to 18 minutes. HRV analysis was performed using Kubios HRV Standard software (Version 3.0.2). The following time-domain measures were obtained: 1) mean RR interval; 2) standard deviation (SD) of RR intervals; 3) coefficient of variation (CoV) for RR intervals, 4) root mean square (RMSSD) of successive differences of RR intervals; 5) NN50, number of successive RR interval pairs that differ more than 50 ms; 6) pNN50, relative number of successive RR interval pairs that differ more than 50 ms; 7) HRV triangular index, the integral of the RR interval histogram divided by the height of the histogram; 8) Triangular interpolation of normal-to-normal intervals (TINN). 9) mean HR; 10) SD of HR; 11) CoV for HR; 12) minimum HR; 13) maximum HR.\n\nFrequency-domain analysis was performed using the Fast Fourier Transform method4 with sampling frequency set at 8 Hz. The power in the repolarization spectrum between 0.04 and 0.4 Hz was defined as total power (TP). The power in the heart rate spectrum was divided into three different frequency bands: very low frequency power (VLF, 0 to 0.04 Hz), low frequency power (LF, 0.04 to 0.15 Hz) and high frequency power (HF, 0.15 to 0.4 Hz).\n\nNon-linear properties of HRV were studied as follow. Poincaré plots are graphical representations of the correlation between successive RR intervals, in which RRn+1 is plotted against RRn. From this plot, the SD of the points perpendicular to the line-of-identity (SD1) describing short-term variability, and the SD of the points along the line-of-identity (SD2) describing the long-term variability, can be determined. The SD2/SD1 ratio is a measure of long-term variability relative to the short-term variability. The approximate entropy provides a measure of the irregularity of the signal. It is computed as follows:\n\nFirstly, a set of length m vectors uj is formed:\n\nuj = (RRj ; RRj+1,…, RRj+m-1); j = 1; 2; …N – m + 1\n\nwhere m is the embedding dimension and N is the number of measured RR intervals. The distance between these vectors is defined as the maximum absolute difference between the corresponding elements:\n\nd(uj, uk) = max {|RRj+n – RRk+n| | n=0, …, m-1}\n\nfor each uj the relative number of vectors uk for which d(uj, uk) ≤ r is calculated. This index is denoted with Cmj (r) and can be written in the form\n\nCjm(r)=nbrof{uk|d(uj,uk)≤r}N−m+1∀k\n\nTaking the natural logarithms gives:\n\nΦm(r)=1N−m+1∑j=1N−m+1lnCjm(r).\n\nThe approximate entropy is then defined as:\n\nApEn(m,r,N)=Φm(r)−Φm+1(r)\n\nLower approximate entropy values reflect a more regular signal, whereas higher values reflect a more irregular signal.\n\nThe sample entropy also provides a measure of signal irregularity but is less susceptible to bias compared to approximate entropy. This is given by:\n\nCjm(r)=nbrof{uk|d(uj,uk)≤r}N−m∀k≠j\n\nAveraging then yields:\n\nCm(r)=1N−m+1∑j=1N−m+1Cjm(r)\n\nThe sample entropy is then given by:\n\nSampEn(m,r,N)=ln(Cm(r)Cm+1(r))\n\nFinally, detrended fluctuation analysis (DFA) was performed to determine long-range correlations in non-stationary physiological time series5, yielding both short-term fluctuation (α1) and long-term fluctuation (α2) slopes. The point at which the slopes α1 and α2 is the crossover point.\n\nStatistical analyses were conducted using Origin Pro 2017. All values were expressed as mean ± standard deviation (SD). Numerical data were compared by one-way analysis of variance (ANOVA). P<0.05 was considered statistically significant and was denoted by * in the figures.\n\n\nResults\n\nRepresentative time series data for RR intervals at rest and during exercise from a single horse are shown in Figure 1A and Figure 1B, respectively, with their frequency distributions shown in Figure 1C and Figure 1D. Their corresponding heart rate time series are shown in Figure 2A and Figure 2B, and frequency distributions in Figure 2C and Figure 2D. Under resting conditions, the mean RR interval was 1392±224 ms, decreasing to 541±84 with exercise (n=7 horses) (Figure 3A). The mean standard deviation (SD) was 65±43 ms (Figure 3B) and coefficient of variation (CoV) was 5±2% (Figure 3C), which decreased to 8±4 ms and 1±1%, respectively, after exercise (ANOVA, P<0.05). Similarly, the root mean square of successive RR interval differences (RMSSD) decreased from 89±92 to 6±4 ms (Figure 3D). The number of interval differences of successive NN intervals greater than 50 ms (NN50) and the proportion derived by dividing NN50 by the total number of NN intervals (pNN50) were decreased from 85±55 to 0±1 ms and from 21±12 to 0±1%, respectively (ANOVA, P<0.05). The HRV triangular index, which is integral of the RR interval histogram divided by the height of the histogram, also decreased from 8.37±2.87 to 2.03±0.45 (ANOVA, P<0.05), as was the triangular interpolation of normal-to-normal intervals (TINN), the baseline width of the RR interval histogram (445± 240 to 40±26 ms; ANOVA, P<0.05). These corresponded to a mean heart rate of 44±8 bpm (Figure 4A), SD was 7±2 bpm (Figure 4B) and CoV (Figure 4C) of 16±4%. With exercise, HR increased to 113±17 bpm, whereas SD, CoV and RMSSD decreased to 4±2 bpm, 3±2% and 6±4 ms, respectively. Finally, the minimum and maximum heart rates (Figure 4D and 4E) at rest were 34±5 and 79±16 bpm, respectively, increasing to 101±16 and 129±15 bpm.\n\nRepresentative time series data for RR intervals at rest (A) and during exercise (B) and the corresponding histograms (C and D) from a single horse.\n\nRepresentative time series data for heart rates at rest (A) and during exercise (B) and the corresponding histograms (C and D) from a single horse.\n\nTime-domain analysis (n=7 horses) yielding mean RR intervals (A), standard deviation (SD) of RR intervals (B), coefficient of variation (CoV) given by SD/mean x 100% (C), root mean square of successive RR interval differences (RMSSD) (D), number of interval differences of successive NN intervals greater than 50 ms (NN50) (E), proportion derived by dividing NN50 by the total number of NN intervals (pNN50) (F), heart rate variability triangular index (HRVTI), the integral of the RR interval histogram divided by the height of the histogram (G) and triangular interpolation of normal-to-normal intervals (TINN), the baseline width of the RR interval histogram (H).\n\nTime-domain analysis yielding mean heart rate (HR) (A), standard deviation (SD) of HR (B), coefficient of variation (CoV) given by standard deviation (SD)/mean x 100% (C), minimum HR (D) and maximum HR (E).\n\nNext, the Fast Fourier Transform method was used for frequency-domain analyses. An example of the power spectrum plot against frequency before and after exercise is shown in Figure 5A and 5B, respectively. Strikingly, the peaks for very low-, low- and high-frequency were not altered by exercise (0.04 ± 0.00 vs. 0.04 ± 0.00 Hz; 0.10 ± 0.03 vs. 0.09±0.03 Hz; 0.19 ± 0.03 vs. 0.18±0.03 Hz, respectively) (Figures 5C to E). Similarly, their percentage powers remained unchanged (2±1 vs. 4±5%; 59±9 vs. 64±20%; 39±10 vs. 32±19%, respectively, P>0.05) (Figures 5F to 5H). By contrast, very low-frequency, low-frequency, and high-frequency powers were significantly reduced from 29±17 to 2±5 ms2, 1138±372 to 22±22 ms2 and 860±564 to 9±6 ms2, respectively (P<0.05) (Figures 6A to 6C), as was the total power (in logarithms) from 7.52±0.52 to 3.25±0.73 (P<0.05) (Figures 6D).\n\nExamples of frequency spectra using the Fast Fourier Transform method for RR time series obtained before (A) and after (B) exercise. Peaks for very low-frequency (VLF) (C), low-frequency (LF) (D) and high-frequency (HF) (E) and their percentage powers (F to H).\n\nVery low-frequency (VLF) (C), low-frequency (LF) (D) and high-frequency (HF) (E) powers and total power in logarithms (D).\n\nPoincaré plots expressing RRn+1 as a function of RRn were constructed, with typical examples from a single horse before and after exercise shown in Figure 7A and 7B. In all of the hearts studied, circular shapes of the data points were observed. The SD perpendicular to the line-of-identity (SD1) and SD along the line-of-identity (SD2) are shown in Figure 7C and 7D, respectively. The mean SD1 and SD2 were 63±65 and 62±17, respectively, corresponding to a SD2/SD1 ratio of 1.33±0.45 (Figure 7E). After exercise, SD1 and SD2 decreased significantly to 5±3 and 9±5, respectively, which corresponded to increased SD2/SD1 ratio to 2.19±0.72 (P<0.05). Moreover, approximate and sample entropy took values of 0.97±0.23 (Figure 7F) and 1.14±0.43 (Figure 7G), respectively. These values were not altered after exercise (0.82±0.22 and 1.37±0.49, respectively; P>0.05). Detrended fluctuation analysis plotting the detrended fluctuations F(n) as a function of n in a log-log scale was performed for the RR intervals (Figure 8A and 8B). This revealed short- (α1) and long-term (α2) fluctuation slopes of 0.76±0.27 (Figure 8C) and 0.16±0.11 (Figure 8D) before exercise. After exercise, α1 was not significantly different (1.18±0.55; P>0.05) but α2 was significantly increased to 0.50±0.16 (P<0.05).\n\nRepresentative Poincaré plots of RRn+1 against RRn before (A) or after exercise (B) from a single horse. Summary data (n=7) for standard deviation (SD) along the line-of-identity (SD1) (C) and SD perpendicular to the line-of-identity (SD2) (D), and the SD2/SD1 ratio (E), approximate entropy (ApEn) (F) and sample entropy (Samp En) (G).\n\nDetrended fluctuation analysis (DFA) plots expressing detrended fluctuations F(n) as a function of n in a log-log scale before (A) and after exercise (B), yielding short-term (C) and long-term (D) fluctuation slopes (α1 and α2, respectively).\n\n\nDiscussion\n\nIn this study, we investigated HRV of equine athletes before and after exercise. The main findings are that 1) variability in heart rate can be detected using time-domain, frequency-domain and non-linear methods; 2) exercise led to reduced HRV as revealed using time-domain, frequency-domain and linear methods, 3) no change in signal entropy was observed after exercise.\n\nThe heart shows variability in their electrical signals both spatially and temporally6–8, and this signal variability can be detected at different levels of complexity, from whole organs down to single ion channels9–11. A certain degree of HRV is present in normal, healthy individuals12–16. However, it can become altered in pathological states and in turn associated with atrial fibrillation, ventricular arrhythmias and sudden cardiac arrest14. HRV has been investigated in the context of aging17, massage18 and pregnancy19. To date, there have only been a handful of studies that have evaluated HRV from equine athletes in the context of exercise20–34. Consistent with previous findings22,29,30, we found that exercise decreased HRV as determined by time-domain methods. Thus, SD of RR intervals, CoV, RMSSD, NN50, pNN50, HRV triangular index, and TINN were all significantly reduced.\n\nFrom frequency-domain analysis, we did not detect significant changes in the low- and high-frequency peaks, or their percentage powers after exercise. These findings are in contrast to some of the previously reported findings. Thus, an increase in high-frequency peak29, higher low-frequency and lower high-frequency components30, and increased high-frequency to low-frequency ratios during gallop31 were observed. By contrast, we observed significantly reductions in the absolute values for very low frequency power, lower frequency power and high frequency power and total power but no change in low frequency/high frequency power ratio after exercise. These findings are consistent with previous demonstrations that power in all bands was reduced by exercise31.\n\nSignificantly, non-linear analyses of RR intervals yielded further insights. Thus, Poincaré plots showed ellipsoid shapes in all of the horses studied at rest with SD2/SD1 ratio close to 1, suggesting similar short-term and long-term variability. By contrast, exercise led to significant decreases in both SD1 and SD2, but an increase in SD2/SD1 ratio, indicated greater long-term variability under these stressed conditions. These findings are consistent with a recent report showing significant reductions in both SD1 and SD222. Furthermore, the present findings also quantified approximate and sample entropy for the first time in equine athletes, demonstrating a degree of entropy present. Entropy is the amount of disorder in a given system and reflects the signal regularity or complexity35,36. However, this was not altered by exercise.\n\nFinally, detrended fluctuation analysis (DFA) was applied, to the best of our knowledge, for the first time in equine athletes. Previosu studies have used this method for investigating long-range correlations in non-stationary physiological time series5. In DFA, the mean fluctuation is plotted against the number of beats on a double logarithmic scale. This would then yield the scaling exponents, α1 and α2, respectively. For uncorrelated data, α takes a value of 0.5. By contrast, the presence of correlation will be reflected by α taking values below or above 0.5. In our study, α1 was 0.76±0.27 and α2 was 0.16±0.11 before exercise. With exercise α2 was increased to 0.50±0.16, suggesting that the long-term correlation was lost. Previous studies have applied DFA to HRV in other species. For example, in rabbits with hypertrophic cardiomyopathy, greater values of the scaling exponent were observed compared to those with the disease37. Moreover, in humans, a reduction in α1 was found during sympathetic activation, indicating a breakdown of the short-term fractal organization of heart rate38. Normal α1 and decreased α2 were observed in patients with atrial fibrillation (AF) compared to those without AF39. Our findings suggest that exercise does indeed alter scaling exponents for long-term correlations, which may be important in diseased states related to sudden cardiac arrest.\n\n\nTranslational outlook, limitations and future directions\n\nCompared to rodent hearts with a significantly higher resting heart rate, equine hearts serve as more representative model system for electrophysiological studies. However, it is known that horses frequently show spontaneous AF due to their high vagal tone and enlarged atria. This needs to be evaluated in future studies, as detailed in other studies40–43. Nevertheless, equine athletes undergo a similar sequence to humans from an athletic life course perspective, through from training to peak performance and retirement from competitive activity. This allows for more insightful investigation into electrophysiological changes in competitive human athletes but not in the general human population, where standard-breed equine models will be more appropriate. Further studies should be conducted on larger equine sample sizes against human counterparts at different life course intervals namely training, peak performance and retirement to better match the electrophysiological changes between the two cardiac models.\n\n\nConclusion\n\nThe present findings report that exercise leads to a decrease in HRV but did not affect signal entropy in horses. Time-domain, frequency-domain and non-linear analyses all provided unique insights into signal variability, regularity and complexity.\n\n\nData availability\n\nThe dataset used in this analysis is available as part of Li et al. (2018)3\n\nPLoS One: S1 Dataset. Excel file of dataset of electrocardiographic intervals in all horses.\n\nhttps://doi.org/10.1371/journal.pone.0194008.s0013\n\nLicense: CC BY 4.0 Attribution",
"appendix": "Grant information\n\nGT is currently supported by a Clinical Assistant Professorship from the Croucher Foundation of Hong Kong.\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\nWe thank the authors of the original dataset for generously sharing their data.\n\n\nReferences\n\nChoy L, Yeo JM, Tse V, et al.: Cardiac disease and arrhythmogenesis: Mechanistic insights from mouse models. Int J Cardiol Heart Vasc. 2016; 12: 1–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBuhl R, Petersen EE, Lindholm M, et al.: Cardiac Arrhythmias in Standardbreds During and After Racing— Possible Association Between Heart Size, Valvular Regurgitations, and Arrhythmias. J Equine Vet Sci. 2013; 33(8): 590–6. Publisher Full Text\n\nLi M, Chadda KR, Matthews GDK, et al.: Cardiac electrophysiological adaptations in the equine athlete-Restitution analysis of electrocardiographic features. PLoS One. 2018; 13(3): e0194008. PubMed Abstract | Publisher Full Text | Free Full Text\n\nErnst G: Hidden Signals-The History and Methods of Heart Rate Variability. Front Public Health. 2017; 5: 265. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPeng CK, Havlin S, Stanley HE, et al.: Quantification of scaling exponents and crossover phenomena in nonstationary heartbeat time series. Chaos. 1995; 5(1): 82–7. PubMed Abstract | Publisher Full Text\n\nTse G, Wong ST, Tse V, et al.: Variability in local action potential durations, dispersion of repolarization and wavelength restitution in aged wild-type and Scn5a+/- mouse hearts modeling human Brugada syndrome. J Geriatr Cardiol. 2016; 13(11): 930–1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZareba W, Moss AJ, le Cessie S, et al.: T wave alternans in idiopathic long QT syndrome. J Am Coll Cardiol. 1994; 23(7): 1541–6. PubMed Abstract | Publisher Full Text\n\nTse G, Wong ST, Tse V, et al.: Cardiac dynamics: Alternans and arrhythmogenesis. J Arrhythm. 2016; 32(5): 411–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSzentandrassy N, Kistamas K, Hegyi B, et al.: Contribution of ion currents to beat-to-beat variability of action potential duration in canine ventricular myocytes. Pflugers Arch. 2015; 467(7): 1431–43. PubMed Abstract | Publisher Full Text\n\nPueyo E, Corrias A, Virág L, et al.: A multiscale investigation of repolarization variability and its role in cardiac arrhythmogenesis. Biophys J. 2011; 101(12): 2892–902. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTse G, Du Y, Hao G, et al.: Quantification of Beat-To-Beat Variability of Action Potential Durations in Langendorff-Perfused Mouse Hearts. Front Physiol. 2018; 9: 1578. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYan BP, Chan CK, Li CK, et al.: Resting and Postexercise Heart Rate Detection From Fingertip and Facial Photoplethysmography Using a Smartphone Camera: A Validation Study. JMIR Mhealth Uhealth. 2017; 5(3): e33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYan BP, Chan CK, Li CK, et al.: Figure Correction: Resting and Postexercise Heart Rate Detection From Fingertip and Facial Photoplethysmography Using a Smartphone Camera: A Validation Study. JMIR Mhealth Uhealth. 2019; 7(1): e11616. PubMed Abstract | Publisher Full Text\n\nLi KHC, White FA, Tipoe T, et al.: Current State of Mobile Phone Apps for Monitoring Heart Rate, Heart Rate Variability, and Atrial Fibrillation: Narrative Review. JMIR Mhealth Uhealth. 2019. Publisher Full Text\n\nBogdan P, Jain S, Marculescu R: Pacemaker control of heart rate variability: A cyber physical system perspective. ACM Trans Embed Comput Syst. 2013; 12(1s): 1–22. Publisher Full Text\n\nBogdan P, Jain S, Goyal K, et al.: Implantable Pacemakers Control and Optimization via Fractional Calculus Approaches: A Cyber-Physical Systems Perspective. 2012 IEEE/ACM Third International Conference on Cyber-Physical Systems. 2012; 23–32. Publisher Full Text\n\nOhmura H, Jones JH: Changes in heart rate and heart rate variability as a function of age in Thoroughbred horses. J Equine Sci. 2017; 28(3): 99–103. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKowalik S, Janczarek I, Kędzierski W, et al.: The effect of relaxing massage on heart rate and heart rate variability in purebred Arabian racehorses. Anim Sci J. 2017; 88(4): 669–77. PubMed Abstract | Publisher Full Text\n\nNagel C, Aurich J, Palm F, et al.: Heart rate and heart rate variability in pregnant warmblood and Shetland mares as well as their fetuses. Anim Reprod Sci. 2011; 127(3–4): 183–7. PubMed Abstract | Publisher Full Text\n\nLenoir A, Trachsel DS, Younes M, et al.: Agreement between Electrocardiogram and Heart Rate Meter Is Low for the Measurement of Heart Rate Variability during Exercise in Young Endurance Horses. Front Vet Sci. 2017; 4: 170. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLoftus L, Marks K, Jones-McVey R, et al.: Monty Roberts' Public Demonstrations: Preliminary Report on the Heart Rate and Heart Rate Variability of Horses Undergoing Training during Live Audience Events. Animals (Basel). 2016; 6(9): pii: E55. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYounes M, Robert C, Barrey E, et al.: Effects of Age, Exercise Duration, and Test Conditions on Heart Rate Variability in Young Endurance Horses. Front Physiol. 2016; 7: 155. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFlethøj M, Kanters JK, Haugaard MM, et al.: Changes in heart rate, arrhythmia frequency, and cardiac biomarker values in horses during recovery after a long-distance endurance ride. J Am Vet Med Assoc. 2016; 248(9): 1034–42. PubMed Abstract | Publisher Full Text\n\nMatsuura A, Maruta H, Iwatake T, et al.: The beneficial effects of horse trekking on autonomic nervous activity in experienced rider with no disability. Anim Sci J. 2017; 88(1): 173–9. PubMed Abstract | Publisher Full Text\n\nvon Lewinski M, Biau S, Erber R, et al.: Cortisol release, heart rate and heart rate variability in the horse and its rider: different responses to training and performance. Vet J. 2013; 197(2): 229–32. PubMed Abstract | Publisher Full Text\n\nMatsuura A, Tanaka M, Irimajiri M, et al.: Heart rate variability after horse trekking in leading and following horses. Anim Sci J. 2010; 81(5): 618–21. PubMed Abstract | Publisher Full Text\n\nSchmidt A, Aurich J, Möstl E, et al.: Changes in cortisol release and heart rate and heart rate variability during the initial training of 3-year-old sport horses. Horm Behav. 2010; 58(4): 628–36. PubMed Abstract | Publisher Full Text\n\nKinnunen S, Laukkanen R, Haldi J, et al.: Heart rate variability in trotters during different training periods. Equine Vet J Suppl. 2006; 38(S36): 214–7. PubMed Abstract | Publisher Full Text\n\nCottin F, Barrey E, Lopes P, et al.: Effect of repeated exercise and recovery on heart rate variability in elite trotting horses during high intensity interval training. Equine Vet J Suppl. 2006; 38(S36): 204–9. PubMed Abstract | Publisher Full Text\n\nVoss B, Mohr E, Krzywanek H: Effects of aqua-treadmill exercise on selected blood parameters and on heart-rate variability of horses. J Vet Med A Physiol Pathol Clin Med. 2002; 49(3): 137–43. PubMed Abstract | Publisher Full Text\n\nPhysick-Sheard PW, Marlin DJ, Thornhill R, et al.: Frequency domain analysis of heart rate variability in horses at rest and during exercise. Equine Vet J. 2000; 32(3): 253–62. PubMed Abstract | Publisher Full Text\n\nKuwahara M, Hiraga A, Kai M, et al.: Influence of training on autonomic nervous function in horses: evaluation by power spectral analysis of heart rate variability. Equine Vet J Suppl. 1999; 31(S36): 178–80. PubMed Abstract | Publisher Full Text\n\nThayer JF, Hahn AW, Pearson MA, et al.: Heart rate variability during exercise in the horse. Biomed Sci Instrum. 1997; 34: 246–51. PubMed Abstract\n\nCottin F, Médigue C, Lopes P, et al.: Effect of exercise intensity and repetition on heart rate variability during training in elite trotting horse. Int J Sports Med. 2005; 26(10): 859–67. PubMed Abstract | Publisher Full Text\n\nPincus SM, Goldberger AL: Physiological time-series analysis: what does regularity quantify? Am J Physiol. 1994; 266(4 Pt 2): H1643–56. PubMed Abstract | Publisher Full Text\n\nPincus SM: Approximate entropy as a measure of system complexity. Proc Natl Acad Sci U S A. 1991; 88(6): 2297–301. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSanbe A, James J, Tuzcu V, et al.: Transgenic rabbit model for human troponin I-based hypertrophic cardiomyopathy. Circulation. 2005; 111(18): 2330–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTulppo MP, Kiviniemi AM, Hautala AJ, et al.: Physiological background of the loss of fractal heart rate dynamics. Circulation. 2005; 112(3): 314–9. PubMed Abstract | Publisher Full Text\n\nKališnik JM, Hrovat E, Hrastovec A, et al.: Severe Cardiac Autonomic Derangement and Altered Ventricular Repolarization Pave the Way to Postoperative Atrial Fibrillation. Innovations (Phila). 2015; 10(6): 398–405. PubMed Abstract | Publisher Full Text\n\nBroux B, De Clercq D, Vera L, et al.: Can heart rate variability parameters derived by a heart rate monitor differentiate between atrial fibrillation and sinus rhythm? BMC Vet Res. 2018; 14(1): 320. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBroux B, De Clercq D, Decloedt A, et al.: Heart rate variability parameters in horses distinguish atrial fibrillation from sinus rhythm before and after successful electrical cardioversion. Equine Vet J. 2017; 49(6): 723–8. PubMed Abstract | Publisher Full Text\n\nGelzer AR, Moïse NS, Vaidya D, et al.: Temporal organization of atrial activity and irregular ventricular rhythm during spontaneous atrial fibrillation: an in vivo study in the horse. J Cardiovasc Electrophysiol. 2000; 11(7): 773–84. PubMed Abstract | Publisher Full Text\n\nKuwahara M, Hiraga A, Nishimura T, et al.: Power spectral analysis of heart rate variability in a horse with atrial fibrillation. J Vet Med Sci. 1998; 60(1): 111–4. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "43955",
"date": "20 Feb 2019",
"name": "Arun V. Holden",
"expertise": [
"Reviewer Expertise Computational biology",
"time series analysis",
"nonlinear dynamics"
],
"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 clearly presented quantitative description of heart rate variability in thoroughbred racehorses, from data sets that are publically available from ref [3] as an xlsx dataset of RR,QT intervals and analysed by standard methods. The results and conclusion are quantitative. The paper does not add anything to the methodology or interpretation of HTV analysis, or to physiological understanding of exercise effects on heart activity, but provides a standard analysis of data from an unusual (for mammalian physiology) animal, and is of potential interest to veterinary medicine. “Equine athletes” is not a term I would search for if I was looking to racehorse ECG analyses. In sports medicine an athletes could be marathon runner, a sprinter, or a weightlifter, all with different cardiovascular profiles. I know it’s used in the previous PlosOne paper, so suggest you replace horse by race horse in the abstract.\n\nFig 1 A, b , 2A,B show RR intervals, rates from the same horse at rest and during exercise. Could “at rest” and “during exercise “ be more specific in the legend/text. A striking feature are the transient accelerations in heart rate at rest, and the slow decline in rate during exercise. Could you comment on these, in the results (are they typical, can they be quantified over the full set of data i.e accelerations in heart rate, in which RR interval dropped by more then x % in y seconds, occurred in % at rest, and y1-y2% decrease in heart rate occurred during the period of exercise; and comment on the possible physiological mechanism. This would add to the paper, and is relevant to your title as it concerns the assumed stationarity of the RR interval sequences you are analysing. Figs 1,2 C,F “frequency”: since you get into the frequency domain later in Fig5 I’d suggest replace frequency by probability (number in bin /no. of intervals) and add a tick marks for linear scale, and unit, so they are now estimates of probability density and have a common scale Time domain statistics: Figs 3 and 4: all OK, but repetitive as the figs repeat the text. I’d prefer one table + some descriptive text.\n\nMethods: Frequency domain analysis. The data in [3] is a sequence of intervals: '…using the FFT with a sampling rate of 8 Hz...' implies you are sampling a continuous signal ie. the V(t), not the point process of R events by their interval sequence (see Kybernetik. 1971 May;8(5):165-71. Alias-free sampling of neuronal spike trains.French AS, Holden AV1.) I think you are constructing a staircase, by plotting the interval against the sum of the intervals ie. time, and sampling this function at 8 Hz. Could you be specific about what you are doing. The power in the repolarization spectrum: I think you mean the total power is the integral of the spectral density from 0.04 and 0.4Hz, repolarization spectrum does not seem to be used in the text or figures. In the legend of fig 5 A, B identify the colours with the frequency bands mentioned in the text. Figs 5A,B PSD scales differ by an order of magnitude.: are the spectra from the same length (number of RR intervals/total time?) of data The spectra in Figs 5A, B look plausible but you need to give more information about how they were constructed [4] is a simple overview. If it were a package (say OriginPro) then give the parameters used (sampling rate, window, type,…,or specify the method in detail What was the total record length? Was it zero-meaned? Was it segmented into 50% overlapping blocks How long each block, how many degrees of freedom, what window function; Hanning, hamming,…. See any text on FFT. Since I’m not too sure of the methods you used to construct the spectra I’m not too sure about the exact meaning of Figs 5 and 6\n\nFig 7: Recurrence plots not Poincare recurrence ( a recurrence plot would be to have a V(t) signal in say x,y.z space and plot its point intersections with a plane, here you’re plotting RR_n+! against RR_n ie a straight recurrence plot, and then quantifing it.\n\nYou explain the sample entropies and DFA but do not give the method (as a computer program say Peng’s code) or package used. If you wrote the code yourself have it as a supplement so the reader could validate it/use it on the data in [3].\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? Yes\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
},
{
"id": "46731",
"date": "23 Apr 2019",
"name": "Haibo Ni",
"expertise": [
"Reviewer Expertise Computational Cardiology"
],
"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, Li and colleagues performed quantitative analyses of heart rate variability in horses based on previously published data in reference 3, and reached a conclusion that is supported by quantitative results.\n\nHere is a list of comments that I believe this MS could benefit from:\nIn Methods – Frequency-domain analysis, how were the cut-off frequencies for these fast, low, and very low frequency determined? Also,‘repolarization spectrum’ appears confusing. It would be helpful if explained in more detail. Please also add more details regarding how FFT was performed, i.e., what software/code and relevant parameters, in order to reproduce these analyses. It is not clear exactly what ‘at rest’ and ‘exercise’ mean in the context of this study, e.g., is ‘at rest’ suggesting HR data taken minutes/hours before exercise? This is more relevant for ‘exercise’ as it is not clear what is the form and duration of exercise, and perhaps these data are taken after exercise? If so, it would be very helpful to clearly describe at what time point these data were taken, since Figures 1 and 2 show gradually slowed HR in the ‘exercise’ column. Other HR-contributing factors like the circadian rhythm may also ideally need to be ruled out if ‘rest’ and ‘exercise’ data are taken during different part of the day. It is also worthwhile to describe the sexes of the horses. Is there any sex-based difference? Figures 1 and 2 are essentially showing the same data and therefore may benefit from merging into a single figure. The authors described the duration of single HR recording (10 to 18 min), were the analyses performed on the whole length of the recordings or segments of them? Please elaborate and perhaps provide a representative plot of whole length recording of HR/RR data? Also, in Figure 2 C-D, there’s no tick/tick labels for the y axis. Figure 5A-B, please add notations for what the color areas stand for. For entropy and detrended fluctuation analyses, please add detailed descriptions of implementation methods. Figure 8A-B appear that the plots do not cover complete dataset. Please modify the axis ranges to improve.\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": []
}
] | 1
|
https://f1000research.com/articles/8-147
|
https://f1000research.com/articles/8-146/v1
|
04 Feb 19
|
{
"type": "Systematic Review",
"title": "Preparedness components of health systems in the Eastern Mediterranean Region for effective responses to dust and sand storms: a systematic review",
"authors": [
"Kiyoumars Allahbakhshi",
"Davoud Khorasani-Zavareh",
"Reza Khani Jazani",
"Zohreh Ghomian",
"Kiyoumars Allahbakhshi",
"Davoud Khorasani-Zavareh",
"Reza Khani Jazani"
],
"abstract": "Background: Dust and Sand Storm (DSS), according to estimates by global reports, will increase dramatically in the Eastern Mediterranean Region (EMR). Numerous health problems caused by DSS will be severely affected regions and vulnerable groups. This study aimed to identify the components of the preparedness of health systems for the DSS phenomenon in EMR. Methods: In this systematic review, the peer-reviewed papers in four electronic databases, including Medline through PubMed, Scopus, ISI Web of Science and the Cochrane library, as well as available grey literature, were searched and selected. The research process was carried out by including papers whose results were related to the potential health effects caused by desert dusts in EMR. Was used the combination of three groups of keywords: the exposure factor, health effects as outcomes, and the countries located in EMR. The focus was on the PRISMA checklist, with no time limitations until December 2017. Finally, through 520 related citations, 30 articles were included. Descriptive and thematic content analyses were evaluated. Results: The preparedness components were divided into three and ten main categories and subcategories, respectively. The three categories covered the areas of DSS hazard identification, planning and policy-making, and risk assessment. Conclusions: Recognition of the health system preparedness factors for DSS in EMR will help policy-makers and managers perform appropriate measures when dealing with this hazard. More studies should be conducted to understand these factors in other parts of the world. Registration: PROSPERO registration number CRD42018093325.",
"keywords": [
"dust and sand storm",
"health",
"preparedness",
"reediness",
"Eastern Mediterranean Region"
],
"content": "Introduction\n\nDust and Sand Storms (DSS), as a natural disaster from the type of meteorological hazards1,2, affects the atmospheric system, air quality, and human health3–6. Over the recent years, the review studies have shown that there is a relation between the occurrence of the DSS and the incidence of human health7,8.\n\nBased on the division by the World Health Organization (WHO), there are 22 countries in the Eastern Mediterranean Region (EMR)9. According to estimates from the Intergovernmental Panel on Climate Change (IPCC) report, EMR is one of the areas that will be strongly affected by this phenomenon in the future and will have potentially harmful health effects, especially on vulnerable groups10,11. Therefore, the best way to minimize the damage caused by the disasters is the preparedness for them12. The constant increasing risk of hazards indicates the need to integrate disaster risk reduction in preparedness, reinforce disaster preparedness, and provide assurance for timely utilization of the capacities13.\n\nThe epidemiological studies conducted on DSS mainly focus on the effects of the phenomenon on mortality and morbidity in humans14–20 or on the nature of major sources of dust phenomenon, the frequency, the duration and changes to the occurrence of the DSS, and the content of particulate matter (PM) suspended in the air8. Considering the DSS challenges on human health, drastic measures should be taken to ensure the health system is prepared for this phenomenon. In this regard, important elements have been proposed in the health system preparedness for disasters including the identification of hazards, the provision of Emergency Operation Plan (EOP), training and education, equipment, Early Warning System (EWS), information exchange, drill and exercise, monitoring, and evaluation13,21–28. On the other hand, there has not been a comprehensive study on the factors affecting the preparedness of the health system for dust phenomenon in EMR. Accordingly, the present study, through systematic review (SR), aims to investigate the factors affect the health system preparedness for the DSS in EMR and identify research gaps on the preparedness of the health system for DSS. The findings of the study, with proposed recommendations, can directly help health policymakers in preparedness promotion for this phenomenon, pave the way for further studies, and add to the richness of the current knowledge.\n\n\nMethods\n\nThis study reports a SR based on the recommendations of the Cochrane and PRISMA guideline29. The PROSPERO registration number is CRD42018093325. A completed PRISMA checklist is available on figshare30.\n\nAll English-language articles related to effective factors on the preparedness of the health system for dust phenomenon in EMR were searched until December 2017. All methods of study, books, and theses that associated with the subject of this research were included. Papers in the form of the letter to the editor and studies that were merely related to the dust subject, and in which no mention of health outcome, were excluded, as were non-English-language articles, as were those without full-text access. Additionally, scientific documents related to non-desert origin (such as volcanic or anthropogenic sources) were not included.\n\nThe international electronic databases investigated by authors (KA, ZGH, DKZ), including English sources from Medline through PubMed, Scopus, ISI web of science, Google Scholar, Cochrane library. The Eastern Mediterranean Health Journal and African Journals OnLine were also searched for published articles in the EMR. Other documents were extracted from reports published by organizations (such as the United Nations).\n\nFollowing consultations with Library & Information Science (LIS) professionals in the health field, the search for articles was done using a combination of three groups of words in the databases mentioned above: The exposure factor (Dust Storm OR Sand Storm OR desert dust), health effects as outcomes (health OR morbidity OR mortality OR disease OR hospital OR respiratory OR cardiovascular OR coccidiomycosis OR meningococcal meningitis OR conjunctivitis OR dermatological OR transport accidents), and member states of the EMR (Middle East OR Afghanistan OR Bahrain OR Djibouti OR Egypt OR Iran OR Iraq OR Jordan OR Kuwait OR Lebanon OR Libyan Arab Jamahiriya OR Morocco OR Oman OR Pakistan OR Occupied Palestinian Territory OR Qatar OR Saudi Arabia OR Somalia OR Sudan OR Syrian Arab Republic OR Tunisia OR United Arab Emirates OR Yemen). These three groups of words were combined with \"AND\" together. The keywords were obtained from previous articles relating to the challenges of the health domain in connection with the DSS. In order to increase the probability of identification of all relevant literature, these keywords were selected based on an agreement of three researchers (KA, ZGH, and DKZ). The search words are used in titles, abstracts, and keywords of used articles. Figure 1 shows the search strategy.\n\nIn order to manage the citations, EndNote software, version X15, was used. All duplicate records became clear and removed. Using the title, abstract and, keywords screening, and given the inclusion criteria, the evaluation of documents was performed by two researchers (KA and ZGH). In the next step, the full text of the remaining articles was analyzed independently by the two researchers considering the inclusion and exclusion criteria and standard quality assessment. The papers were later examined in order to observe the points of the criteria for inclusion/exclusion by two other scholars (DKZ and RKJ). The quality of all articles was evaluated using the Cochrane handbook to evaluate the bias (Table 1)31. In this SR, the data extraction sheet was designed in two main parts. The first section covers the general specification of articles containing the ID, hyperlinks, title, electronic database, first author, publication year, country or region, method, population, and year/years of study. The second part was used to identify the main findings of articles, research suggestions related to this SR, and the preparedness components of the health system for the DSS. All extracted data were evaluated by members of the research team to verify accuracy and completeness.\n\n\nResults\n\nIn general, through the implementation of the research strategy, the number of 520 records was found. In the last stage, 30 unique articles were obtained based on the inclusion/exclusion criteria (Figure 1). Data on the risk evaluation of bias are provided in Table 1. Final articles were examined in two parts, including descriptive and thematic content analysis. In Table 2, more details are presented about the imported literature to the SR. Findings indicate that since 2008, researchers have published more articles about the effects of the DSS on public health. Most articles related to Iran (17 articles, 56%). None of them had a qualitative approach. In thematic content analysis, based on literature review of studies that identify factors affecting the health system preparedness for various hazards21–28 and the multistage analysis by the research group, the preparedness components of health system for DSS in EMR were extracted and classified (Table 3).\n\nTM, Total Mortality; CM, Cardiovascular Mortality; RM, Respiratory Mortality; HARD, Hospital Admission for Respiratory Diseases; HACD, Hospital Admission for Cardiovascular Diseases; PHC, Primary Health Centers; DuSNIFF, Dust Storm Network-based Integrated System of Forecast and Forewarning.\n\nNA, not applicable.\n\nHealth problems caused by DSS. Studies conducted in EMR showed that the prevalence of respiratory diseases (RD)17,34–43 and respiratory mortality (RM) is directly related to DSS17,40,41. Some of these studies focused on the incidence of asthma and pulmonary dysfunction in schoolchildren44. On the other hand, dust storms result in hospitalization due to cardiac diseases37,40,41,45 and cardiac mortalities (CM)40. Some of the various fungal species in EMR dust storms cause infectious diseases in human beings46 that increase or decrease allergic and asthma diseases47. Exposure to carcinogenic metals in PM10 increases the risk of cancer in citizens48. Other problems reported by the researchers were death and trauma due to the occurrence of storms (at a speed of 110 km/h) with dust particles49. Researchers found that as the concentration of DSS increases, the death caused by road traffic accidents (RTAs) decreases50. Studies conducted in Qatar showed that inhaling cyanotoxins found in DSS could have devastating effects on human health51,52. In contrast to these results, a number of researchers found that in general DSS had no effect on asthma, RD, RM18,45,53, CM18, and RTAs54.\n\nComposition of DSS. The most abundant of bacteria observed in the dust particles were Bacillus spp.42,55,56, Staphylococcus spp., Streptomyces spp., Micrococcus spp.56 and Escherichia coli42. Also, most types of fungi were (Mycosporium spp.55 Penicillium, Aspergillus flavus, Cladosporium, Alternaria, Rhizopus, and Cladosporium)46,47, C. albicans42. The pollen grains caused allergies, identified from dust storms, in descending order, included Chenopodiaceous, Graminea, Pine, Artemisia, Palmae, Olea, and Typha42. No viruses were found in the dust particle samples42. On the other hand, the studies showed that the concentration of metals in dust increases in dusty days35. The main elements contained in the PM10 include (Na, Ca, Mg, Al, Fe)48,57. Among the carcinogenic metals in PM10 are Ni, Cr, As, and Cd57. According to studies conducted in Jordan and Saudi Arabia, dust samples contained considerable amounts of radioactivity58,59. Among other compounds reported in the desert dust are cyanobacteria toxins51,52.\n\nEconomic losses on health. The economic losses caused by the respiratory problems in Zabol, Iran, are estimated at about US$66 million32. The total damage estimated from DSS to the health system in Iran and Iraq amounted to US$306 million33.\n\nEducation and training. The researchers emphasized that there should be health care recommendations for all affected individuals by DSS to reduce the vulnerability of population at risk, especially susceptible groups such as older persons, children, and cardiovascular and respiratory patients17,36,60. Also, providing community-based training is an important role in the proper functioning of the people after receiving a warning message related to DSS49.\n\nResearch. Scientific findings focused on this issue that further epidemiological studies should be conducted to identify chemical compounds and microorganisms in PM10, the baseline incidence values for each country, and the potential effects of DSS on health, especially long-term effects18,35,37,39,43,52–55,57,61. Moreover, there was no qualitative research among the studies reviewed.\n\nCompilation of new and local indexes of air quality. Based on the assessment of PM-related health risks, researchers found that it is necessary to design new standards for local ambient air quality in the EMR61,62. Only limited countries such as Saudi Arabia, Bahrain and Jordan have an air pollution index relevant to their country61,63.\n\nComprehensive database development. To sum up the studies, provision of a comprehensive database of air pollutants and the effects of natural hazards, such as DSS, on health facilities is crucial for the benefit of policymakers and people57,64.\n\nPrediction and warning. Based on the findings of the studies conducted in EMR, the deployment of EWS such as the network-based integrated system of forecast and forewarning (DuSNIFF) can be a good base and framework for a timely warning to the population at risk49,65. Conversely, other researchers reported that there is no need for a warning to the emergency department of hospitals in the event of DSS occurrence53.\n\nUse of hospital safety assessment tools. According to the assessment of Farsi Hospital Safety Index (FHSI), as a preparedness tool66,67, during different years in Iran, DSS was assessed as one of the highest probability of occurrence and health effects68.\n\nUse models for health risk assessment of DSS. Based on the studies, health policymakers can find a better understanding of the health effects associated with PM10 peak times by using the findings of analytic models such as AirQ69 and generalized additive model (GAM)18,39.\n\n\nDiscussion\n\nPreparedness for DSS in large-scale management and at the community level is one of the essential measures in the EMR. Provision of the necessary information such as the burden of diseases caused by DSS can be used in policy-making to focus on pre-disaster planning such as preparedness. The Hyogo Framework for Action (HFA) as an international strategy put emphasis on preparedness in order to produce effective response measures at all levels, to prioritize disaster risk reduction, and to reduce the background risk factors70. Preparedness needs to be preserved and dynamic and ongoing efforts21. Conducting research and producing scientific evidence relevant to disease burdens resulting from DSS can help to improve the health system preparedness for DSS71. Given that health disorders caused by climate change affect the pattern and changing the burden of disease in the community, therefore, having basic guidelines on preparedness will make optimal use of resources in health service delivery72,73. The health managers in EMR must develop their readiness based on the recognition of the burden of acute and long-term diseases and different dimensions of phenomenon. DSS preparedness requires a clear understanding and assessment of the country's situation. Health centers also need to be prepared regarding the personnel, equipment, medicine, and infrastructure.\n\nThe trained people that have greater understanding are more aware of the risks of hazards and in the event of a disaster, they act more appropriately74,75. One of the top priorities of the Sendai Framework for disaster risk 2015–2030 is the understanding of disasters in all their dimensions to allow appropriate measures to be taken for disaster preparedness. This framework is the result of consultations and intergovernmental negotiations that were encouraged by the United Nations Office for Disaster Risk Reduction13. Training, as one of the effective factors in the promotion of a disaster-preparedness culture, is essential in two levels: community-based education and training of health providers. Planning and doing regular drills in various scales is essential for the promotion of general and specialized education levels of the organizations and the people affected by this phenomenon. To enhance the understanding of disaster risk among managers and the community, awareness-raising programs should focus on capacity development through sharing previous experiences and lessons about preparedness for DSS. Also, the development of indigenous knowledge should be considered. For instance, in desert area, and not available to a mask, using of keffiyeh76,77 is recommended for protection from DSS.\n\nWith the purpose of planning for preparedness, regional policymakers must consider local considerations when using air quality indexes. Tsiouri et al.63 and Murena78 noted that different geographical regions have specific climatic conditions; therefore, this issue has an impact on atmospheric pollutants that affect human health as well as population responses to air pollution. As a result, the localization and adaptation of air pollution and its indicators can take place throughout the world. With the purpose of planning for preparedness, regional policymakers have to take into account that cannot be ignored local considerations when using air quality indexes63,78. To reduce the concerns, in light of valid regional evidence, the use of these standards should be reviewed.\n\nSome studies in EMR18,45,50,53,54 and other regions of the world have shown that there is no association between the occurrence of dust and increasing health problems79–82. However, it is important to conclude that with increasing frequency, intensity and geographic expansion of DSS, it is necessary to ensure a timely and valid warning to vulnerable populations and groups16,83. Provision of advanced and accurate warning systems requires continuous efforts to improve air quality modeling and prediction73. Moreover, the results of various studies in the world show that early health warnings to vulnerable people about air pollution can reduce emergency visits to health centers through the reduction of outdoor activity during dusty days16,83.\n\nThe trans-boundary nature of DSS, unlike many natural hazards, is not limited to a specific geographic area; therefore, regional cooperation is needed to prepare for this phenomenon. In this regard, Kuwaiti scholars have suggested that it is essential to establish a regional committee54. The health system of the countries involved in DSS give priority to the development of regional health memorandums of understanding (MoU). In the framework of these MoUs, countries can strengthen regional and global collaboration as with UN and WMO to transfer modern technologies for prediction of DSS occurrences, exchange of medical knowledge, allocation of financial and technical assistance, combat against desertification, share of information and successful practical experiences, to form a regional credit union, and to build training workshops.\n\nIt can be noted, among the strengths of this study, the literature analysis performed with carefully assessing and have been done several times by the research team. Their research area was about Health in Disasters and Emergencies. The focus of this SR was on the EMR, as one of the most challenging areas, which has large DSS sources and creates regional health problems. A comprehensive study with this goal, so far, has not been conducted in this region. However, in this SR, only English articles related to EMR were included, the number of articles related to dust and health field was limited, and the full text of some studies was not accessible.\n\n\nConclusion\n\nThe burden of diseases caused by dust in EMR shows the need to undertake measures for government preparedness to protect the health of affected. Given that DSS is a large-scale hazard, to gain preparedness, countries should move towards regional and international cooperation. The health system needs to develop a comprehensive plan of readiness to improve the effectiveness of the response measures. Also, regular exercises in all scales are a very important component. To increase public health recommended the development of dust-health EWS. Promotion of preparedness culture and the increase of public awareness about the effects of DSS through public media are suggested. In preparedness programs, the participation of the community is recommended. Health workers should receive regular training on cardiovascular and respiratory problems. Further quantitative and qualitative researches to identify the nature of DSS and adaptive factors can help bridge the gap between scientific findings and preparedness measures. Although we addressed preparedness in this study, there should be a comprehensive plan to manage the hazard and to consider all the loops of the disaster risk management cycle. In general, this study can help policy-makers of the health system in disaster risk management to identify factors that are effective in preparedness for DSS and to take the necessary preparedness measures.\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\nReporting guidelines\n\nA completed PRISMA checklist is available on figshare, DOI: https://doi.org/10.6084/m9.figshare.758183330.\n\nLicense: CC0 1.0 Universal.",
"appendix": "Grant information\n\nThe authors declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThis study is the result of the dissertation of the PhD in the School of Public Health and Safety, Shahid Beheshti University of Medical Sciences, Tehran, Iran. Current study was approved by the Ethics Committee in Biomedical Research of Shahid Beheshti University of Medical Sciences, Tehran, Iran, with the ethical code of IR.SBMU.RETECH.1396.1149.\n\n\nReferences\n\nEM-DAT: The International Disaster Database, Centre for Research on the Epidemiology of Disasters- RECRED. Reference Source\n\nThe Disasters classification used in EM-DAT is based on and adapted from the he IRDR Peril Classification and hazard Glossary. DATA Project Report #2, March 2014. Reference Source\n\nShao Y, Wyrwoll KH, Chappell A, et al.: Dust cycle: An emerging core theme in Earth system science. Aeolian Res. 2011; 2: 181–204. Publisher Full Text\n\nMahowald NM, Kloster S, Engelstaedter S, et al.: Observed 20th century desert dust variability: Impact on climate and biogeochemistry. Atmos Chem Phys. 2010; 10(22): 10875–93. Publisher Full Text\n\nDe Longueville F, Hountondji YC, Henry S, et al.: What do we know about effects of desert dust on air quality and human health in West Africa compared to other regions? Sci Total Environ. 2010; 409(1): 1–8. PubMed Abstract | Publisher Full Text\n\nGiannadaki D, Pozzer A, Lelieveld J: Modeled global effects of airborne desert dust on air quality and premature mortality. Atmos Chem Phys. 2014; 14(2): 957–68. Publisher Full Text\n\nZhang X, Zhao L, Tong DQ, et al.: A systematic review of global desert dust and associated human health effects. Atmosphere. 2016; 7(12): 158. Publisher Full Text\n\nGoudie AS: Desert dust and human health disorders. Environ int. 2014; 63: 101–13. PubMed Abstract | Publisher Full Text\n\nThe WHO Regional Office for the Eastern Mediterranean. Reference Source\n\nPachauri RK, Reisinger A: IPCC fourth assessment report. IPCC, Geneva. 2007; 2007. . Reference Source\n\nMessage from the Regional Director. Reference Source\n\nRana AK, Pandey DJ: Higher Education in Disaster Management: Educating Built Environment Professions to Reduce Damage. the international emergency management society journal.71.\n\nUNISDR (United Nations International Strategy for Disaster Reduction): 2015: Sendai framework for disaster risk reduction 2015–2030. 2015; Accessed 6 Jun 2015. Reference Source\n\nMartinelli N, Olivieri O, Girelli D: Air particulate matter and cardiovascular disease: a narrative review. Eur J Intern Med. 2013; 24(4): 295–302. PubMed Abstract | Publisher Full Text\n\nTao Y, An X, Sun Z, et al.: Association between dust weather and number of admissions for patients with respiratory diseases in spring in Lanzhou. Sci Total Environ. 2012; 423: 8–11. PubMed Abstract | Publisher Full Text\n\nTam WW, Wong TW, Wong AH, et al.: Effect of dust storm events on daily emergency admissions for respiratory diseases. Respirology. 2012; 17(1): 143–8. PubMed Abstract | Publisher Full Text\n\nKhaniabadi YO, Daryanoosh SM, Amrane A, et al.: Impact of Middle Eastern Dust storms on human health. Atmos Pollut Res. 2017; 8(4): 606–13. Publisher Full Text\n\nAl-Taiar A, Thalib L: Short-term effect of dust storms on the risk of mortality due to respiratory, cardiovascular and all-causes in Kuwait. Int J Biometeorol. 2014; 58(1): 69–77. PubMed Abstract | Publisher Full Text\n\nMiri A, Ahmadi H, Ghanbari A, et al.: Dust storms impacts on air pollution and public health under hot and dry climate. Int J Energy Environ. 2007; 2(1): 101–5. Reference Source\n\nNeophytou AM, Yiallouros P, Coull BA, et al.: Particulate matter concentrations during desert dust outbreaks and daily mortality in Nicosia, Cyprus. J Expo Sci Environ Epidemiol. 2013; 23(3): 275. PubMed Abstract | Publisher Full Text\n\nCoppola DP: Introduction to international disaster management. Elsevier. 2006. Reference Source\n\nAsghar S, Alahakoon D, Churilov L: A comprehensive conceptual model for disaster management. J Human Assist. 2006; 1360(0222): 1–15. Reference Source\n\nKhan H, Vasilescu LG, Khan A: Disaster management cycle-a theoretical approach. J Manage Mark. 2008; 6(1): 43–50. Reference Source\n\nPelfrey WV: The cycle of preparedness: Establishing a framework to prepare for terrorist threats. J Homel Secur Emerg. 2005; 2(1). Publisher Full Text\n\nDisaster preparedness cycle. Reference Source\n\nKent R: Disaster Preparedness - 2nd Edition (DHA/UNDRO - DMTP - UNDP). 66. Reference Source\n\nWorld Health Organization: Risk reduction and emergency preparedness: WHO six-year strategy for the health sector and community capacity development. 2007. Reference Source\n\nWorld Health Organization: Health sector emergency preparedness guide: Making a difference to vulnerability. Geneva: WHO. 1998; 114. Reference Source\n\nMoher D, Liberati A, Tetzlaff J, et al.: Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009; 151(4): 264–9. PubMed Abstract | Publisher Full Text\n\nAllahbakhshi K, Khorasani-Zavareh D, Jazani RK, et al.: PRISMA 2009 checklist.word.doc. figshare. Figure. 2019. http://www.doi.org/10.6084/m9.figshare.7581833.v3\n\nHiggins JPT, Green S: Cochrane handbook for systematic reviews of interventions version 5.1. 0. The Cochrane Collaboration. 2011; 2015. Reference Source\n\nMiri A, Ahmadi H, Panjehkeh N, editors, et al.: Dust Storms Impact on Air pollution and public Health: A Case Study in Iran. WSEAS International Conferences; 2007. Reference Source\n\nMeibodi AE, Abdoli G, Taklif A, et al.: Economic Modeling of the Regional Polices to Combat Dust Phenomenon by Using Game Theory. Proc Econ Financ. 2015; 24: 409–18. Publisher Full Text\n\nGeravandi S, Sicard P, Khaniabadi YO, et al.: A comparative study of hospital admissions for respiratory diseases during normal and dusty days in Iran. Environ Sci Pollut Res Int. 2017; 24(22): 18152–9. PubMed Abstract | Publisher Full Text\n\nNaimabadi A, Ghadiri A, Idani E, et al.: Chemical composition of PM10 and its in vitro toxicological impacts on lung cells during the Middle Eastern Dust (MED) storms in Ahvaz, Iran. Environ Pollut. 2016; 211: 316–24. PubMed Abstract | Publisher Full Text\n\nNeisi A, Vosoughi M, Idani E, et al.: Comparison of normal and dusty day impacts on fractional exhaled nitric oxide and lung function in healthy children in Ahvaz, Iran. Environ Sci Pollut Res Int. 2017; 24(13): 12360–71. PubMed Abstract | Publisher Full Text\n\nKhaniabadi YO, Fanelli R, De Marco A, et al.: Hospital admissions in Iran for cardiovascular and respiratory diseases attributed to the Middle Eastern Dust storms. Environ Sci Pollut R. 2017; 24(20): 1–9. PubMed Abstract | Publisher Full Text\n\nAmarloei A, Jonidi Jafari A, Mahabadi HA, et al.: Investigation on the lung function of general population in Ilam, west of Iran, as a city exposed to dust storm. Glob J Health Sci. 2015; 7(3): 298–308. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThalib L, Al-Taiar A: Dust storms and the risk of asthma admissions to hospitals in Kuwait. Sci Total Environ. 2012; 433: 347–51. PubMed Abstract | Publisher Full Text\n\nMarzouni MB, Alizadeh T, Banafsheh MR, et al.: A comparison of health impacts assessment for PM10 during two successive years in the ambient air of Kermanshah, Iran. Atmos Pollut Res. 2016; 7(5): 768–74. Publisher Full Text\n\nGoudarzi G, Daryanoosh S, Godini H, et al.: Health risk assessment of exposure to the Middle-Eastern Dust storms in the Iranian megacity of Kermanshah. Public Health. 2017; 148: 109–16. PubMed Abstract | Publisher Full Text\n\nAl-Dabbas MA, Abbas MA, Al-Khafaji RM: Dust storms loads analyses—Iraq. Arab J Geosci. 2012; 5(1): 121–31. Publisher Full Text\n\nKhaefi M, Geravandi S, Hassani G, et al.: Association of particulate matter impact on prevalence of chronic obstructive pulmonary disease in Ahvaz, southwest Iran during 2009–2013. Aerosol Air Qual Res. 2017; 17(1): 230–7. Publisher Full Text\n\nBener A, Abdulrazzaq YM, Al-Mutawwa J, et al.: Genetic and environmental factors associated with asthma. Hum Biol. 1996; 68(3): 405–14. PubMed Abstract\n\nEbrahimi SJ, Ebrahimzadeh L, Eslami A, et al.: Effects of dust storm events on emergency admissions for cardiovascular and respiratory diseases in Sanandaj, Iran. J Environ Health Sci Eng. 2014; 12(1): 110. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSoleimani Z, Goudarzi G, Naddafi K, et al.: Determination of culturable indoor airborne fungi during normal and dust event days in Ahvaz, Iran. Aerobiologia. 2013; 29(2): 279–90. Publisher Full Text\n\nShabankarehfard E, Ostovar A, Farrokhi S, et al.: Air- and Dust-Borne Fungi in Indoor and Outdoor Home of Allergic Patients in a Dust-Storm-Affected Area. Immunol Invest. 2017; 46(6): 577–89. PubMed Abstract | Publisher Full Text\n\nAlghamdi MA: Characteristics and Risk Assessment of Heavy Metals in Airborne PM10 from a Residential Area of Northern Jeddah City, Saudi Arabia. Pol J Environ Stud. 2016; 25(3): 939–949. Publisher Full Text\n\nMoradian MJ, Rastegarfar B, Rastegar MR, et al.: Tehran dust storm early warning system: corrective measures. PLoS Curr. 2015; 7: pii: ecurrents.dis.14f3c645eb2e2003a44c6efd22c23f5e. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlmasi A, Bakhshi S, Pirsaheb M, et al.: Effects of air pollution caused by particulate matter (PM10) on tourism industry and road accidents-case study: Kermanshah, Iran (2008-2013). Acta Med Mediterr. 2016; 32: 1951–4. Reference Source\n\nRicher R, Banack SA, Metcalf JS, et al.: The persistence of cyanobacterial toxins in desert soils. J Arid Environ. 2015; 112: 134–9. Publisher Full Text\n\nMetcalf JS, Richer R, Cox PA, et al.: Cyanotoxins in desert environments may present a risk to human health. Sci Total Environ. 2012; 421–422: 118–23. PubMed Abstract | Publisher Full Text\n\nAlangari AA, Riaz M, Mahjoub MO, et al.: The effect of sand storms on acute asthma in Riyadh, Saudi Arabia. Ann Thorac Med. 2015; 10(1): 29–33. PubMed Abstract | Free Full Text\n\nMohammad MSA: Dust storm phenomena and their environmental impacts in Kuwait. University of Glasgow; 1989. Reference Source\n\nNourmoradi H, Moradnejadi K, Moghadam FM, et al.: The Effect of Dust Storm on the Microbial Quality of Ambient Air in Sanandaj: A City Located in the West of Iran. Glob J Health Sci. 2015; 7(7 Spec No): 114. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSoleimani Z, Goudarzi G, Sorooshian A, et al.: Impact of Middle Eastern dust storms on indoor and outdoor composition of bioaerosol. Atmos Environ. 2016; 138: 135–43. Publisher Full Text\n\nHosseini G, Teymouri P, Shahmoradi B, et al.: Determination of the Concentration and Composition of PM10 during the Middle Eastern Dust Storms in Sanandaj, Iran. J Res Health Sci. 2015; 15(3): 182–8. PubMed Abstract\n\nAlharbi A, El-Taher A: Measurement of Natural Radioactivity and Radiation Hazard Indices for Dust Storm Samples from Qassim Region, Saudi Arabia. Life Sci J. 2014; 11(9). Reference Source\n\nHamadneh HS, Ababneh ZQ, Hamasha KM, et al.: The radioactivity of seasonal dust storms in the Middle East: the May 2012 case study in Jordan. J Environ Radioact. 2015; 140: 65–9. PubMed Abstract | Publisher Full Text\n\nAlolayan MA, Brown KW, Evans JS, et al.: Source apportionment of fine particles in Kuwait City. Sci Total Environ. 2013; 448: 14–25. PubMed Abstract | Publisher Full Text\n\nJassim MS, Coskuner G: Assessment of spatial variations of particulate matter (PM10 and PM2.5) in Bahrain identified by air quality index (AQI). Arab J Geosci. 2017; 10(1): 19. Publisher Full Text\n\nLim CC, Thurston GD, Shamy M, et al.: Temporal variations of fine and coarse particulate matter sources in Jeddah, Saudi Arabia. J Air Waste Manag Assoc. 2018; 68(2): 123–138. (just-accepted). PubMed Abstract | Publisher Full Text | Free Full Text\n\nTsiouri V, Kakosimos KE, Kumar P: Concentrations, sources and exposure risks associated with particulate matter in the Middle East Area—a review. Air Qual Atmos Health. 2015; 8(1): 67–80. Publisher Full Text\n\nArdalan A, Mowafi H, Khoshsabeghe HY: Impacts of natural hazards on primary health care facilities of Iran: a 10-year retrospective survey. PLoS Curr. 2013; 5: pii: ecurrents.dis.ccdbd870f5d1697e4edee5eda12c5ae6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShaw R, Izumi T, Shi P, et al.: Asia science technology status for disaster risk reduction. Integrated Research on Disaster Risk (IRDR); 2016. Reference Source\n\nArdalan A, Kandi M, Talebian MT, et al.: Hospitals safety from disasters in I.R.iran: the results from assessment of 224 hospitals. PLoS Curr. 2014; 6: pii: ecurrents.dis.8297b528bd45975bc6291804747ee5db. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOrganization WH: Hospital safety index: Guide for evaluators. 2015. Reference Source\n\nArdalan A, Keleh MK, Saberinia A, et al.: 2015 estimation of hospitals safety from disasters in IR Iran: the results from the assessment of 421 hospitals. PLoS One. 2016; 11(9): e0161542. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaleki H, Sorooshian A, Goudarzi G, et al.: Temporal profile of PM10 and associated health effects in one of the most polluted cities of the world (Ahvaz, Iran) between 2009 and 2014. Aeolian Res. 2016; 22: 135–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nISDR U, editor: Hyogo framework for action 2005-2015: building the resilience of nations and communities to disasters. Extract from the final report of the World Conference on Disaster Reduction (A/CONF 206/6). The United Nations International Strategy for Disaster Reduction Geneva. 2005. Reference Source\n\nGuha-Sapir D, Vos F, Below R, et al.: Annual disaster statistical review 2011: the numbers and trends. Centre for Research on the Epidemiology of Disasters (CRED). 2012. Reference Source\n\nWoodruff RE, McMichael T, Butler C, et al.: Action on climate change: the health risks of procrastinating. Aust N Z J Public Health. 2006; 30(6): 567–71. PubMed Abstract | Publisher Full Text\n\nBlashki G, Armstrong G, Berry HL, et al.: Preparing health services for climate change in Australia. Asia Pac J Public Health. 2011; 23(2 Suppl): 133S–43S. PubMed Abstract | Publisher Full Text\n\nPerry RW, Lindell MK, Tierney KJ: Facing the unexpected: Disaster preparedness and response in the United States. Joseph Henry Press; 2001. Publisher Full Text\n\nAsfaw A, Admassie A: The role of education on the adoption of chemical fertiliser under different socioeconomic environments in Ethiopia. Agric Econ. 2004; 30(3): 215–28. Publisher Full Text\n\nDamluji NN: Imperialism reconfigured: the cultural interpretations of the keffiyeh. 2010. Reference Source\n\nKeffiyeh. accessed on 30 October 2018. Reference Source.\n\nMurena F: Measuring air quality over large urban areas: development and application of an air pollution index at the urban area of Naples. Atmos Environ. 2004; 38(36): 6195–202. Publisher Full Text\n\nWiggs GFS, O'Hara S, Wegerdt J, et al.: The dynamics and characteristics of aeolian dust in dryland Central Asia: possible impacts on human exposure and respiratory health in the Aral Sea basin. Geophys J Roy Astron Soc. 2003; 169(2): 142–57. Publisher Full Text\n\nBennion P, Hubbard R, O'Hara S, et al.: The impact of airborne dust on respiratory health in children living in the Aral Sea region. Int J Epidemiol. 2007; 36(5): 1103–10. PubMed Abstract | Publisher Full Text\n\nYang CY, Cheng MH, Chen CC: Effects of Asian dust storm events on hospital admissions for congestive heart failure in Taipei, Taiwan. J Toxicol Environ Health A. 2009; 72(5): 324–8. PubMed Abstract | Publisher Full Text\n\nSchwartz J, Norris G, Larson T, et al.: Episodes of high coarse particle concentrations are not associated with increased mortality. Environ Health Perspect. 1999; 107(5): 339–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee H, Honda Y, Lim YH, et al.: Effect of Asian dust storms on mortality in three Asian cities. Atmos Environ. 2014; 89: 309–17. Publisher Full Text"
}
|
[
{
"id": "43965",
"date": "13 Mar 2019",
"name": "Mohamad Taghi Moghadamnia",
"expertise": [
"Reviewer Expertise Disaster 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\nIn this article, the authors survey preparedness components of health systems in the Eastern Mediterranean Region for effective responses to dust and sand storms using a systematic review approach. It addresses an important topic, especially the preparedness of health system in Eastern Mediterranean Region. Regarding climate change, increasing temperatures and dust storms, awareness of the readiness and response of health providers is important. In this systematic review, the authors attempted to study the studies carried out in this regard. Below are more specific comments:\nIn the abstract, the authors have paid too much attention to the methods, and the results are very limited and incomplete. The keyword “reediness” is wrong, it would be better to write \"readiness”. A large part of the material written at the beginning of the discussion section is related to the introduction, and it would be better to move that section. It is recommended that the authors, after a brief description of the subject framework, focused on the findings of their study and then compare them with the findings of others. In the conclusion, the authors must clearly and explicitly state the message of their 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? Not applicable\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": []
},
{
"id": "46125",
"date": "25 Mar 2019",
"name": "Randah Ribhi Hamadeh",
"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 The paper tackles an important topic on preparedness for dust and sand storm, a common phenomenon in the EMR that has not been much studied. The authors have done a systematic review on the topic and reviewed 30 articles in a comprehensive manner. I have some specific comments:\n\nAbstract\nExpand results section and reduce methods. Line 2: “Numerous health problems caused by DSS will be severely affected regions”: Something missing in the statement. Line 2, results: The preparedness components were divided into three and ten main categories and subcategories: Revise to “The preparedness components were divided into three main categories and ten subcategories”.\nIntroduction\nThe introduction is concise and organized, however I feel that that the last lines “The findings of the study, with proposed recommendations, can directly help health policymakers in preparedness promotion for this phenomenon, pave the way for further studies, and add to the richness of the current knowledge” better be moved to discussion.\n.Results\n“Using of keffiyeh” in end of results, better to write what it means in brackets (Arab Headdress).\nDiscussion\nMove some parts of the discussion to the introduction where they are more relevant particularly from the first paragraph. The authors can also mention in the limitations that some articles might have been published in local scientific journals that are not indexed and thus were missed from the review.\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? Not applicable\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-146
|
https://f1000research.com/articles/8-141/v1
|
01 Feb 19
|
{
"type": "Research Note",
"title": "(Re)emergence of A(H1N1)pdm09 influenza viruses with pandemic markers in the 2018/2019 flu season in the USA",
"authors": [
"Slobodan Paessler",
"Veljko Veljkovic",
"Slobodan Paessler"
],
"abstract": "During the 2009 pandemic, the Centers for Disease and Control and Prevention (CDC) estimated that 43 to 89 million cases of swine flu were reported during a 1-year span, with 1799 deaths in 178 countries worldwide. Now, nine years later, A(H1N1)pdm09 influenza viruses, which caused this pandemic, have reemerged and become the dominant subtype for the current 2018/19 flu season in the USA. The in silico analysis of A(H1N1)pdm09 viruses isolated in USA in October and November 2018, performed using the electronic biology platform “wEB”, showed that these viruses carry previously identified pandemic markers suggesting their increased pandemic potential. Possible consequences of these findings are discussed.",
"keywords": [
"influenza virus",
"pandemic",
"swine flu",
"mutations",
"electronic biology"
],
"content": "Introduction\n\nEach flu season represents a serious public health threat and the challenge for health workers, vaccine producers and researchers due to unpredictable behavior of influenza A viruses. Vaccination is the most effective way to protect against seasonal influenza viruses. As a consequence of the high variability of influenza viruses, it is difficult to select vaccine candidates or predict the vaccine effectiveness (VE) for the upcoming season. Accordingly, the VE against the dominant A strain H3N2 in Australia in 2017 was very low (about 10%), and this country experienced record-breaking numbers of influenza-related hospitalizations and deaths (http://www.health.gov.au/internet/main/publishing.nsf/Content/cda-ozflu-2017.htm). The World Health Organization (WHO) selected the same vaccine strains for the United States for the flu season 2017–2018 and based on data coming from Australia 6 months before the beginning of the flu season, a similarly low VE against the H3N2 viruses was anticipated. This prediction of the potentially low VE in the USA, which was made by leading vaccine experts1, and aggressively promoted by press media, resulted in a very low vaccination rate in the USA. As a consequence, according to the CDC, 2017-2018 was the worst flu season on record, with an estimated 79,000 deaths, including total of 185 pediatric deaths. This number exceeds the previously highest number of flu-associated deaths in children reported during a regular flu season (171 during the 2012-2013 season). Approximately 80% of these deaths occurred in unvaccinated children.\n\nPrevious comparison of H3N2 viruses from 2017 from Australia with viruses collected during the pre-flu season 2017–2018 in the USA was performed using the electronic biology platform “wEB”, developed using methods described previously2. Our analysis demonstrated significant differences between these two groups of viruses3 allowing us to correctly predict higher VE against H3N2 virus for the flu season 2017–2018 in US than in Australia3. This prediction was in contrast with the prediction of VE of 10% based on Australian data1, and it was confirmed at the end of the flu season. CDC officials reported an overall VE of 36%, with VE of 25% against the H3N2 strain. The Armed Forces Health Surveillance Branch Air Force (AFHSB-AF) found that the vaccine provided substantially better protection to military dependents treated at the U.S. Air Force School of Aerospace Medicine, with an overall adjusted VE of 51% and a VE of 37% against H3N2. The Naval Health Research Center (NHRC) determined that the vaccine had even greater effectiveness among the civilians who received care for febrile respiratory illnesses, with an overall adjusted VE of 55% and VE of 52% against H3N2.\n\nIn contrast to the flu season in 2017–2018, which was dominated by H3N2 strains, the 2018–2019 flu season in the USA is heavily dominated by circulating influenza A(H1N1)pdm09 viruses between October 2018 and January 2019 (to access this data, visit https://gis.cdc.gov/grasp/fluview/fluportaldashboard.html and select Season 2018–19). This change in the dominant flu subtype was followed by an early beginning of the flu season and increased hospitalization and death rates for children and healthy young people in the beginning of the flu season for 2018–2019.\n\nHerein, we analyzed sequences of HA of viruses A(H1N1)pdm09 collected in the USA during months of October and November of 2018 using our previously described electronic biology platform (wEB). The most recent analysis revealed that some of the (re)emerging viruses contain mutations previously identified as pandemic markers4. Moreover, the percentage of the viruses belonging to this population is significantly higher in October–November 2018 than in all past nine years. We also demonstrate that most of A(H1N1)pdm09 viruses have changed informational properties previously proposed to correlate with their biological and immunological properties. Possible consequences of these findings are discussed.\n\n\nMethods\n\nWe analyzed the hemagglutinin subunit 1 (HA1) of (i) 33197 A(H1N1)pdm09 viruses collected between May 2009 and December 2017, (ii) 224 human pH1N1 viruses collected in US from October to November 2018 (GISAID database) and 28 human pH1N1 viruses collected in Canada during the same period and (iii) vaccine virus A/Michigan/45/2015 for 2018/2019 flu season. All sequences were stored in the publicly open database GISAID.\n\nThe ISM is a virtual spectroscopy method for the study of the long-range protein-protein interaction. According to this method, described in detail elsewhere2, 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 potentials, determining the electronic properties of amino acid/nucleotides for their intermolecular interactions. The signal obtained is than decomposed in periodical function by Fourier transformation. The result is a series of frequencies and their amplitudes. The obtained frequencies correspond to the distribution of structural motifs with defined physico-chemical characteristics responsible for biological function of the sequence. When compared, proteins that share the same biological or biochemical function(s) possess the code/frequency pairs specific for their common biological properties. The method is insensitive to the location of the motifs and does not require previous alignment of the sequence.\n\nThe ISM served as a base for development of the phylogenetic algorithm for the Informational Spectrum-based Phylogenetic Analysis (ISTREE)5.\n\nThe phylogenetic tree of the HA1 influenza proteins is generated with the ISM-based phylogenetic algorithm ISTREE, previously described in detail elsewhere5. In the presented analysis, we calculated the distance matrix with the distance measure between sequences X1 and X2 defined as:\n\n\n\nwhere A1(F1) and A2(F1) are amplitudes on frequency F1=0.295; A1(F2) and A2(F2) are amplitudes on frequency F2=0.055 in informational spectra on sequences X1 and X2 respectively.\n\n\nResults\n\nPreviously, we showed that the IS frequency F(0.055) of HA1 is responsible for interaction between H1N1/N2 and swine protein(s)/receptor(s) while the IS frequency F(0.295) for the same HA subunit is responsible for the interaction with human protein(s)/receptor(s)4. This suggests that the acquisition of mutations in HA1, leading to increased ratio of these amplitudes at these two frequencies A(0.29536)/A(0.055), are essential for adaptation of swine A/H1N1 viruses to humans, and this is currently used for real-time monitoring of A/H1N1 viruses. It has been also suggested that positions 94, 196 and 274 in HA1 of A(H1N1)pdm09 are hotspots for advantageous mutations for human adaptation of A(H1N1)pdm094. In Figure 1 and Table 1 we present the fraction of A(H1N1)pdm09 viruses with mutations in these hotspots that were collected between May 2009 and December 2017, and viruses isolated in between October and November 2018 (this period corresponds to the beginning of the flu season in the USA). Of note is that all of these mutations in 2018 appeared in A(H1N1)pdm09 viruses collected in October and November of 2018 in the USA.\n\n*28 redundant HA1 with mutation D274N from viruses collected in Hungary are given as one representative sequence\n\nIn Figure 2a we present the consensus informational spectrum (CIS) of A(H1N1)pdm09 viruses isolated in the period of September–November 2018 in the USA. The dominant peak in this CIS corresponds to the frequency F(0.281). Viruses collected in the same period in Canada are characterized with the IS frequency F(0.295) (Figure 2b), which was previously identified as the hallmark of pandemic A(H1N1)pdm094.\n\nThe consensus informational spectrum of A(H1N1)pdm09 viruses with pandemic markers which are collected in October and November 2018 in (a) the USA and (b) Canada.\n\nPreviously, we showed that the amplitude on characteristic frequencies in IS of HA1 from influenza A viruses is indicative of the vaccine efficacy against these viruses3,6. Accordingly, in Figure 3 we present the ISM-based phylogenetic tree of 224 A(H1N1)pdm09 viruses isolated in October and November of 2018 in the USA and the vaccine virus A/Michigan/45/2015. As presented, 101 (45 %) viruses are co-cauterized with the vaccine virus suggesting that the current vaccine can at least efficiently protect against this fraction of analyzed viruses.\n\n\nDiscussion\n\nWe previously showed that the ratio of amplitudes on the characteristic frequencies in IS of HA1 of influenza viruses determines adaptation of animal viruses to human7,8. In silico analysis of pandemic A(H1N1)pdm09 viruses revealed positions 94, 196 and 274 in HA1 as hotspots for mutations, which could increase infectivity of these viruses4. At the very beginning of the last pandemic, when this analysis was performed (May 2009), only three viruses from three countries (Spain, Italy and USA) in the GISAID database contained the mutations in these hot-spots. At the end of the pandemic (late 2010) sequences of 350 viruses with these mutations were deposited in GISAID. This suggests better adaptation of some viruses to humans during pandemic. The fraction of the viruses with these pandemic markers sharply decreased at the end of the pandemic in 2011. However, this viral population continued to slowly grow until 2017 while reaching the same end-pandemic level of 2010 in October and November of 2018 (Figure 1 and Table 1). It is also important to note that all viruses with pandemic markers isolated in October and November of 2018 are from the USA only, and that the fraction of these viruses is significantly higher than at the end of the 2009 pandemic (Table 1). This is a warning sign that the pandemic potential of A(H1N1)pdm09 viruses in the USA has increased. Of note also is that 7 of 9 of these viruses are isolated in patients 5 to 13 years old, suggesting that children could be more susceptible to infection with these viruses than adults.\n\nEach subtype of influenza A viruses is characterized with specific frequency9. Results presented in Figure 2 show that HA1 from US viruses collected at the beginning of the 2018/2019 flu season are characterized with the IS frequency F(0.281). On the contrary, all other A(H1N1)pdm09 viruses isolated in the last 10 years are characterized with IS frequency F(0.295). This strongly suggests possible changes in the interaction profile of these viruses with host proteins. For now, it is not possible to predict how this change will influence pathological and immunological properties of A(H1N1)pdm09 virus.\n\nResults presented in Figure 3 demonstrated that 101 of 224 (45%) A(H1N1)pdm09 viruses isolated in October and November 2018 in the USA are clustered in the ISM-based phylogenetic tree together with the current vaccine virus A/Michigan/45/2015. This suggests that the efficacy of the current seasonal flu vaccine could be about 50%, which corresponds to common protection against A(H1N1)pdm09 viruses in the last years. The problem could arise if viruses which are not compatible with the vaccine take over during the flu season. Of great concern are viruses with very high amplitude ratio A(0.29536)/A(0.055) that also carry pandemic markers (A/New Mexico/25/2018, A/California/70/2018, A/Texas/134/2018). In our opinion these viruses have increased pandemic potential and could represent precursors of the new pandemic virus.\n\nIn conclusion, presented results show: (i) that some pre-seasonal and seasonal A(H1N1)pdm09 viruses collected in October and November 2018 in the USA carry pandemic markers, indicating possibility of their evolution toward new pandemic viruses; (ii) that these US viruses have changed informational properties which determine their interacting profiles with the host, and; (iii) that about 50% of circulating viruses could escape the current flu vaccine and could evolve toward new pandemic viruses. Taking into account presented results, as well as the fact that A(H1N1)pdm09 viruses are dominant in the flu season 2018/2019, causing already high hospitalization and mortality rates in children and young healthy people, further monitoring of evolution of these viruses is urgently needed.\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": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nPaules CI, Sullivan SG, Subbarao K, et al.: Chasing Seasonal Influenza - The Need for a Universal Influenza Vaccine. N Eng J Med. 2018; 378(1): 7–9. PubMed Abstract | Publisher Full Text\n\nVeljkovic V, Paessler S: Possible repurposing of seasonal influenza vaccine for prevention of Zika virus infection [version 2; referees: 2 approved]. F1000Res. 2016; 5: 190. 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; referees: 2 approved]. F1000Res. 2017; 6: 2067. 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\nPerovic V: Novel algorithm for phylogenetic analysis of proteins: application to analysis of the evolution of H5N1 influenza viruses. J Mathem Chem. 2013; 51(8): 2238–2255. Publisher Full Text\n\nPaessler S, Veljkovic V: Using electronic biology based platform to predict flu vaccine efficacy for 2018/2019 [version 2; referees: 2 approved, 1 approved with reservations]. F1000Res. 2018; 7: 298. PubMed Abstract | Publisher Full Text | Free 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\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"
}
|
[
{
"id": "45904",
"date": "03 Apr 2019",
"name": "Guadalupe Ayora-Talavera",
"expertise": [
"Reviewer Expertise Influenza"
],
"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 described the analysis of HA sequences from H1N1pdm09 viruses reported since 2009 to the last influenza season 2018-2019 in the USA. The analysis was performed using the electronic biology platform \"wEB\" and the ISM method. The aim of the study was to determine pandemic markers that could favor the (re)emergency of the H1N1pdm09 virus. They emphasize the role of three hotspots as predictors of pandemic potential according to the analysis performed.\nThe authors refer to residues 94, 196 and 274. They mention in figure 1 and table 1 the proportion of viruses that have mutated in these positions. It would be good to complement table 1 with the amino acid changes that have occurred in these hotspots.\nWhen the authors say “that the pandemic potential of A(H1N1)pdm09 viruses in the USA has increased” what do they mean? They mention three viruses with pandemic markers, the same as 94, 196 and 274? Or different? On what evidence they base the assumption that these viruses could represent precursors of new pandemic viruses?\nMinor comments: The authors should mention what HA numbering are using, the H1 or H3 I recommend being cautious about the terms used as “(re)emergence” or “precursor of a new pandemic”. It is desirable always to include any biological data available to reinforce any conclusion that could have an impact.\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? Partly",
"responses": [
{
"c_id": "4541",
"date": "04 Apr 2019",
"name": "Veljko Veljkovic",
"role": "Author Response",
"response": "Interaction between virus and the host involves two steps: (i) recognition and targeting between virus proteins and receptor (long-range interactions - distances >5A) and (ii) chemical binding between virus and receptor (short-range interactions - distances <5A). The ISM allows analysis of the first step. Mutations, which increase amplitudes on the frequenciesresponsible for long-range interaction between virus and receptor, increase efficacy of the virus-host interaction, and from this point of view increase pandemic potential of influenza viruses. This concept has been experimentally proven (see ref. 8). Based on this concept mutations in positions 94, 196 and 274 are characterized as potential pandemic markers (see refs. 4 and 7) because they increase the likelihood of “finding the receptor”."
}
]
},
{
"id": "46162",
"date": "03 Apr 2019",
"name": "Alexandra S. Gambaryan",
"expertise": [
"Reviewer Expertise influenza viruses",
"receptors",
"host restriction."
],
"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\nInformational spectrum method (ISM) was used for analyzing the HA sequences of the influenza A/H1N1 viruses. The ISM is a virtual spectroscopy method for the study of the long-range protein-protein interaction. 252 human pH1N1 viruses collected in US and Canada from October to November 2018 were compared with 33197 H1N1 pdm09 viruses of 2009-2017 years and with vaccine virus A/Michigan/45/2015 for 2018/2019 flu season. Based on this analysis, it was concluded that: 45% of analyzed viruses of 2018 are co-cauterized with the vaccine virus A/Michigan/45/2015 for 2018/2019 flu season suggesting that this vaccine can at least efficiently protect against this fraction of circulated viruses. It has been also suggested that these viruses carry previously identified pandemic markers (namely substitution in positions 94, 196 and 274 in HA1 suggesting their increased pandemic potential. The authors declare that the acquisition of these mutations in HA1 are essential for adaptation of swine A/H1N1 viruses to humans, by virtue of the interaction with human protein(s)/receptor(s).\nThe main question that can be presented to the authors is the advantage of the Informational spectrum method in comparison with the generally accepted analysis of genomic sequences. The subdivision of circulating viruses into subpopulations is detected by building evolutionary trees. To predict the effectiveness of the vaccine, you can compare the sequence of circulating viruses with the sequence of the vaccine strain in the field of antigenic determinants.\n\nIdentification of mutations that are markers of pandemic viruses is more reliably carried out by direct analysis of sequences, than by changing the averaged characteristics. Moreover, the statement that positions 94, 196 and 274 are pandemic markers is based only on the previous work of the authors of this article, and was not confirmed by other researchers.\nEven more doubtful is the claim that the positions 94, 196 and 274 are responsible for the interaction with the human receptor. The problem of the interaction of influenza viruses with the host receptors has been well studied. Are known the amino acids that are located in the receptor-binding site of HA and form hydrogen bonds with specific atoms of the receptor molecule. The authors do not quote these works, and do not seem to know them.\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? 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? No",
"responses": [
{
"c_id": "4542",
"date": "04 Apr 2019",
"name": "Veljko Veljkovic",
"role": "Author Response",
"response": "Interaction between virus and the host involves two steps: (i) recognition and targeting between virus proteins and receptor (long-range interactions - distances >5A) and (ii) chemical binding between virus and receptor (short-range interactions - distances <5A). The ISM allows analysis of the first step. Mutations, which increase amplitudes on the frequenciesresponsible for long-range interaction between virus and receptor, increase efficacy of the virus-host interaction, and from this point of view increase pandemic potential of influenza viruses. This concept has been experimentally proven (see ref. 8). Based on this concept mutations in positions 94, 196 and 274 are characterized as potential pandemic markers (see refs. 4 and 7) because they increase the likelihood of “finding the receptor”.All remarks from this Referee concern the second step of interaction (chemical binding) between virus and receptor and for this reason are not relevant for results presented in this article (see previous comment)."
}
]
}
] | 1
|
https://f1000research.com/articles/8-141
|
https://f1000research.com/articles/8-139/v1
|
31 Jan 19
|
{
"type": "Software Tool Article",
"title": "Epidemic curves made easy using the R package incidence",
"authors": [
"Zhian N. Kamvar",
"Jun Cai",
"Juliet R.C. Pulliam",
"Jakob Schumacher",
"Thibaut Jombart",
"Jun Cai",
"Juliet R.C. Pulliam",
"Jakob Schumacher"
],
"abstract": "The epidemiological curve (epicurve) is one of the simplest yet most useful tools used by field epidemiologists, modellers, and decision makers for assessing the dynamics of infectious disease epidemics. Here, we present the free, open-source package incidence for the R programming language, which allows users to easily compute, handle, and visualise epicurves from unaggregated linelist data. This package was built in accordance with the development guidelines of the R Epidemics Consortium (RECON), which aim to ensure robustness and reliability through extensive automated testing, documentation, and good coding practices. As such, it fills an important gap in the toolbox for outbreak analytics using the R software, and provides a solid building block for further developments in infectious disease modelling. incidence is available from https://www.repidemicsconsortium.org/incidence.",
"keywords": [
"epicurve",
"incidence",
"epidemics",
"outbreaks",
"R"
],
"content": "Introduction\n\nResponses to infectious disease epidemics use a growing body of data sources to inform decision making (Cori et al., 2017; Fraser et al., 2009; WHO Ebola Response Team et al., 2014; WHO Ebola Response Team et al., 2015). While new data—such as whole genome pathogen sequences—are increasingly useful complements to epidemiological data (Gire et al., 2014), epidemic curves—which describe the number of new cases through time (incidence)—remain the most important source of information, particularly early in an outbreak. Specifically epidemic curves(often referred to as ‘epicurves’) represent the number of new cases per time unit based on the date or time of symptom onset.\n\nWhile conceptually simple, epicurves are useful in many respects. They provide a simple, visual outline of epidemic dynamics, which can be used for assessing the growth or decline of an outbreak (Barrett et al., 2016; Fitzgerald et al., 2014; Jernberg et al., 2015; Lanini et al., 2014; Nhan et al., 2018) and therefore informing intervention measures (Meltzer et al., 2014; WHO Ebola Response Team et al., 2014; WHO Ebola Response Team et al., 2015). In addition, epicurves also form the raw material used by a range of modelling techniques for short-term forecasting (Cori et al., 2013; Funk et al., 2018; Nouvellet et al., 2018; Viboud et al., 2018) as well as in outbreak detection algorithms from syndromic surveillance data (Farrington & Andrews, 2003; Unkel et al., 2012).\n\nBecause of the increasing need to analyse various types of epidemiological data in a single environment using free, transparent and reproducible procedures, the R software (R Core Team, 2017) has been proposed as a platform of choice for epidemic analysis (Jombart et al., 2014). But despite the existence of packages dedicated to time series analysis (Shumway & Stoffer, 2010) as well as surveillance data (Höhle, 2007), a lightweight and well-tested package solely dedicated to building, handling and plotting epidemic curves directly from linelist data (e.g. a spreadsheet where each row represents an individual case) is still lacking.\n\nHere, we introduce incidence, an R package developed as part of the toolbox for epidemics analysis of the R Epidemics Consortium (RECON) which aims to fill this gap. In this paper, we outline the package’s design and illustrate its functionalities using a reproducible worked example.\n\n\nMethods\n\nThe philosophy underpinning the development of incidence is to ‘do the basics well’. The objective of this package is to provide simple, user-friendly and robust tools for computing, manipulating, and plotting epidemic curves, with some additional facilities for basic models of incidence over time.\n\nThe general workflow (Figure 1) revolves around a single type of object, formalised as the S3 class incidence. incidence objects are lists storing separately a matrix of case counts (with dates in rows and groups in columns), dates used as breaks, the time interval used, and an indication of whether incidence is cumulative or not (Figure 1). The incidence object is obtained by running the function incidence() specifying two inputs: a vector of dates (representing onset of individual cases) and an interval specification. The dates can be any type of input representing dates including Date and POSIXct objects, as well as numeric and integer values. The dates are aggregated into counts based on the user-defined interval representing the number of days for each bin. The interval can also be defined as a text string of either \"week\", \"month\", \"quarter\", or \"year\" to represent intervals that can not be defined by a fixed number of days. For these higher-level intervals, an extra parameter—standard—is available to specify if the interval should start at the standard beginning of the interval (e.g. weeks start on Monday and months start at the first of the month). incidence() also accepts a groups argument which can be used to obtain stratified incidence. The basic elements of the incidence object can be obtained by the accessors get_counts(), get_dates(), and get_interval().\n\nThe raw data is depicted in the top left as either a vector of dates for each individual case (typical usage) or a combination of both dates and a matrix of group counts. The incidence object is created from these where it checks and validates the timespan and interval between dates. Data subsetting and export is depicted in the upper right. Data visualization is depicted in the lower right. Addition of log-linear models is depicted in the lower left.\n\nThis package facilitates the manipulation of incidence objects by providing a set of handler functions for the most common tasks. The function subset() can be used for isolating case data from a specific time window and/or groups, while the [ operator can be used for a finer control to subset dates and groups using integer, logical or character vectors. This is accomplished by using the same syntax as for matrix and data.frame objects, i.e. x[i, j] where x is the incidence object, and i and j are subsets of dates and groups, respectively.\n\nThe function pool() can be used to merge several groups into one, and the function cumulate() will turn incidence data into cumulative incidence. To maximize interoperability, incidence objects can also be exported to either a matrix using get_counts() or a data.frame using as.data.frame(), including an option for a ‘long’ format which is readily compatible with ggplot2 (Wickham, 2016) for further customization of graphics.\n\nIn line with RECON’s development guidelines, the incidence package is thoroughly tested via automatic tests implemented using testthat (Wickham, 2011), with an overall coverage nearing 100% at all times. We use the continuous integration services travis.ci and appveyor to ensure that new versions of the code maintain all existing functionalities and give expected results on known datasets, including matching reference graphics tested using the visual regression testing implemented in vdiffr (Henry et al., 2018). Overall, these practices aim to maximise the reliability of the package, and its sustainable development and maintenance over time.\n\nMany different approaches can be used to model, and possibly derive predictions from incidence data (e.g. Cori et al., 2013; Nouvellet et al., 2018; Wallinga & Teunis, 2004), and are best implemented in separate packages (e.g. Cori et al., 2013). Here, we highlight three simple functionalities in incidence for estimating parameters via modeling or bootstrap and the two specialized data classes that are used to store the models and parameter estimates.\n\nAs a basic model, we implement the simple log-linear regression approach in the function fit(), which can be used to fit exponential increase or decrease of incidence over time by log-transforming case counts and applying a linear regression on these transformed data. The log-linear regression model is of the form log(y) = r × t + b where y is the incidence, r is the growth rate, t is the number of days since the start of the outbreak, and b is the intercept. This approach estimates a growth rate r (the slope of the regression), which can in turn be used for estimating the doubling or halving time of the epidemic, and with some knowledge of the serial interval, for approximating the reproduction number, R0 (Wallinga & Lipsitch, 2007).\n\nIn the presence of both growing and decreasing phases of an epidemic, the date representing the peak of the epidemic can be estimated. In incidence, this can be done in two ways. The function estimate_peak() uses multinomial bootstrapping to estimate the peak, assuming that a) reporting is constant over time, b) the total number of cases is known, and c) the bootstrap never samples zero-incidence days. This function returns the estimated peak with a confidence interval along with the boostrap estimates. Alternatively, the function fit_optim_split() can be used to detect the optimal turning point of the epidemic and fit two separate models on either side of the peak. This is done by maximizing the combined mean adjusted R2 value from the two models (Figure 1, Figure 5).\n\nThe fit() function returns an incidence_fit object and the fit_optim_split() function returns an incidence_fit_list object, which is a specialized object designed to contain an unlimited number of (potentially nested) incidence_fit objects. While the incidence package returns incidence_fit objects containing log-linear models by default, they can be constructed from any model from which it’s possible to extract the growth rate (r) and predict incidence along the model. Both object classes can be plotted separately or added to an existing epicurve using the function add_incidence_fit() (Figure 5).\n\nThe minimal system requirements for successful operation of this package is R version 3.1.\n\n\nUse cases\n\nTwo worked examples are used to demonstrate the functionality and flexibility of the incidence package. The first example illustrates how to compute and manipulate stratified weekly incidence directly from a line-list, while the second example shows how to import pre-computed daily incidence and fit a log-linear model to estimate growth rate (r) and doubling time for the growing phase1.\n\nIn this first example, we use the dataset ebola_sim_clean in the outbreaks package, which provides a linelist for a fictitious outbreak of Ebola Virus Disease (EVD) that matches some key epidemiological properties (e.g. serial intervals, reproduction numbers) of the West African Ebola outbreak of 2014–2015 (WHO Ebola Response Team et al., 2014).\n\n1) Importing data\n\nFirst, we load the dataset ebola_sim_clean from the outbreaks package. The dataset contains 5,829 cases of 9 variables, among which the date of symptom onset ($date_of_onset) and the name of the hospital ($hospital) are used for computing the weekly epicurves stratified by hospitals.\n\n\n\n2) Building the incidence object\n\nThe weekly incidence stratified by hospitals is computed by running the function incidence() on the Date variable dat1$date_of_onset with the arguments interval = 7 and groups = dat1$hospital. The incidence object i.7.group is a list with class of incidence for which several generic methods are implemented, including print.incidence() and plot.incidence(). Typing incidence object i.7.group implicitly calls the specific function print.incidence() and prints out the summary of the data and its list components. The 5,829 cases (the total number of cases stored in the $n component) with dates of symptom onset ranging from 2014-04-07 to 2015-04-27 (spanning from 2014-W15 to 2015-W18 in terms of the ISO 8601 standard for representing weeks) are used for building the incidence object i.7.group. The $counts component contains the actual incidence for defined bins, which is a matrix with one column per group. Here $count is a matrix with 56 rows and 6 columns as groups by hospital with 6 factor levels are specified. The bin size in number of days is stored in the $interval component. In this example, 7 days suggests that weekly incidence is computed, while by default, daily incidence is computed with the argument interval = 1. The $dates component contains all the dates marking the left side of the bins, in the format of the input data (e.g. Date, integer, etc.). The $timespan component stores the length of time (in days) for which incidence is computed. The $cumulative component is a logical indication whether incidence is cumulative or not.\n\nThe generic plot() method for incidence objects calls the specific function plot.incidence(), which makes an incidence barplot using the ggplot2 package. Hence, customization of incidence plot can benefit from the powerful graphical language from ggplot2.\n\n\n\nNote that when weekly incidence is computed from dates, like in this example, the ISO 8601 standard weeks are used by default with the argument standard = TRUE in the incidence() function. Under this situation, an extra component of $isoweek is added to the incidence object i.7.group to store those weeks in the ISO 8601 standard week format “yyyy-Www”, and the $dates component stores the corresponding first days of those ISO weeks. Meanwhile the x-axis tick labels of the weekly incidence plot are in the ISO week format “yyyy-Www” (see Figure 2) rather than in the date format “yyyy-mm-dd” as the argument labels_iso_week in the plot() function is by default TRUE when plotting the ISO week-based incidence objects.\n\n3) Manipulate the incidence object\n\nIn the above visualisation, it can be difficult to see what the dynamics were in the early stages of the epidemic. If we want to see the first 18 weeks of the outbreak in the four major hospitals, we can use the [ operator to subset the rows and columns, which represent weeks and hospitals, respectively, in this particular incidence object.\n\n\n\nHere, because of the few numbers of cases in the first few weeks, we have also highlighted each case using show_cases = TRUE (Figure 3). We’ve also used a different color palette to differentiate between the subsetted data and the full data set.\n\nAs shown in Figure 2, the missing hospital name (NA) is treated as a separate group, resulting from the default of the argument na_as_group = TRUE in the incidence() function. This argument can be set to FALSE to not include data with missing groups in the object.\n\nThe datasets zika_girardot_2015 and zika_sanandres_2015 used in the second example are also from the outbreaks package. These datasets describe the daily incidence of Zika virus disease (ZVD) in, respectively, Girardot and San Andres Island, Colombia from September 2015 to January 2016. For details on these datasets, please refer to Rojas et al. (2016).\n\n1) Import pre-computed daily incidence\n\nzika_girardot_2015 and zika_sanandres_2015 are data frames with the same variables date and cases. In order to obtain a more complete picture of the epidemic dynamics of ZVD in Colombia, we merge these two data.frames into a single one, dat2, by variable date. As dat2 is already pre-computed daily incidence rather than a vector of dates such as those in example 1, we can directly convert it into an incidence object grouped by geographical locations, i.group, by using the as.incidence() function. This shows the flexibility of the incidence package in making incidence objects. Using the pool() function, the daily incidence stratified by locations, i.group, can be collapsed into an incidence object without groups, i.pooled. The stratified and pooled daily incidence plots of ZVD in Colombia are shown in Figure 4, from which we can see that the epidemic of ZVD occurred earlier in San Andres Island than in Girardot.\n\n\n\n(A) stratified and (B) pooled daily incidence plots of ZVD in Colombia, September 2015 to January 2016.\n\nAs shown in Figure 4B, the pooled daily incidence in Colombia shows approximately exponential phases before and after the epidemic peak. Therefore, we fit two log-linear regression models around the peak to characterize the epidemic dynamics of ZVD in Colombia. Such models can be separately fitted to the two phases of the epicurve of i.pooled using the fit() function, which, however, requires us to know what date should be used to split the epicurve in two phases (see the argument split in the fit() function). Without any knowledge on the splitting date, we can turn to the fit_optim_split() function to look for the optimal splitting date (i.e. the one maximizing the average fit of both models) and then fit two log-linear regression models before and after the optimal splitting date.\n\n\n\nThe returned object fos is a list with 4 components. The $split component suggests that the optimal splitting date is 2015-11-15. The $fit component is an incidence_fit_list containing two incidence_fit objects named ‘before’ and ‘after’. These each contain the information extracted from the fitted log-linear regression models. Printing the $fit component shows a daily growth rate r of 0.067 and its 95% confidence interval (CI) ([0.059, 0.074]), and a doubling time of 10.4 days (95% CI, [9.31, 11.8]) during the first phase, and a daily decreasing rate r of -0.048 (95% CI, [-0.054, -0.042]), and a halving time of 14.4 days (95% CI, [12.7, 16.6]) during the second.\n\nThe predictions and their 95% CIs from the two incidence_fit objects, ‘before’ and ‘after’, can be added to the existing incidence plot of i.pooled using the piping-friendly function add_incidence_fit(). As shown in Figure 5, based on visual comparison of models and data, these two log-linear regression models provide a decent approximation for the actual dynamics of the epidemic (adjusted R2 = 0.83 and 0.77 for the increasing and decreasing phases, respectively).\n\n\nConclusion\n\nThis article has described the package incidence and its features—which include three lightweight data classes and utilities for data manipulation, plotting, and modeling. We have shown that an incidence object can flexibly be defined at different datetime intervals with any number of stratifications and be subset by groups or dates. The most important aspects of this package are use-ability and interoperability. For both field epidemiologists and academic modellers, the data received are often in the form of line-lists where each row represents a single case. We have shown that these data can easily be converted to an incidence object and then plotted with sensible defaults in two lines of code.\n\nWe have additionally shown that because the data are aggregated into a matrix of counts, it becomes simple to perform operations related to peak-finding, model-fitting, and exportation (e.g. using as.data.frame()) into different formats. Thus, because it has built-in tools for aggregation, visualisation, and model fitting, the incidence package is ideal for rapid generation of reports and estimates in outbreak response situations where time is a critical factor.\n\n\nSoftware availability\n\nincidence available from: https://www.repidemicsconsortium.org/incidence Code to reproduce all figures can be found by running demo (\"incidence-demo\", package = \"incidence\") from the R console with the incidence package installed.\n\nSource code available from: https://github.com/reconhub/incidence\n\nArchived source code as at time of publication: https://doi.org/10.5281/zenodo.2540217 (Jombart et al., 2019)\n\nSoftware license: MIT\n\n\nData availability\n\nDatasets used in the worked examples are from the outbreaks package:\n\nebola_sim_clean: https://github.com/reconhub/outbreaks/blob/master/data/ebola_ sim_clean.RData\n\nzika_girardot_2015: https://github.com/reconhub/outbreaks/blob/master/data/zika_ girardot_2015.RData\n\nzika_sanandres_2015: https://github.com/reconhub/outbreaks/blob/master/data/zika_sanandres_2015.RData",
"appendix": "Grant information\n\nThe authors acknowledge financial support from the Global Challenges Research Fund (GCRF) for the project ‘RECAP – research capacity building and knowledge generation to support preparedness and response to humanitarian crises and epidemics’ managed through RCUK and ESRC (ES/P010873/1), from the UK Public Health Rapid Support Team, which is funded by the United Kingdom Department of Health and Social Care, and from the National Institute for Health Research - Health Protection Research Unit for Modelling Methodology.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe would like to thank Michael Höhle for discussion about the caveats for estimate_peak(), the R developer community for constantly improving our working environment, github for hosting our project, travis, appveyor and codecov for providing free continuous integration resources, and the RECON community.\n\n\nFootnotes\n\n1 Negative values of r in incidence are reported as halving times instead of doubling times and decreasing phase instead of growing phase\n\n\nReferences\n\nBarrett P, Chaintarli K, Ryan F, et al.: An ongoing measles outbreak linked to a suspected imported case, Ireland, April to June 2016. Euro Surveill. 2016; 21(27). PubMed Abstract | Publisher Full Text\n\nCori A, Donnelly CA, Dorigatti I, et al.: Key data for outbreak evaluation: building on the Ebola experience. Philos Trans R Soc Lond B Biol Sci. 2017; 372(1721): pii: 20160371. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCori A, Ferguson NM, Fraser C, et al.: A new framework and software to estimate time-varying reproduction numbers during epidemics. Am J Epidemiol. 2013; 178(9): 1505–1512. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFarrington P, Andrews N: Outbreak detection: Application to infectious disease surveillance. In Monitoring the Health of Populations. Oxford University Press, New York. 2003. Publisher Full Text\n\nFitzgerald M, Thornton L, O'Gorman J, et al.: Outbreak of hepatitis A infection associated with the consumption of frozen berries, Ireland, 2013--linked to an international outbreak. Euro Surveill. 2014; 19(43): pii: 20942. PubMed Abstract | Publisher Full Text\n\nFraser C, Donnelly CA, Cauchemez S, et al.: Pandemic potential of a strain of influenza A (H1N1): early findings. Science. 2009; 324(5934): 1557–1561. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFunk S, Camacho A, Kucharski AJ, et al.: Real-time forecasting of infectious disease dynamics with a stochastic semi-mechanistic model. Epidemics. 2018; 22: 56–61. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGire SK, Goba A, Andersen KG, et al.: Genomic surveillance elucidates Ebola virus origin and transmission during the 2014 outbreak. Science. 2014; 345(6202): 1369–1372. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHenry L, Sutherland C, Hong D: vdiffr: Visual regression testing and graphical diffing. 2018. Reference Source\n\nHöhle M: surveillance: An R package for the monitoring of infectious diseases. Comput Stat. 2007; 22(4): 571–582. Publisher Full Text\n\nJernberg C, Hjertqvist M, Sundborger C, et al.: Outbreak of Salmonella Enteritidis phage type 13a infection in Sweden linked to imported dried-vegetable spice mixes, December 2014 to July 2015. Euro Surveill. 2015; 20(30): pii: 21194. PubMed Abstract | Publisher Full Text\n\nJombart T, Aanensen DM, Baguelin M, et al.: OutbreakTools: a new platform for disease outbreak analysis using the R software. Epidemics. 2014; 7(0): 28–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJombart T, Kamvar ZN, Cai J, et al.: reconhub/incidence 1.5 (Version 1.5). Zenodo. 2019. http://www.doi.org/10.5281/zenodo.2540217\n\nLanini S, Capobianchi MR, Puro V, et al.: Measles outbreak on a cruise ship in the western Mediterranean, February 2014, preliminary report. Euro Surveill. 2014; 19(10): pii: 20735. PubMed Abstract | Publisher Full Text\n\nMeltzer MI, Atkins CY, Santibanez S, et al.: Estimating the future number of cases in the Ebola epidemic--Liberia and Sierra Leone, 2014-2015. MMWR Suppl. 2014; 63(3): 1–14. PubMed Abstract\n\nNhan LNT, Hong NTT, Nhu LNT, et al.: Severe enterovirus A71 associated hand, foot and mouth disease, Vietnam, 2018: preliminary report of an impending outbreak. Euro Surveill. 2018; 23(46): 1800590. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNouvellet P, Cori A, Garske T, et al.: A simple approach to measure transmissibility and forecast incidence. Epidemics. 2018; 22: 29–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nR Core Team: R: A language and environment for statistical computing. 2017. Reference Source\n\nRojas DP, Dean NE, Yang Y, et al.: The epidemiology and transmissibility of zika virus in girardot and san andres island, colombia, september 2015 to january 2016. Euro Surveill. 2016; 21(28). PubMed Abstract | Publisher Full Text | Free Full Text\n\nShumway RH, Stoffer DS: Time Series Analysis and Its Applications: With R Examples. Springer Science & Business Media, 2010. Reference Source\n\nUnkel S, Farrington CP, Garthwaite PH, et al.: Statistical methods for the prospective detection of infectious disease outbreaks: a review. J R Stat Soc Ser A Stat Soc. 2012; 175(1): 49–82. Publisher Full Text\n\nViboud C, Sun K, Gaffey R, et al.: The RAPIDD ebola forecasting challenge: Synthesis and lessons learnt. Epidemics. 2018; 22: 13–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWallinga J, Lipsitch M: How generation intervals shape the relationship between growth rates and reproductive numbers. Proc Biol Sci. 2007; 274(1609): 599–604. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWallinga J, Teunis P: Different epidemic curves for severe acute respiratory syndrome reveal similar impacts of control measures. Am J Epidemiol. 2004; 160(6): 509–516. PubMed Abstract | Publisher Full Text\n\nWickham H: ggplot2: elegant graphics for data analysis. UseR! Springer. 2016. Reference Source\n\nWickham H: testthat: Get started with testing. R J. 2011; 3(1): 5–10. Reference Source\n\nWHO Ebola Response Team, Aylward B, Barboza P, et al.: Ebola virus disease in West Africa--the first 9 months of the epidemic and forward projections. N Engl J Med. 2014; 371(16): 1481–1495. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWHO Ebola Response Team, Agua-Agum J, Ariyarajah A, et al.: West African Ebola epidemic after one year--slowing but not yet under control. N Engl J Med. 2015; 372(6): 584–587. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "45526",
"date": "03 Apr 2019",
"name": "Benjamin M. Bolker",
"expertise": [
"Reviewer Expertise epidemic modeling",
"biostatistics",
"evolution of virulence"
],
"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 have written an R package that does some useful tasks in aggregating and manipulating incidences and plotting them. For the most part the paper is clearly written and technically correct, and the package seems well-constructed. The abstract says that the package is \"solely dedicated to building, handling and plotting epidemic curves\", although it also fits simple epidemic models to incidence data.\nOverall I think this package is sensible and well-designed. My only major concern is with the log-linear fitting procedures; on the one hand, the authors say clearly that there are many fitting procedures which cannot (and should not) all be squeezed into this package, and I suspect that they don't intend these fits to be taken as very precise measures of the growth rate. On the other hand, I'm quite concerned by the possibility that users will take growth rates and confidence intervals estimated by these simplistic fits too seriously. There are a number of delicate issues around the estimation of epidemic growth rates:\nAssume log-Normal incidence (as done by a log-linear fit), or make other distributional assumptions (e.g. Poisson or negative binomial)?\n\nIf using exponential or log-linear fits, what time window should one use to capture enough of the beginning of the epidemic but not bias the growth rate downward by capturing the saturation phase of the epidemic (Ma et al., 20141)?\n\nShould one allow for the influence of both process and measurement error (King et al., 20152)?\n\nIs growth exponential or sub-exponential (Viboud et al., 20163)?\n\nI think I would have preferred that, if they were not going to go deeply into this area, that the authors instead provide some sort of non-parametric smooth fit (perhaps constraining the changes to be monotonically increasing before the peak and decreasing after the peak) rather than oversimplifying in this way.\nThe minimal change that is needed to the document is a stronger set of caveats/warnings to users that the log-linear model may have major shortcomings in some circumstances, and should not be accepted unquestioningly.\nThe package suggests that the fit() function is extendable to allow other fitting methods (\"While the incidence package returns incidence_fit objects containing log-linear models by default, they can be constructed from any model from which it’s possible to extract the growth rate (r) and predict incidence along the model\"); I wasn't able to figure out how to achieve that goal.\nThe main use of the package is for converting from line lists to aggregated incidence data. It would be useful (I can't tell if it's possible) to easily be able to aggregate data that are already in date/incidence form to coarser scales, or to approximately disaggregate incidence data.\nI had trouble fitting epidemic curves to data sets that only included incidence up to the peak or shortly beyond (it seems that the model was automatically trying to estimate a decline rate as well as an increase rate). This seems like an odd choice for a tool that people may be using to work with data from emerging epidemics that have not yet peaked.\n\nMinor comments:\np. 3, paragraph 1: missing space before (\n\np. 3, paragraph 3: maybe R \"language\" or \"environment\" rather than \"software\" (R defines itself as \"a language and environment for statistical computing and graphics\")?\n\nLast paragraph of intro: comma before \"which aims to fill this gap\"?\n\nMethods, paragraph 2: Probably don't need to say \"numeric and integer\" values, \"numeric\" would suffice (is.numeric(1L) is TRUE in R).\n\nMethods, paragraph 2: \"can not\" -> \"cannot\".\n\nHow do subsets and indexing work with dates?\n\nFigure 1 caption: delete \"both\"?\n\n\"created from these where it checks and validates\" -> \"created from these components after checking and validating\" ?\n\n\"[Fitting] of log-linear models is depicted in the lower left\"? (Technically, the addition of log-lin models to the plot is depicted along the bottom edge, or in the lower right).\n\np. 5: \"boo[t]strap\".\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": "45525",
"date": "25 Apr 2019",
"name": "Quirine ten Bosch",
"expertise": [
"Reviewer Expertise infectious disease modeling",
"vector-borne disease",
"multi-host systems"
],
"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 introduce a package that aids in processing linelist data (i.e., in which each row represents a case) for building, handling, and plotting epidemic curves. This tool fills a gap within the larger epidemics toolbox of the R Epidemics Consortium in ‘getting the basics right’, or otherwise put, getting data common to outbreak situations in the right format for further analysis.\n\nI agree with the authors that this is a helpful tool. This is particularly true during outbreaks, when linelist data need to be processed quickly and disseminated to relevant actors. The procedures are straightforward and well presented. The examples given in this work are well chosen and provide a good starting point for working with this package. I only have a few comments pertaining to i) the loglinear model implementation and presentation and ii) the integration with other packages.\nWhile the package is meant to focus on processing and visualization of data, the authors have added a modeling-capability that estimates epidemic growth rates. The respective function contains an option for estimating the peak of the outbreak and fit the exponential increase or decrease during the epidemic. Basing the time window for exponential growth on the timing of the peak will underestimate the growth rate as growth will no longer be exponential just before the peak of the epidemic. A more careful description in the MS is needed to highlight this and other shortcomings of this approach. Other methods to estimate the best time window, such as those used in the R0-package1, could prove helpful and be implemented in the package. Further, the choice of fitting the growth rates to the aggregated data rather than the two distinct regions (Girardot and San Andres) is a bit uncomfortable. Particularly so because, as the authors acknowledge, there seem to be two quite distinct outbreaks.\n\nThe authors highlight the gap within the epidemics toolbox is filled. It would be nice to see a little bit more context on what packages can easily work with their data structures and what analyses can hence readily be done.\nMinor comments:\nPage 7: group_names(1.7.group) should be i.7.group\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": "48153",
"date": "07 May 2019",
"name": "Bertrand Sudre",
"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\nMajor comments\nThe proposed package is addressing some of the topmost descriptive elements of any epidemiological data set, namely a systematic time-place-person description. With regards to epidemiological curves, a limited number of dedicated packages addressing these aspects were available at the time of this package release (mainly: epitools, https://cran.r-project.org/web/packages/epitools/epitools.pdf [last update: October 26, 2017] and EpiCurve, https://cran.rstudio.com/web/packages/EpiCurve/EpiCurve.pdf [last update: April 24, 2018]). The alternative was a tailor-made, and time consuming, customization based on existing dedicated packages (for instance using customized bar charts geoms from ggplot2). While the data storage and representation is well addressed by the authors, the proposed package offers: i) some basic utilities for outbreak description across time, ii) basic tool for outbreak modelling and iii) a standard for data storage to enhance interoperability between released projects and packages from the R epidemic consortium. In the introduction a short overview of the alternative tools mentioned above should be provided to the reader, together with the new added-values of the current package. It would be an asset in order to ensure a benchmarking analysis with pre-existing resources.\n\n“Here, we introduce incidence, an R package developed as part of the toolbox for epidemics analysis of the R Epidemics Consortium (RECON) which aims to fill this gap. In this paper, we outline the package’s design and illustrate its functionalities using a reproducible worked example.” According to RECON website: package incidence corresponds to “Computation, handling, visualisation and simple modelling of incidence”. It is honourable that the name of a package “incidence” is the choice of the creators of course. It is true that incident cases are all individuals who change to non-disease status from disease, so in this way “incidence” could refer to the occurrence of new cases. In recent modelling papers, the term incidence has been associated to count time series under the same approach, so-called “incidence time series”1. Nevertheless, the current name can be misleading for a certain number of epidemiologists. The reason is that in epidemiology the term incidence is traditionally associated to a measure of morbidity, so-called 'Incidence proportion' (or attack rate or risk; for more information see: https://www.cdc.gov/ophss/csels/dsepd/ss1978/lesson3/section2.html). The latter comprised the numerator (= count of case used for a raw epi-curve) but as well a denominator representing the population at risk during the selected time interval. At the first glance, the target audience reading the package name might believe that the package is dedicated to incidence calculation rather than epidemiological curve graphic representation and some basic modelling utilities. Indeed, the package is presented as being able to compute, handle and visualize time-related count data through epi-curve and additional derived features which are not related to measure of incidence (proportion) strictly speaking. You may wish to consider adding such features and include new capabilities to this package which can cover calculation/representation of incidence in epidemiology. For instance, adding a slot for population data to populate the denominator of an incidence proportion calculation. This can be completed by dedicated graphic outputs with points for each incidence values and line through all data points (geom line). Factor-specific incidence rate (and corresponding CI) can be considered as further extensions (factor: sex-, age-, or any other factor). The advantage is the possibility to overlay and compare several incidence line charts coming for different locations (e.g. attack rate for different health districts or for several population group). It is a suggestion and it is acceptable that the authors would keep the package focus on basic utilities and epicurve representation.\n\nWith regards to the structure of the article, it would be easier to start with an example based on simple line listing (see comment on figure 1) with simple epicurve without stratification (with different timing hour, day. week), then move to more complex representation (stratification with various colour in the legend; +/- facetting), and then an example of tailor-made polished figure (see code below). In doing so the figure 1 which would start for a line-list format would be easier to understand.\nMinor comments\nSection: Author’s affiliation\n\nComment 1:\n“Department of Infectious Disease Epidemiology , School of Public Health, Imperial\" - Space to remove after epidemiology “Epidemiological Modelling and Analysis (SACEMA),, Stellenbosch University, Stellenbosch” - Double coma to remove\nSection: Introduction\n\nComment 2:\n“Responses to infectious disease epidemics use a growing body of data sources to inform decision making (Cori et al., 2017; Fraser et al., 2009; WHO Ebola Response Team et al., 2014; WHO Ebola Response Team et al., 2015). While new data—such as whole genome pathogen sequences—are increasingly useful complements to epidemiological data (Gire et al., 2014), epidemic curves—which describe the number of new cases through time (incidence)—remain the most important source of information, particularly early in an outbreak.” - Rephrase avoiding four \"—\" in the same sentence.\nPartial genome sequencing is useful to complement epidata/infect. disease epi. If you wish to highlight the WGS/extensive sequencing, consider several aspects:\nConfirm change in infectiousness or epidemiology (CFR, change in clinical presentation, shift in risk group.) Relationships between cases (mapping Lassa, Ebola transmission chains ...) Emergence and patterns of transmission (vector, host, reservoir)\nSuggested examples from the recent literature:\nKraemer MUG, Cummings DAT, Funk S, Reiner RC, Faria NR, Pybus OG, Cauchemez S. Reconstruction and prediction of viral disease epidemics. Epidemiol Infect. 2018 Nov 5:1-7. doi: 10.1017/S0950268818002881. [Epub ahead of print] PubMed PMID: 30394230; PubMed Central PMCID: PMC6398585.2\n\nSiddle KJ et al. Genomic Analysis of Lassa Virus during an Increase in Cases in Nigeria in 2018. N Engl J Med. 2018 Nov.3 Kafetzopoulou LE et al, Metagenomic sequencing at the epicenter of the Nigeria 2018 Lassa fever outbreak. Science. 2019 Jan 4;363(6422):74-77.4\nComment 3:\n“which describe the number of new cases through time (incidence)—remain” -Proposed change of 'through' to 'over'\nComment 4:\n“important source of information, particularly early in an outbreak.” - Not only. This can be helpful to look at the magnitude and pattern (recurrent environmental sources, detect outliers ...).\nComment 5:\n“Specifically epidemic curves(often” - Missing space.\nComment 6:\n“[…] often referred to as ‘epicurves’) represent the number of new cases […]” - New cases = incident case in epidemiology. Might be more accurate to say, \"incident cases of a disease” and add within bracket “(corresponding to the vertical y-axis)”. “[…] per time unit based […]” - Consider 'time interval' instead of 'unit' as ‘unit’ can be misunderstood (e.g. unit = hour, min, sec). The interval is at the user's discretion (3 hours = 'time interval’). “[…]: on the date or time of symptom onset” - Consider replacing symptoms by “time of illness onset among cases” (and add within bracket “(corresponding to the vertical x-axis)”.\nComment 7:\n“provide a simple, visual outline of epidemic dynamics, which can be used for assessing the growth or decline of an” - That is the main purpose in practice. You may consider being more assertive, by replacing \"can\" with \"is\". Very well-known added-values. No need for six references to support this statement.\nComment 8:\n“[…] by a range of modelling techniques for short-term forecasting (Cori et al., 2013; Funk et al., 2018; Nouvellet et al., 2018; Viboud et al., 2018)” - Could you please consider developing with few sentences pointing to ad hoc examples in relation with the features of the cited references.\nComment 9:\n“[…] as well as in outbreak detection algorithms from syndromic surveillance data” - Consider more references that are recent. Note such analytical frameworks are not restricted to syndromic surveillance but to any regularly collected count data from epidemiological time-series in health surveillance system (syndromic or not).\nComment 10:\n“[…] But despite the existence of packages dedicated to time series analysis (Shumway & Stoffer, 2010) as well as surveillance data (Höhle, 2007)” - Time series analysis is a generic term, consider “time series of epidemiological data” to be more accurate. Numerous time-series packages are available for count data that can be \"re-used\" for human epi.\nComment 11:\n“a lightweight and well-tested package solely dedicated to building, handling and plotting epidemic curves directly from linelist data (e.g. a spreadsheet where each row represents an individual case) is still lacking.” - Perhaps “lightweight” is something relative in package development and might be changed. More importantly, as mentioned above, please, consider mentioning other previous packages supporting specifically epi-curve creation (for instance: epitools, EpiCurve). These packages can manage aggregated/no aggregated data with and without factor. Current presented package has an added-value is its interoperability, the presence of simple modelling tools and further graph customization. Stricto sensu, epicurve were able to be done in R form user customization of ggplot2 and cited package.\nSection: Methods\nComment 12:\n“[...] some additional facilities for basic models of incidence over time.” - Consider “additional features” instead.\nComment 13:\n“[...] an indication of whether incidence is cumulative or not” - Consider “whether the case count is cumulative”.\nComment 14:\n“[…] representing dates including Date and POSIXct objects”. - The EpiCurve package supports hourly data. This intends to cope with some peculiar situation in which an hourly epi-curve can be of interest (acute food intoxication or environmental source). Could give an example with hourly data (it should be fairly easy as POSIXct is date format in the ggplot2 framework). See an example here: https://www.cambridge.org/core/journals/epidemiology-and-infection/article/unusually-high-illness-severity-and-short-incubation-periods-in-two-foodborne-outbreaks-of-salmonella-heidelberg-infections-with-potential-coincident-staphylococcus-aureus-intoxication/FED73E4BCEFC109DA327F2909B80BBA7/core-reader\nComment 15:\n“The dates are aggregated into counts based on the user-defined interval representing the number of days for each bin.”- Consider 'user-defined time interval' instead of user-defined interval. “number of days for each bin” - “days” is too restrictive. It is just the chosen time interval, it can be the number of weeks, etc.\nComment 16:\n“also accepts a groups argument which can be used to obtain stratified incidence“ -Consider changing stratified incidence by “epidemiological curve\".\nComment 17:\n“The basic elements of the incidence object can be obtained by the accessors get_counts(), get_dates(), and get_interval().\" - Please, number the number of basic elements for clarity purpose.\nComment 18:\n“The function subset() can be used for isolating case data from a specific time window and/or groups, while the [ operator can be used for a finer control to subset dates and groups using integer, logical or character vectors.” - If several functions are to be presented, it is easier to use bullet points to structure the reading. Consider the removal 'for isolating case data from a specific' and changing with 'to define'. “[“ change to 'indexing operator, to follow the classical denomination https://cran.r-project.org/doc/manuals/r-release/R-lang.pdf.\nComment 19:\n“Figure 1. Generalized workflow from incidence object construction to modeling and visualization.” - The first example in the paper is using a line-list format as data inputs but showed a stratified graphic. To be consistent and easier to follow for the reader, this figure illustrates the flow of such of data type, knowing that line-list is the primary source of epidemiological surveillance. A solution present in the figure is the two types of data inputs (non-aggregated/aggregated count). It would be easier, from a reader point of view, to capture the data flow from the line-lit to the final products proposed by this package.\nComment 20:\n“The function pool() can be used to merge several groups into one,” - Consider ending the sentence and explaining what it does. “and the function cumulate() will turn incidence data into cumulative incidence” - Consider changing “cumulative incidence” to “cumulative count of cases”.\nComment 21:\n[…]: an option for a ‘long’ format which is readily compatible with ggplot2 (Wickham, 2016) for further customization of graphics.” - Would it be possible to mention how date format is exported? This might good to elaborate a bit more about date and user's customization with ggplot2. This can be addressed later in the manuscript (see proposal for a theme).\nComment 22:\n“In line with RECON’s development guidelines, the incidence package is thoroughly” - Add a hyperlink/ref. to the website: https://www.repidemicsconsortium.org/resources/guidelines/\nSection: Modelling utilities\nComment 23:\n“Here, we highlight three simple functionalities in incidence for estimating parameters via modelling or bootstrap and the two specialized data classes that are used to store the models and parameter estimates.” - Consider structuring the following section according to the five elements mentioned (=three functions [fit() , estimate_peak() , fit_optim_split()] and two specialized data classes) using for instance bullet points/subtitles. For each function, the goal, data input, statistical methods and output object(s) can be grouped in single section.\nComment 24:\n“we implement the simple log-linear regression approach in the function fit(), which can […]\" - Please add more information about the structure of the 'incidence_fit objects containing log-linear models'\nComment 25:\n“[…] fit exponential increase or decrease of incidence over time by log-transforming case counts … “ - Can be simplified to “fit exponential increase or decrease using a linear regression over time on log-transformed case counts…“.\nComment 26:\n“where y is the incidence, r is the growth rate, t\" - Replace “incidence” with number of new case/incident case.\nComment 27:\n“serial interval” - Consider adding “serial interval of the infectious agent”\nComment 28:\n“uses multinomial bootstrapping to estimate the peak, assuming” - Some explanation about the method and references would be desirable.\nComment 29:\n“Both object classes can be plotted separately or added to an existing epicurve using the function add_incidence_fit() (Figure 5).\" - The customization of the epicurve is well described. However, it is not mentioned how to change the layout of the model outcome and confidence interval. Indeed, some users might wish to use an alternative ggplot2 geometric object such as geom_range with a shaded grey semi-transparent band instead of two dotted lines. It would an added-value to provide some capacities or explanation and an example of customization of the layout of the “incidence_fit objects”.\nSection: Use cases\nComment 30:\n“Two worked examples are used to demonstrate the functionality and flexibility of the incidence package. The first example illustrates how to compute and manipulate stratified weekly incidence directly from a line-list”. - Consider “The first example illustrates how to create directly from a line-list incidence object in order draw an epicurve of the weekly number of cases with or without stratification on patient characteristics”.\nComment 31:\n“while the second example shows how to import pre-computed daily incidence and fit a log-linear model to estimate growth rate (r) and doubling time for the growing phase.” - Footnote to be included in the section about the function for more clarity.\n\nExample 1: computing and manipulating stratified weekly incidence\nComment 32:\n“The weekly incidence stratified by hospitals is computed by running the function incidence() on the Date variable dat1$date_of_onset with the arguments interval = 7 and groups = dat1$hospital.\" - Consider rephrasing. For instance: “the epicurve with the weekly number of cases by hospital can be computed from the line listing dataframe object (dat1) using the function incidence() on i) the date variable (dat1$date_of_onset), ii) by specifying the argument interval of seven days in order to aggregate the number case per week (interval = 7) and iii) including a the line-listing variable for stratification in the argument groups, in this case the hospital name (groups = dat1$hospital).\nComment 33:\n“Here $count is a matrix with 56 rows and 6 columns as groups by hospital” - Missing s at the end of $counts.\nComment 34:\n“The generic plot() method for incidence objects calls the specific function plot.incidence(), which makes an incidence barplot using the ggplot2 package. Hence, customization of incidence plot can benefit from the powerful graphical language from ggplot2.” - A short explanation and command line to explain how to access the code of the method would be welcome (notably the plot). This would help users to understand which ggplot2 geometric object(s) is used for the bar plot and incidence_fit lines. This would be an asset to understand how to proceed with further customization (within the aesthetic, theme or faceting specifications). This can be proposed at the end through several examples.\nComment 35:\n# plot incidence object my_theme <-theme_bw(base_size = 12) + theme(panel.grid.minor = element_blank()) + theme(axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5, color = \"black\")) plot(i.7.group, border = \"white\") + my_theme + theme(legend.position = c(0.8, 0.75))\nThe current example assumes that all users are familiar with ggplot2, notably how to customize the non-data components through the theme. We might suggest to introduce what is a theme in ggplot2 and what it does, for instance “Themes allows modification (content and layouts) of non-data components such as titles, axis labels, legends (position and aspect …), graphics grid lines and backgrounds (Modify components of a theme; ref: https://ggplot2.tidyverse.org/reference/theme.html)”. In addition, mention that theme specifications can “overwrite” some layout specification in the other part of the ggplot2 function. The use in this example of theme_bw appears to clarify what the default built-in theme is (complete themes: https://ggplot2.tidyverse.org/reference/ggtheme.html , other themes https://cran.r-project.org/web/packages/ggthemes/ggthemes.pdf). To help the reader, you might consider the arguments as displayed in “Modify components of a theme” order as much as possible:\nmy_theme <-theme_bw(base_size = 12) + theme ( panel.grid.minor = element_blank, axis.text.x = element_text(angle = 90, hjust = 1, vjust = 0.5, color = \"black\"), legend.position = c(0.8, 0.75) )\nFor this first example, consider setting the legend outside of the main graphic frame, especially for the first figure (instance legend.position=\"bottom\", legend.box = \"horizontal\"). For the axis labels, the Y is family the “Num. of case of EVD. per week” in the Figure 2.\nComment 36:\ni.7.sub <- i.7.group[1:18, grep(\"Hospital\", group_names(1.7.group))] hosp_colors <-c(\"#899DA4\", \"#C93312\", \"#FAEFD1\", \"#DC863B\") - Add an intermediate line to illustrate that in this period selection (e.g. print i.7), cases are coming from four hospitals, to hosp_colors would a four colours vector. Keep the legend out of the main graphic frame (top or bottom).\nComment 37:\n“Here, because of the few numbers of cases in the first few weeks, we have also highlighted each case using show_ cases = TRUE (Figure 3)”. - The figure is displaying the individual case which is the standard representation of an epicurve according to number of training programs (US CDC, EPIET …). This is of primary importance in outbreak investigation to able to delineate case number in an easy-to-read visual and often stratified representation by case characteristics (sex, exposure, location …). This type of representation is adopted by default in package EpiCurve (see https://cran.r-project.org/web/packages/EpiCurve/vignettes/EpiCurve.pdf) as well as US CDC training example of epicurve (https://www.cdc.gov/training/quicklearns/createepi/). Representation of case as square can be achieved with the current package as illustrated in at the end of the package vignette (see: vignette (\"customize_plot\", package=\"incidence\") in the section Applying the style of European Programme for Intervention Epidemiology Training (EPIET). It’s understandable that square form can be changed in case of a large dataset making the size of the intervals for the x- and y- axes different. It would be an asset for the target audience and the current manuscript to contain a full example of an epicure with case made under square representation (somehow close to the example cited above in the vignette) combined with a full theme following simple and standard representation.\nComment 38\nConsider change in the layout of Figure 2: 'Number of EVD cases stratified by hospital ... between week XX and week YY'. Y axis title: 'Number of cases' X axis title: 'Week of onset of illness' X axis ticks should be made visible for all weeks even the label displayed for any other week.\nComment 39:\nWould it be possible to give some more information to which ggplot2 geometry are used when show_ cases = TRUE is specified? This is of importance to provide the reader with a clue about how both borderline and colour content of each square can be further customized.\nComment 40:\nAs standard the label of the X axis time interval is displayed on the left side of the bin. Formally, the ideal position would be right below the bin (as illustrated in the CDC example above and numerous of published epicurve). In the EpiCurve package, the standard representation is not ideal as the X axis tick is right in the middle of the bin. The position of the label can be customized by the user, but would it be possible to look at the option to place the time label under each bin instead of to the left bin tick mark (see an example of ad hoc customization below). Of note, such label position should support figure export resizing.\nComment 41:\nFigure 4. (A) stratified and (B) pooled daily incidence plots of ZVD in Colombia, September 2015 to January 2016. Please consider comments made above for the other figure on figure title and axis labels:\n“Epicurve of the daily number of Zika virus disease between Sep and Jan 2016”. Panel (A) stratified by location and (B) pooled across locations. “ Y axis title: 'Number of cases' X axis title: ' Week of onset of illness'\nSome readers might wonder if the faceting capabilities of ggplot2 would be supported or not. An evident alternative is to prepare separated epicurve for the both locations and further combined them using specific package (ggarrange for instance). Would it be possible to add information on this point, and if supported, provide an alternative presentation of the both epicurves using two vertically aligned panels?\nComment 42:\n“Without any knowledge on the splitting date, we can turn to the fit_optim_split() function to look for the optimal splitting date”. - Could you please reconsider the phrasing as “Without any knowledge on the splitting date” which seems not logical when looking retrospectively to an outbreak epicurve and visually identifying the peak of the epidemic wave.\nComment 43:\nlibrary('magrittr') - Provide a short explanation about the dependency(ies) with this package.\nComment 44:\n“The predictions and their 95% CIs from the two incidence_ft objects, ‘before’ and ‘after’, can be added to the existing incidence plot of i.pooled using the piping-friendly function add_incidence_fit().” - Provide more explanation on “the piping-friendly function” aspect. Provide an example of layout customization of the two log-linear regression models (linetype, colour, size).\nComment 45:\nPlease find below a proposal for an advanced and custom-made layout epicurve representation using the Zika dataset. In this example, some of the important layout feature of an epicurve are available: i) no space around the limits of both x and Y axes, ii) representation of each case with a square, iii) x label under each bin and make the graphic as light as possible based on Tufte’s advice (less grid as possible, visible labels …). In the graphic below, an incidence_fit object can be added for the pooled location with a specific customization to give a complete overview to the package capabilities.\n\nWorking example based on Zika diease dataset:\n\nhead(zika_girardot_2015, 20) head(zika_sanandres_2015, 20) dat2 <- merge(zika_girardot_2015, zika_sanandres_2015, by = \"date\", all = TRUE) # # combine two datasets into one names(dat2)[2:3] <- c(\"Girardot\", \"San Andres\") # rename variables dat2[is.na(dat2)] <- 0 # replace NA with 0 i.group <- as.incidence(x = dat2[, 2:3], dates = dat2$date) # convert pre-computed incidence in data.frame into str(i.group)\n\n# Graphic\n\ni.group_zoom_in <- i.group[10:60]\n\nplot(i.group_zoom_in, n_breaks = nrow(i.group_zoom_in), border = \"grey90\", show_cases=TRUE) +\n\ntheme_bw() +\n\nscale_y_continuous(expand = c(0, 0), limits=c(0,max(i.group_zoom_in$counts +1))) +\n\nscale_x_date(date_breaks = \"1 day\", date_labels = \"%b %d\",expand = c(0, 0), limits= c(as.Date(min(i.group_zoom_in$dates)) , as.Date(max(i.group_zoom_in$dates+1)))) + # scale_fill_discrete(name = \"Location:\") +\n\n#scale_x_continuous(expand = c(0, 0),limits= c(as.Date(min(zomm_in$dates)) , as.Date(max(zomm_in$dates)))) +\n\nlabs(title = \"Number of Zika disease cases\",\n\nsubtitle = \"Girardot and San Andres municipalities, Colombia. Period: 6 Sep to 11 Oct 2015.\" ,\n\nx=\"Week of onset of illness\",\n\ny=\"Number of cases\",\n\nfill=\"Municipalities:\") +\n\ntheme(panel.border = element_rect(colour = \"white\"),\n\npanel.grid.major.y = element_line(colour = \"grey70\", linetype=\"dotted\", size =0.5),\n\naxis.line = element_line(colour = \"black\", size = 0.7),\n\naxis.text.x = element_text(angle = 45, hjust = 0.8, size = 6),\n\naxis.ticks.x = element_line(size = rel(2)),\n\naxis.ticks.y = element_line(size = rel(2)),\n\naxis.title.y = element_text(margin=margin(0,10,0,0), size=10),\n\naxis.title.x = element_text(margin=margin(10,0,0,0), size=10),\n\nplot.title = element_text(size = 12,face = \"bold\"),\n\nplot.subtitle = element_text(size = 9),\n\nlegend.position=\"bottom\",\n\nlegend.box = \"horizontal\") +\n\ncoord_equal()\nLink to plot available here.\n\nIs the rationale for developing the new software tool clearly explained? Yes. The rational is sounds and the new package “Incidence” is fulfilling the objectives cited. It allows streamlining the manipulation and representation for epidemiological data for epidemiological representation. It's expected that this new package would reach a wide audience of epidemiologists and epidemiological data analysts working with R.\n\nIs the description of the software tool technically sound? Yes. However, some clarifications proposed in the detailed review would enhance the description of the software tool features.\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? Overall, yes. Minor adjustments are proposed to allow reader to better access code (e.g. provide example to access full code of incidence (S3) class incidence and its subsequent methods (plot, …)) and offer additional clarification to improve graphic customization (notably around the incidence_fit object and further integration of ggplot2 faceting functionality).\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool? Yes. Several practical and realistic examples are provided by the authors. All examples were tested and further tests with epidemiological additional datasets were conducted and allow to reproduce accurately tool behaviours. In addition to the paper review, the R package documentation (description file, recent reference manual and vignettes) were thoroughly reviewed to assess any discrepancies between the peer-review publication and package documentation on CRAN.\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes.\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/8-139
|
https://f1000research.com/articles/7-879/v1
|
22 Jun 18
|
{
"type": "Research Article",
"title": "Streptococcus pneumoniae serotype epidemiology among PCV-10 vaccinated and unvaccinated children at Gertrude’s Children’s Hospital, Nairobi County: a cross-sectional study",
"authors": [
"Michael Walekhwa",
"Margaret Muturi",
"Revathi Gunturu",
"Eucharia Kenya",
"Beatrice Kabera",
"Michael Walekhwa"
],
"abstract": "Background: Streptococcus pneumoniae (SPn) serotype replacement and emergence of multidrug resistant SPn has exacerbated the need for continuous regional serotype surveillance. We investigated SPn serotypes circulating among children ≤5 years in Nairobi County. Methods: Streptococcus pneumoniae stocks stored at −70°C in brain heart infusion medium were thawed at room temperature for 30 minutes. In total, 10 µl of the stored SPn cells were suspended in 50 µl PBS and gently vortexed. About 10 µl of the suspended cells were added on to a glass slide and mixed with 10 µl pooled antisera. The glass slide was swirled gently while observing for any reaction. The process was repeated with individual groups under various antisera pools. Those serotypes that did not belong to any pool were typed directly until a positive agglutination reaction was observed. The cells/PBS/serotype-specific antisera mixture on the glass slide were covered with a coverslip and observed under a phase contrast microscope at ×100 objective lens with oil emulsion. Results: Out of the 206 subjects sampled, 20.39% (n=42) were found to be carriers of SPn. About 52% (n=22) of the SPn carriers had received the recommended dose of PCV-10, while 48% (n=20) of the carriers had not. Almost all (n=41; 19.90% of subjects) isolates contained non-vaccine type SPn serotypes, while n=1 of the serotypes (in 0.49% of subjects) were untypeable. Serotypes 28F, 6A, 11A, 3 and 7C were prevalent in both vaccinated and unvaccinated children, whereas serotypes 23A, 17F, 35F, 48, 13 and 35B, and 23B, 20, 19B, 21, untypeable, 15B and 39 were found among unvaccinated and vaccinated groups, respectively. Conclusions: All SPn serotypes isolated from the subjects sampled were non PCV-10 vaccine type. Therefore Kenyan children receiving PCV-10 vaccine are not protected.",
"keywords": [
"Streptococcus pneumoniae",
"serotypes",
"Nairobi",
"Quellung reaction",
"Optochin test",
"Bile solubility"
],
"content": "Introduction\n\nStreptococcus pneumoniae (SPn) is a highly invasive, Gram-positive, extracellular bacterial pathogen (Mitchell & Mitchell, 2010). It is a major cause of morbidity and mortality globally, causing more deaths than any other infectious disease (Jones et al., 2010). SPn is classified into serogroups (denoted by numbers and letters, e.g. 18c, 23f) (Kellogg et al., 2001). There are over 90 known serotypes and the prevalence of different serotypes varies by regions of the world (Hackel et al., 2013). Different serotypes exhibit differing potentials to cause disease and may cause different syndromes in different age groups (Harboe et al., 2009).\n\nSome strains also have a greater potential to develop antibiotic resistance (Song et al., 2012). The 13 most common serotypes of SPn pneumonia cause 80–93% of serious pneumococcal disease in children (Johnson et al., 2010). According to the World Health Organization (WHO) and UNICEF Global Action Plan for the Prevention and Control of Pneumonia, pneumonia kills more children than any other illness in the world (WHO & UNICEF, 2009). Given the high burden of under-five mortality associated with pneumonia, control efforts are critical to achieving Sustainable Development Goal 3 (Colglazier, 2015).\n\nWHO and UNICEF estimates indicate that over 800,000 children under 5 years of age die from pneumococcal disease each year (O’Brien et al., 2009). In Kenya, an estimated one in every five children under 5 years of age dies from this disease every year (WHO, 2013).\n\nSPn vaccines protect against several severe forms of pneumococcal disease, such as meningitis, pneumonia and bacteremia (Feldman & Anderson, 2014). These vaccines will not protect against these conditions if they are caused by agents other than SPn or from other strains of SPn that are not contained in the vaccine (Moffitt & Malley, 2011). The 10-valent pneumococcal conjugate vaccine (PCV10) was introduced into the Kenya Expanded Program on Immunization (KEPI) in February 2011 with a 2+1 schedule (at 6, 10, 14 weeks) without catch-up vaccinations (Hammitt et al., 2014). The vaccine covers 1, 4, 5, 6b, 7f, 9V, 14, 18c, 19f and 23f SPn serotypes.\n\nCurrently over 90 different serotypes have been identified, six of them very recently (Weinberger et al., 2011). Various SPn serotypes with antigenic similarities are classified under the same groups (9A, 9L, 9N and 9V) while those lacking antigenic similarities are given numbers only (1, 2, 3, 4 and 5). The degree of interaction (cross-reactivity) between various SPn groups may vary. For instance, serotypes 6A and 6B have identical chemical composition except for one of the bonds between two sugars yet they are highly cross-reactive but serotypes 19F and 19A are less reactive.\n\nPneumococcal conjugate (PCVs) and polysaccharide (PPVs) vaccines are designed according their virulence mechanisms and how they generally interact with the human immune system (Castañeda-Orjuela et al., 2012). The WHO has advised that all children ≤5 years should be immunized against pneumococcal disease and continuous surveillance done to keep out the disease especially in the developing world (Vandenbos et al., 2013). The need for continuous surveillance has been exacerbated by the acute emergence of multi-drug resistant SPn strains and escalated child mortality and morbidity due to pneumococcal disease, despite the availability of PCVs and PPVs. (Väkeväinen et al., 2010). This study therefore sought to establish the SPn serotypes among vaccinated and unvaccinated children ≤5 years of age in Nairobi County, Kenya.\n\n\nMethods\n\nThis study was conducted among children ≤5 years attending the outpatient department of Gertrude's Children's Hospital, Muthaiga, and its satellite clinic in Githongoro, Nairobi County between May 2017 and February 2018. Subjects were clinically assessed by a physician and those who presented with pneumococcal disease symptoms recommended to the study nurse for recruitment. Gertrude's Children’s Hospital is the largest standalone health care facility specializing in pediatric care in East and Central Africa. The hospital is accredited by the Joint Commission on International Accreditation (JCIA). SPn isolation and stocking was done at Gertrude's Children’s Hospital Main Laboratory and capsular serotyping done at KEMRI Wellcome Trust, Kilifi, Kenya.\n\nThis was a descriptive cross-sectional study. Streptococcus pneumoniae serotype epidemiology among PCV-10 vaccinated and unvaccinated children between 6 months and 5 years of age was measured. Children who had no history of any chronic disease and whose parents or legal guardians consented to the study were systematically recruited. Children whose parents or legal guardians declined to give consent and those with any known immunosuppressive conditions were excluded from the study.\n\nTo determine the minimum sample size, the formula developed by Chow et al. (2007) was used, with a prevalence rate of 16% (Agweyu et al., 2014).\n\n\n\nWhere n= desired minimal sample size; z= standard normal deviation (1.96, from the tailed normal table); p̂= prevalence rate; and m= the desired degree of accuracy at a 95% confidence level of 0.05. This gave a sample size of 206.\n\nNasopharyngeal swabs were per nasally collected using Copan flocked swabs and temporarily suspended in Armies medium for transportation to the main laboratory. Each swab was inoculated onto a selective gentamicin with 5% sheep blood agar (BA) plate. All swabs were plated within 24 h of collection. The plates were incubated at 37°C in a 5% CO2 atmosphere and examined at 16–24 h and then again at 40–48 h for growth of SPn. Isolates were identified as SPn by colony morphology (Mucoid, draughtsman appearance, α-haemolysis) and susceptibility to optochin (positive, ≥14 mm zone of inhibition; negative, <14 mm zone of inhibition). Plates with colonies akin to SPn morphological features but with optochin clearance zones below 14 mm were further subjected to solubility in bile salts (positive, bile soluble; negative, bile insoluble).\n\nThe isolation of a single colony indicated carriage. Single colonies were picked using sterile inoculating loops and evenly plated on BA. After 24–48 h, enough inoculum was stocked in brain heart infusion (BHI) agar with 5% sheep blood (Ultralab East Africa, Ltd), gently vortexed and stored at –70°C for serotyping.\n\nCapsular serotyping was done using the Quellung reaction test. Frozen vials containing SPn stocks stored at -70°C were thawed at room temperature for about 30 min. Next, 10 µl of the stored SPn cells were suspended in 50 µl PBS and gently vortexed. Subsequently, 10 µl of the suspended cells were added on to a glass slide and mixed with 10 µl pooled antisera (Statens Serum Institute, cat. No. 16744). The glass slide was swirled gently while observing for any agglutination reaction until a positive reaction was observed with various pooled antisera. The process was repeated with individual groups under various antisera pools.\n\nAfter that, 10 µl of the suspended cells in PBS were added to a glass slide and mixed with various SPn serotype-specific antisera included in the antisera pools that gave a positive reaction. This was done until a positive reaction with the particular serotype specific antisera was observed. Those serotypes that did not belong to any pool were typed directly until a positive agglutination reaction was observed. The cells/PBS/serotype-specific antisera mixture on the glass slide were covered with a cover slip and observed under a phase contrast microscope with a ×100 objective lens with oil emulsion.\n\n\nResults\n\nOut of n=206 (100%) of the subjects sampled, n=97 (47.1%) were male and n=109 (52.9%) were female. In total, 68 (33.0%) of the children studied were within the age bracket of 6–12 months, 47 (22.8%) were between the ages of 13–24 months, 46 (22.3%) were between the ages of 25–36 months, 17 (8.3%) were between the ages of 37 and 48 months and 28 (13.6%) were between the ages of 49 and 60 months. Out of the total number of subjects (n=206) sampled, 20.39% (n=42) were found to be carriers of SPn; 52% (n=22) of the SPn carriers had received the recommended dose of PCV-10 immunization, while 48% (n=20) had not. All isolates (n=42; 20% of subjects) contained non-vaccine-type SPn serotypes, while n=1 (0.49% of the subjects) of the serotypes were untypeable (Table 1). In total, 18 different SPn serotypes were found in this population. They include: 28F (8 instances), 6A (5 instances), 3 (4 instances), 23B (3 instances), 20 (3 instances), 23A (3 instances), 19B (2 instances), 17F (2 instances), 7C (2 instances), 11A (2 instances), 35F (1 instance), 15B (1 instance), untypeable (1 instance), 48 (1 instance), 35B (1 instance), 21 (1 instance), 39 (1 instance) and 13 (1 instance).\n\nThe percentage of SPn carriage status among PCV-10 vaccinated and unvaccinated children is shown.\n\nVarious serotypes were found to be prevalent in different age groups. For instance, out of the 42 serotypes found, 9 (23.53%) were prevalent among children at 6–12 months of age (n=16). They include: 28F (4 instances), 11A (2 instances), 23A (2 instances), 3 (2 instances), 6A (2 instances), 17F (1 instance), 35F (1 instance), 7C (1 instance) and untypeable (1 instance). There were 7 (16.67%) serotypes prevalent among children at 13–24 months (n=8), including: 20 (2 instances), 21 (1 instance), 39 (1 instance), 28F (1 instance), 35B (1 instance), 17F (1 instance) and 13 (1 instance). There were 8 (19%) serotypes found among children of 25–36 months of age (n=12), including: 23B (3 instances), 19B (2 instances), 3 (2 instances), 20 (1 instance), 28F (1 instance), 7C (1 instance), 23A (1 instance) and 48 (1 instance). There were 3 (7%) serotypes prevalent among children at 37–48 months old (n=4), including: 6A (2 instances), 15B (1 instance) and 28F (1 instance).\n\nThere were 2 (4.76% of the total) serotypes prevalent among children at 49–60 months (n=2): 6A (1 instance) and 28F (1 instance) (Table 2). Out of the 42 isolates (found in 20.39% of subjects), serotype 28F was the most prevalent (3.88% of the total), followed by 6A (2.43%), 3 (1.94%) and 20, 23A and 23B all at 1.46% (n=3). Each of the serotypes 7C, 11A, 17F and 19B represented 0.97% (n=2) of the total serotypes, while serotypes: 13, 21, 39, untypeable, 48, 15B, 35B and 35F represented 0.49% (n=1) each of the total serotypes found (Figure 1 and Figure 2). In total 51% (n=106) of the total sampled subjects were confirmed to have received a full dose of the PCV-10 vaccination as per the recommended schedule of immunization at 6, 10 and 14 weeks. Approximately 11% (n=12) of the immunized children were carriers of SPn in their nasopharyngeal region; 10% (n=10) of the non-immunized group were also carriers (Table 3). Serotypes 28F (5 instances), 23A (3 instances), 6A (3 instances), 17F (2 instances), 11A (1 instance), 3 (1 instance), 35F (1 instance), 48 (1 instance), 13 (1 instance), 35 (1 instance) and 7C (1 instance) were prevalent among the 9.71% (n=20) of the total sample group that had not received PCV-10 immunization. Serotypes 3 (3 instances), 28F (3 instances), 23B (3 instances), 20 (3 instances), 19B (2 instances), 6A (2 instances), 21 (1 instance), 11A (1 instance), 7C (1 instance), untypeable (1 instance), 15B (1 instance) and 39 (1 instance) were prevalent among the 10.68% (n=22) of the total sample group that received immunization (Table 4).\n\nThe SPn serotypes as found among PCV-10 vaccinated and unvaccinated children of varying age groups is shown.\n\nThis figure shows the prevalence of various SPn serotypes among PCV-10-vaccinated and –unvaccinated children.\n\nThe percentage prevalence of SPn serotypes among PCV-10 vaccinated and unvaccinated children is shown.\n\nNGR, no growth observed; SPn, Streptococcus pneumoniae; NA, not applicable.\n\n\nDiscussion\n\nThis study found that 20.39% of all children studied, from both the PCV-10 vaccinated and unvaccinated groups, were carriers of SPn. While this is a significant reduction from the pre-vaccine era, it is still high compared to malaria, diarrhea and HIV/AIDS (Feikin et al., 2010). In total, n=41 of the serotypes found were non-vaccine type (in 19.90% of the subjects), with one additional untypeable serotype. This is a very important finding as it explains the high level of child morbidity and mortality due to pneumococcal disease despite the availability of PCV-10.\n\nWhile these findings agree partially with those of Jacobs et al. (2008), where a significant decrease in the vaccine type SPn serotypes found in isolates was observed, a 97.6% (n=41) decrease is, at the very least, surprising. This trend may be attributed to the increased level of antimicrobial misuse by a greater percentage of the study population (Domenech de Cellès et al., 2011). 10-valent pneumococcal conjugate vaccine contains 10 different serotypes, which include: 1, 4, 5, 6B, 7F, 9V, 14, 18C, 19F, 23F (Slotved et al., 2016). None of these 10 serotypes was found in the study population yet this is the vaccine currently included in KEPI, targeting the same population.\n\nStreptococcus pneumoniae carriage decreased with age as 11.65% (n=24) were obtained from children aged between 6–24 months and 8.74% (n=18) from children >24 months. The study results demonstrated a linear relationship between child age and SPn carriage. A similar study done elsewhere reported findings that partly agree with this and partly disagrees (Hill et al., 2008). The former being attributable to development of SPn-specific IgG antibodies due to vaccination and during that window before most children start attending school (Corscadden et al., 2013). Unlike findings from a study by de Paz et al. (2015), serotype 28F was the most prevalent and was present in all five age groups profiled. This is a likely scenario of serotype replacement as SPn attempts to evade the action of the immune system and eventually shares the resistant genes within the microbial community, especially in the nasopharyngeal region (Donati et al., 2010).\n\nSerotypes 28F, 6A, 11A, 3 and 7C were prevalent in both vaccinated and unvaccinated children, whereas serotypes 23A, 17F, 35F, 48, 13, 35B and 23B, 20, 19B, 21, untypeable, 15B, 39 were found among unvaccinated and vaccinated groups respectively. There exists different antigenic features between and within various strains of SPn (Song et al., 2012). While the majority, if not all, pneumococcal serotypes are capable of causing disease, the frequency with which they are isolated varies (Kalin, 1998). In this case, vaccination would only be partially effective and, if so, due to inter-strain antigenic characteristics.\n\nWhile trying to evade the action of the immune system, SPn has a tendency to exchange resistant genes and other antigenic correlates at the nasopharyngeal region (Johnston et al., 2014). Resistance to antimicrobial agents is occasioned by among other factors, misuse of antibiotics (Dinsbach, 2012). This is largely due to a lack of properly enforced antibiotic use regulations by the authorities.\n\n\nData availability\n\nDataset 1. List of basic demographic information for each subject, with the size of the optochin clearance zone and serotype of Streptococcus pneumoniae, if found. DOI: http://doi.org/10.5256/f1000research.14387.d207458 (Walekhwa et al., 2018).",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThe study was in part funded by the National Commission of Science Technology & Innovation (NACOSTI) Kenya. Ref Nos. NACOSTI/RCD/ST & I/7TH CALL/PhD/148.\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\nSpecial thanks to Prof. Anthony Scott, Dr. Claire Gordon and Ms. Angela Karani of Kemri Wellcome Trust, Kilifi for their professional input. The laboratory and technical team: Ann Karanu, Elizabeth Mbithe, Shadrack Mutua and Alfred Too. Your technical input was amazing! The Gertrude's Hospital clinical and laboratory team (Linet Okose, Barasa Kasili, Charles Muia, Lilive Njagi, Carolyne Thumbi, and Peninah Chege), you made this possible. Thanks to Japheth Wambani Rapando for the review of my work.\n\n\nReferences\n\nAgweyu A, Kibore M, Digolo L, et al.: Prevalence and correlates of treatment failure among Kenyan children hospitalised with severe community-acquired pneumonia: a prospective study of the clinical effectiveness of WHO pneumonia case management guidelines. Trop Med Int Health. 2014; 19(11): 1310–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCastañeda-Orjuela C, Alvis-Guzmán N, Velandia-González M, et al.: Cost-effectiveness of pneumococcal conjugate vaccines of 7, 10, and 13 valences in Colombian children. Vaccine. 2012; 30(11): 1936–1943. PubMed Abstract | Publisher Full Text\n\nChow S, Wang H, Shao J: Sample Size Calculations in Clinical Research. New York. 2nd ed. Taylor & Francis Group. 2007. Reference Source\n\nColglazier W: SUSTAINABILITY. Sustainable development agenda: 2030. Science. 2015; 349(6252): 1048–50. PubMed Abstract | Publisher Full Text\n\nCorscadden KJ, Kirkham LA, Thornton RB, et al.: High pneumococcal serotype specific IgG, IgG1 and IgG2 levels in serum and the middle ear of children with recurrent acute otitis media receiving ventilation tubes. Vaccine. 2013; 31(10): 1393–1399. PubMed Abstract | Publisher Full Text\n\nde Paz HD, Selva L, Muñoz-Almagro C: Pneumococcal Vaccination and Consequences. Streptococcus Pneumoniae. 2015; 41–57. Publisher Full Text\n\nDomenech de Cellès M, Opatowski L, Salomon J, et al.: Intrinsic epidemicity of Streptococcus pneumoniae depends on strain serotype and antibiotic susceptibility pattern. Antimicrob Agents Chemother. 2011; 55(11): 5255–5261. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDinsbach NA: Antibiotics in dentistry: Bacteremia, antibiotic prophylaxis, and antibiotic misuse. Gen Dent. 2012; 60(3): 200–207; quiz 208–9. PubMed Abstract\n\nDonati C, Hiller NL, Tettelin H, et al.: Structure and dynamics of the pan-genome of Streptococcus pneumoniae and closely related species. Genome Biol. 2010; 11(10): R107. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFeikin DR, Jagero G, Aura B, et al.: High rate of pneumococcal bacteremia in a prospective cohort of older children and adults in an area of high HIV prevalence in rural western Kenya. BMC Infect Dis. 2010; 10: 186. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFeldman C, Anderson R: Review: Current and new generation pneumococcal vaccines. J Infect. 2014; 69(4): 309–25. PubMed Abstract | Publisher Full Text\n\nHackel M, Lascols C, Bouchillon S, et al.: Serotype prevalence and antibiotic resistance in Streptococcus pneumoniae clinical isolates among global populations. Vaccine. 2013; 31(42): 4881–4887. PubMed Abstract | Publisher Full Text\n\nHammitt LL, Akech DO, Morpeth SC, et al.: Population effect of 10-valent pneumococcal conjugate vaccine on nasopharyngeal carriage of Streptococcus pneumoniae and non-typeable Haemophilus influenzae in Kilifi, Kenya: Findings from cross-sectional carriage studies. Lancet Glob Health. 2014; 2(7): e397–e405. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarboe ZB, Thomsen RW, Riis A, et al.: Pneumococcal serotypes and mortality following invasive pneumococcal disease: A population-based cohort study. PLoS Med. 2009; 6(5): e1000081. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHill PC, Cheung YB, Akisanya A, et al.: Nasopharyngeal Carriage of Streptococcus pneumoniae in Gambian Infants: A Longitudinal Study. Clin Infect Dis. 2008; 46(6): 807–814. PubMed Abstract | Publisher Full Text\n\nJacobs MR, Good CE, Beall B, et al.: Changes in serotypes and antimicrobial susceptibility of invasive Streptococcus pneumoniae strains in Cleveland: a quarter century of experience. J Clin Microbiol. 2008; 46(3): 982–990. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJohnson HL, Deloria-Knoll M, Levine OS, et al.: Systematic evaluation of serotypes causing invasive pneumococcal disease among children under five: The pneumococcal global serotype project. PLoS Med. 2010; 7(10): pii: e1000348. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJohnston C, Campo N, Bergé MJ, et al.: Streptococcus pneumoniae, le transformiste. Trends Microbiol. 2014; 22(3): 113–9. PubMed Abstract | Publisher Full Text\n\nJones RN, Jacobs MR, Sader HS: Evolving trends in Streptococcus pneumoniae resistance: Implications for therapy of community-acquired bacterial pneumonia. Int J Antimicrob Agents. 2010; 36(3): 197–204. PubMed Abstract | Publisher Full Text\n\nKalin M: Pneumococcal serotypes and their clinical relevance. Thorax. 1998; 53(3): 159–162. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKellogg JA, Bankert DA, Elder CJ, et al.: Identification of Streptococcus pneumoniae revisited. J Clin Microbiol. 2001; 39(9): 3373–3375. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMitchell AM, Mitchell TJ: Streptococcus pneumoniae: virulence factors and variation. Clin Microbiol Infect. 2010; 16(5): 411–8. PubMed Abstract | Publisher Full Text\n\nMoffitt KL, Malley R: Next generation pneumococcal vaccines. Curr Opin Immunol. 2011; 23(3): 407–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nO’Brien KL, Wolfson LJ, Watt JP, et al.: Burden of disease caused by Streptococcus pneumoniae in children younger than 5 years: global estimates. Lancet. 2009; 374(9693): 893–902. PubMed Abstract | Publisher Full Text\n\nSlotved HC, Dalby T, Hoffmann S: The effect of pneumococcal conjugate vaccines on the incidence of invasive pneumococcal disease caused by ten non-vaccine serotypes in Denmark. Vaccine. 2016; 34(6): 769–774. PubMed Abstract | Publisher Full Text\n\nSong JH, Dagan R, Klugman KP, et al.: The relationship between pneumococcal serotypes and antibiotic resistance. Vaccine. 2012; 30(17): 2728–37. PubMed Abstract | Publisher Full Text\n\nVäkeväinen M, Soininen A, Lucero M, et al.: Serotype-specific hyporesponsiveness to pneumococcal conjugate vaccine in infants carrying pneumococcus at the time of vaccination. J Pediatr. 2010; 157(5): 778–83.e1.PubMed Abstract | Publisher Full Text\n\nVandenbos F, Gal J, Radicchi B: [Vaccination coverage against influenza and pneumococcus for patients admitted to a pulmonary care service]. Rev Mal Respir. 2013; 30(9): 746–751. PubMed Abstract | Publisher Full Text\n\nWalekhwa M, Muturi M, Gunturu R, et al.: Dataset 1 in: Streptococcus pneumoniae serotype epidemiology among PCV-10 vaccinated and unvaccinated children at Gertrude’s Children’s Hospital, Nairobi County: a cross-sectional study. F1000Research. 2018. Data Source\n\nWeinberger DM, Harboe ZB, Flasche S, et al.: Prediction of serotypes causing invasive pneumococcal disease in unvaccinated and vaccinated populations. Epidemiology. 2011; 22(2): 199–207. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWHO: Pneumonia Factsheet. WHO Media Centre. 2013; 1–2.\n\nWHO & UNICEF: Global Action Plan for Prevention and Control of Pneumonia (GAPP) Technical Consensus statement. B World Health Organ. 2009; 86(5): 1–23. Reference Source"
}
|
[
{
"id": "39126",
"date": "25 Oct 2018",
"name": "Bartholomew N. Ondigo",
"expertise": [
"Reviewer Expertise Immunoparasitology"
],
"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\nTitle: To include Kenya\nCorresponding author contacts need to be indicated\n\nAbstract: Background: It needs to be indicated that circulating SPn serotypes are being investigated after the introduction of 10-valent pneumococcal conjugate vaccine (PCV10) in 2011.\n\nThe methods need to be a summary of the techniques used in data collection. As it is written it appears to be copy pasted from the method section. Indicate number of children assessed. For instance:\n\nMaterials: Two hundred and six children attending and not-attending in ……20-2009 were studied. Materials for study were pharyngeal swabs and sputum. Identification was performed using optochin disks, Quellung reaction,……agglutination on the glass, viewed under a phase contrast microscope and the sent to KEMRI- Kilifi for further confirmative identification tests.\n\nThe conclusion seem alarming and need to suggest or indicate. Therefore Kenyan children receiving PCV-10 vaccine are not protected – Revise to something like:- This study highlights the importance of monitoring and evaluation to provide epidemiological information to determine the effectiveness of PCV10 in Kenya’s Public health services.\n\nRepeating ideas should be deleted - causing more deaths than any other infectious disease vs. kills more children than any other illness in the world.\n\nIntroduction need to be shortened, preferably to three paragraphs.\n\nMethods: Were the children admitted or outpatient? All the source of equipment used and consumables need to be indicated, for instance incubator etc. The Research Ethics Committee that approved the study need to be indicated.\nSoftware used for calculation of %s need to be indicated?\n\nResults: Headings need to be introduced that are in agreement with the content. This will help the reader to when reading. Suggested possible headings include:\n\nDemography of the Study Participants Prevalence of SPn carriage status among PCV-10 vaccinated and unvaccinated children Prevalence of SPn carriage status by age – You probably have several age groups on this for instance <1, 2 -4 years, 4 – 5 years.\n\nAuthors need to clarify on the following:\n\nHow many pneumococcus serotypes were identified? Which serotype was most frequent? Can you please arrange them in a descending order?\n\nDiscussion: Adequate in content. Consider serotype replacement to enrich your discussion.\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? No\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "4348",
"date": "31 Jan 2019",
"name": "Michael Walekhwa",
"role": "Author Response",
"response": "Dear Dr. Ondigo, Many thanks for your review of this article. I have read through and keenly updated the areas you highlighted during your review. Kindly revisit. Many thanks"
}
]
},
{
"id": "39125",
"date": "01 Nov 2018",
"name": "Jackie K. Obey",
"expertise": [
"Reviewer Expertise Medicinal parasitology (Malariology)",
"immunoparasitology and medical entomology",
"antimicrobial activity of plant extracts"
],
"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 carried out by the authors on Streptococcus pneumoniae is extremely important for Kenya. It addresses a major health problem that has been addressed by other authors in the past and for which a lasting solution is currently being sought. The study is detailed and was able to employ modern techniques to assess the carriage state of children at The Gertrude's Children Hospital, Nairobi, Kenya. The methods used were appropriate and in line with the study's objectives. The sample size was appropriately determined and descriptive statistics have been used appropriately to describe the results obtained by the authors.\nThe title of the study however includes the word 'epidemiology' and this may lead the reader to think that the authors would have tried to determine factors that influenced the establishment of the research problem at the study site. The authors have not determined those factors or risks that lead to Kenyan children being vaccinated, yet not being protected by the PCV-10 vaccine. Those findings would then have given an idea of the risk of expose of children to Streptococcus pneumoniae disease and given the hospital and government the strategies to employ for preventive measures against the disease. The conclusion is brief and does not give recommendations to the Government of Kenya or The Gertrude's Children Hospital on what to do after the findings of the study.\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": "39482",
"date": "15 Nov 2018",
"name": "Felix Dube",
"expertise": [
"Reviewer Expertise Medical microbiology"
],
"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 reports pneumococcal carriage in a cohort of Kenyan children. This has significant implication when it comes to the evaluation of the impact of PCV10 on the population structure of pneumococcus.\nSpecific comments:\nPneumococcus is a normal commensal, not \"highly invasive\" as is reported in intro.\n\nThe authors must avoid the use of non-standard nomenclature such as SPn. Further, there is inconsistent use of S. pneumoniae and pneumococcus throughout the text. These cannot be used interchangeably.\n\nThe introduction needs to be reworked and repetitions avoided, i.e. in paragraph 1 and 5, the authors talk about 90 serotypes. The last sentence of paragraph 5 does not include references.\n\nThe study was conducted between 2017 and 2018, it would really have benefited from the WHO working group report on pneumococcal carriage studies1. Most importantly, 79% (164/206) being non-viable seriously indicates problems in the experimental design. The authors don't say anything about broth enrichment in order to improve recovery of the pneumococcus. Amies media is not ideal for pneumococcus compared to STGG. A 10ul innoculum is very little, did the authors attempt to use bigger volumes especially for the samples with no growth? A 2% BA plate is used as primary culture then selective media, if possible, do this in parallel.\n\nNeed to be more consistent in reporting proportions.\n\nWhat was the PCV10 vaccine coverage at each timepoint? Also authors need to report non-pcv as non-PCV10 because some of the serotypes the report as \"non-vacccine\" such as serotype 3 are included in PCV13.\n\nRepetition of results. The authors do not report any metadata looking at risk factors for carriage, hence the \"epi\" in title must fall out.\n\nThe authors repeatedly report \"decrease\" in carriage post PCV, but they really can’t say this without data on Pre-PCV10.\n\nAvoid sweeping statement to infer serotype replacement if they actually don't show this evidence.\n\n\"While this is a significant reduction from the pre-vaccine era, it is still high compared to malaria, diarrhea and HIV/AIDS\" doesn't make sense, what are the authors referring to?\n\n\"This is a very important finding as it explains the high level of child morbidity and mortality due to pneumococcal disease despite the availability of PCV-10.\" How do you arrive at this if working with carriage and not invasive disease isolates. This and other strong conclusions need to be avoided, you surely cant with 42 isolates.\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?\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": []
}
] | 1
|
https://f1000research.com/articles/7-879
|
https://f1000research.com/articles/8-128/v1
|
30 Jan 19
|
{
"type": "Review",
"title": "The role of sclerostin and dickkopf-1 in oral tissues – A review from the perspective of the dental disciplines",
"authors": [
"Mohammad Samiei",
"Klara Janjić",
"Barbara Cvikl",
"Andreas Moritz",
"Hermann Agis",
"Mohammad Samiei",
"Klara Janjić",
"Barbara Cvikl",
"Andreas Moritz"
],
"abstract": "Wnt signaling is of high relevance in the development, homeostasis, and regeneration of oral tissues. Therefore, Wnt signaling is considered to be a potential target for therapeutic strategies. The action of Wnt is tightly controlled by the inhibitors sclerostin (SOST) and Dickkopf (DKK)-1. Given the impact of SOST and DKK-1 in hard tissue formation, related diseases and healing, it is of high relevance to understand their role in oral tissues. The clinical relevance of this knowledge is further underlined by systemic and local approaches which are currently in development for treating a variety of diseases such as osteoporosis and inflammatory hard tissue resorption. In this narrative review, we summarize the current knowledge and understanding on the Wnt signaling inhibitors SOST and DKK-1, and their role in physiology, pathology, and regeneration in oral tissues. We present this role from the perspective of the different specialties in dentistry, including endodontics, orthodontics, periodontics, and oral surgery.",
"keywords": [
"Sclerostin",
"Dickkopf-1",
"Wnt Pathway",
"Regeneration",
"Periodontology",
"Endodontology"
],
"content": "Introduction\n\nResearch and development of novel treatment strategies for regenerative dentistry has become crucial to improve oral health due to the growing numbers of dental and maxillofacial problems which require treatment. Therefore, cell-therapy modalities, biologicals, and gene therapy approaches are in development with the aim to provide novel tools for the clinical demands (Dissanayaka et al., 2014; Fretwurst et al., 2018; Itoh et al., 2018; Kaigler et al., 2015; Nevins et al., 2005; Plonka et al., 2017; Taut et al., 2013). This research needs to be guided by an appropriate understanding of the cell biological mechanisms underlying pathological changes and regeneration in the oral tissue, including the periodontium and the dental tissue. Due to the major role of Wnt signaling and the respective regulators in development and healing, research in regenerative medicine and dentistry has evaluated the feasibility of targeting the pathway for therapeutic approaches (Florio et al., 2016; Heiland et al., 2010; Taut et al., 2013; Yu et al., 2018).\n\nThe Wnts comprise a family of at least 19 lipid-modified glycoproteins and short-range ligands. Wnts can activate canonical and non-canonical pathways of signaling in the cells (Mohammed et al., 2016; Tang et al., 2009; Yang et al., 2016). The canonical Wnt path involves the interaction of Wnt with frizzled (Frz) and low density lipoprotein receptor-related protein 5/6 (LRP5/6). Thereby β-catenin accumulation is induced leading to the translocation of β-catenin into the nucleus where target genes are activated (Mohammed et al., 2016; Tang et al., 2009; Yang et al., 2016) (Figure 1).\n\nScheme of the Wnt signaling pathway (A) which is inhibited by sclerostin (SOST, B) and dickkopf-1 (DKK-1, C). Adopted from (Yorgan & Schinke, 2014) and modified.\n\nWnt signaling regulates various cellular functions which include cell migration, proliferation, differentiation, apoptosis, and morphogenesis. Given this broad involvement of Wnt signaling it is not surprising that it has a key role in organ development, regeneration and homeostasis of tissues (Logan & Nusse, 2004; Sarkar & Sharpe, 2000; Seo et al., 2012). In the maxillofacial region, Wnt signaling modulates morphological patterns in tooth development including teeth number, shape, size, and positioning (Han et al., 2011; Kim et al., 2013; Kratochwil et al., 2002; Liu et al., 2008; Sarkar et al., 2000; Wang et al., 2009; Zhang et al., 2013). Maturation of dental mesenchyme into odontoblasts and cementoblasts is also modulated by Wnt signaling which highlights the relevance to dentistry (Janjić et al., 2018; Zhang et al., 2013). Given this dominant role of Wnts, Wnt signaling needs to be tightly controlled involving inhibitors.\n\nSclerostin (SOST) and dickkopf-1 (DKK-1) are the primary inhibitors controlling the Wnt signaling pathways (Figure 1). These inhibitors can directly bind to LRP5/6 and inhibit the activation of LRP5- and LRP6- related signaling. Dkk-1 binds to a larger region on LRP5 and LRP6 extracellular surface and thereby can inhibit binding with other Wnts than SOST. (Ahn et al., 2011; Bourhis et al., 2010; Bourhis et al., 2011; Cheng et al., 2011; Florio et al., 2016; Li et al., 2005; van Bezooijen et al., 2004) Thereby SOST and DKK-1 play critical roles in the formation of hard tissue and associated diseases. SOST and DKK-1 are consequently considered potential therapeutic targets for regenerative approaches (Janjić et al., 2018; Li et al., 2005; Semënov et al., 2005; Taut et al., 2013).\n\nThe glycoprotein SOST is mainly secreted by osteocytes, the mechanosensor cells of the bone, and has a major effect on bone and dental tissue (Kuchler et al., 2014; Pflanz et al., 2017; Saag et al., 2017; Taut et al., 2013; Winkler et al., 2003; Yu et al., 2018). In recent years, SOST production in oral tissues has gained more attention and led to in vitro and in vivo studies on the regulation of SOST (Pflanz et al., 2017; Taut et al., 2013; Yu et al., 2018). SOST production in mineralizing periodontal ligament cells is increased; this observation suggests that targeting of SOST might support periodontal regeneration. SOST has also been demonstrated to have a role in the dimension of the periodontal ligament (Kuchler et al., 2014). This further supports the hypothesis of a key role of SOST in oral tissues. DKK-1, another antagonist of Wnt signaling, modulates the dimension of oral tissues, including the periodontium and teeth (Bao et al., 2013; Jäger et al., 2010; MacDonald et al., 2007).\n\nAccording to the findings of 10 years analysis of publications regarding SOST and DKK-1 in medical fields (Figure 2), it becomes evident that there is a growing set of literature on SOST and DKK-1. Within dentistry the majority of publications on SOST and DKK-1 is from the field of oral surgery. This review aims to provide a synopsis of the existing knowledge on the Wnt signaling inhibitors SOST and DKK-1 from endodontics, orthodontics, periodontics, and oral surgery perspectives.\n\nThe number of publications on sclerostin (SOST, A) and dickkopf-1 (DKK-1, C) per year as found in Pubmed.org (Search terms SOST or DKK-1 with and without Medicine or Dentistry). The total number of publications on SOST (B) and DKK-1 (D) in Dentistry and the respective specialties Endodontics, Periodontics, Orthodontics, and Oral Surgery as found in Pubmed.org (Search terms SOST with and without Dentistry, Endodontics, Periodontics, Orthodontics, or Oral Surgery).\n\n\nEndodontics perspective\n\nResearch and developments in endodontics lead to the continuous improvement of the field, as the introduction of new techniques and biomaterials drive profound progress in endodontic clinical practice. In particular regenerative endodontics is an exciting dental specialty. Treatment protocols of infected root canals in immature teeth were established which allow continued apical closure and root maturation (Chueh & Huang, 2006; Huang, 2008; Huang et al., 2008). In these regenerative endodontic procedures, after treatment with ethylenediaminetetraacetic acid (EDTA) signaling factors embedded in the dentin matrix such as TGF-β are exposed and released into the canal space (Galler et al., 2015). These factors can guide the migration of dental pulp stem cells and support differentiation into odontoblast-like cells, leading to the production of mineralized matrix also termed reparative dentin (Nakashima et al., 1994; Rutherford et al., 1993; Sloan & Smith, 1999; Tziafas et al., 1998). Interestingly TGF-β has the capacity to induce SOST expression (Gruber et al., 2017; Manokawinchoke et al., 2015); therefore, it is possible that dentin conditioning can interfere with the regulation system of Wnt signaling (Gruber et al., 2017; Manokawinchoke et al., 2015).\n\nClinically applied regenerative protocols, such as revascularization, utilize the regenerative capacity of endogenous stem cells. During treatment, over instrumentation into the periapical tissues causes stem cells to migrate and enter into the canal system via the blood clot. This process leads to the treatment of immature teeth with pulp necrosis by replacing dentin, root structures and pulp-dentin complex cells (Diogenes et al., 2016). However, revascularization can be accompanied by intracanal calcification (Song et al., 2017). It is unclear how Wnts and their inhibitors SOST and DKK-1 are involved in this process. Therefore, studies which investigate the regulation of SOST and DKK-1 in pulp cells and tissue under conditions which are present in the early phase of healing are of clear relevance (Janjić et al., 2018).\n\nResults of Zhang et al. indicate that Wnt/β-catenin signaling is required for odontoblastic differentiation and also promotes proliferation of pre-odontoblasts and odontogenesis during root development (Zhang et al., 2013). Based on these findings, the integrity of Wnt/β-catenin signaling in odontoblasts is vital for proliferation and differentiation through root formation. Targeted deletion of β-catenin in odontoblasts leads to incomplete incisors and rootless molars. Furthermore, β-catenin deficiency disrupts the differentiation of odontoblasts and cementoblasts. (Zhang et al., 2013) Epithelial expression of DKK-1 or epithelium-specific inactivation of β-catenin causes abnormal tooth development at the early bud stage (Andl et al., 2002; Chen et al., 2009; Han et al., 2011). Mesenchyme-specific inhibition of β-catenin indicates the critical role of Wnt/β-catenin signaling in the potential mesenchymal odontogenic activation throughout early tooth growth (Chen et al., 2009).\n\nAfter tooth development odontoblasts secrete dentin and the pulp chamber system narrows with age (Arana-Chavez & Massa, 2004; Foster et al., 2013; Lee et al., 2013; Sakai et al., 2010). While dentin gradually thickens, the pulp chamber space is reduced and a massive bone loss can be observed during aging. (Boskey & Coleman, 2010; Carvalho & Lussi, 2017; Gabet & Bab, 2011) In response to injuries and stress the formation of mineralized tissue by odontoblasts, also termed reparative dentinogenesis, is stimulated. The odontoblasts which are involved in this repair process are clearly responsive to Wnt (Zhao et al., 2018). SOST knockout mice demonstrate dramatically enhanced formation of reparative mineralized bridges and increased mineralization in dental pulp cells compared with wild-type mice (Collignon et al., 2017). These findings are related to an increased SOST expression in wild type cells. Further more these results show that SOST deficiency accelerates reparative dentinogenesis after pulp damage and therefore inhibition of SOST may provide a promising therapeutic strategy to improve the healing of injured pulps (Collignon et al., 2017).\n\nExpression of DKK-1 is up-regulated in a rodent model of induced periapical lesions, suggesting that DKK-1 is involved in the inflammatory processes and bone resorption in periapical lesions (Zhang et al., 2014). Interestingly several potential therapeutic approaches rely on the modulation of Wnt signaling (Ishimoto et al., 2015; Lee et al., 2016; Zhao et al., 2018), including several plant-derived molecules. Baicalein promotes odontoblastic differentiation and angiogenesis of human dental pulp cells. It was suggested that via the inhibition of DKK-1 Baicalein can contribute to regenerative endodontics and dental pulp repair. (Lee et al., 2016).\n\nGiven the importance of hypoxia-induced signaling in the early phase of pulp healing we investigated the production of SOST and DKK-1 in dental pulp cells upon treatment with hypoxia or the hypoxia mimetic agent L-mimosine in monolayer, spheroid, and tooth slice cultures. Our results show that the response with regard to SOST and DKK-1 production depends on the culture model.\n\nTaken together, the literature highlights a major role of SOST and DKK-1 in tooth development and as a potential target for regenerative strategies.\n\n\nOrthodontics perspective\n\nOrthodontic tooth movement is a result of external force applications to the teeth, an intervention which involves remodeling in dental and surrounding oral tissues, such as periodontal ligament, alveolar bone, and gingiva (Antoun et al., 2017; Krishnan & Davidovitch, 2006). The applied forces induce displacement of the teeth in the periodontal ligament space, thereby establishing sites where the tissue is compressed and sites were traction establishes (Antoun et al., 2017; Krishnan & Davidovitch, 2006). Thereby, modeling of the alveolar bone through the processes of bone resorption and bone formation is induced leading to clinical changes in the position of the tooth. (Krishnan & Davidovitch, 2006; Krishnan & Davidovitch, 2009; Reitan, 1967)\n\nMechanical forces can regulate the expression of SOST by osteocytes; for instance the expression of SOST is reduced when loading is increased, and increased when loading is decreased leading to an increase in bone formation or bone loss, respectively (Moustafa et al., 2012; Robling et al., 2008). This feature of SOST indicates that it is a crucial protein in bone formation under mechanical stimulation (Agholme et al., 2011; Mantila Roosa et al., 2011; Robling et al., 2008; ). Given that SOST is produced by osteocytes and cementocytes in alveolar bone and in cellular cementum, respectively, an involvement of SOST in tooth movement is likely (Jäger et al., 2010).\n\nDuring orthodontic tooth movement, one of the essential processes is remodeling of the periodontal ligament. Periostin is one of the factors which increased in the periodontal ligament during initial stages of orthodontic tooth movement (Wilde et al., 2003), a matricellular protein and collagen-rich tissues which are highly expressed in periosteum under persistent mechanical stress (Horiuchi et al., 1999). Periodontal cells have shown to respond to these mechanical forces with an increase of TGF-β which can stimulate SOST (Gruber et al., 2017; Manokawinchoke et al., 2015). The increase in the levels of SOST in the compression side and the decrease in the tension side during tooth movement can be suggestive of an interplay between periostin from Sharpey’s fibres and SOST in alveolar bone during orthodontic therapy (Nishiyama et al., 2015).\n\nCementoblasts are highly differentiated cells from the mesenchymal lineage which generate cementum (Bosshardt, 2005). Diseases that influence bone characteristics often interfere with the properties of the cementum, which highlights the similarities of these tissues. Although SOST is expressed in both osteocytes and cementocytes in dental tissues (Jäger et al., 2010), it is still not clear if cementocytes can have similar mechanical stress sensing capacity as osteocytes. However, SOST which is released by osteocytes in the alveolar bone may affect the function of cementoblasts. This process might be modulated upon orthodontic treatment. Research on the role of SOST can also help to understand the impact of therapeutics on the periodontium. Strontium supports differentiation of cementoblasts (Bao et al., 2014) with one possible underlying mechanism being that strontium decreases the production of SOST. Thus, strontium was proposed for regenerative approaches to support cementum production in cases of root resorption. (Bao et al., 2014) However, if SOST or DKK-1 can serve effectively as targets for orthodontic approaches and cementum repair requires further studies.\n\n\nPeriodontics perspective\n\nPeriodontitis is an inflammatory disease of the teeth supporting tissues, has a multifactorial etiology. The inflammation created by specific microorganisms extends deep into the tissues and causes the destruction of the tooth supporting connective tissue and alveolar bone. This progressive process leads to the pathological impairment of collagen fibres, loss of periodontal ligament and alveolar bone recession (Armitage, 2004; Highfield, 2009; Pihlstrom et al., 2005). Evaluation of SOST and DKK-1 in chronic periodontitis patients showed that both of these inhibitors were up-regulated in the periodontal tissues of these subjects (Napimoga et al., 2014).\n\nThere are hints from animal studies that inflammation can not only trigger bone resorption, but also decrease bone formation. This anti-anabolic effects seem to be mediated by increased levels of DKK-1 (Heiland et al., 2010). In vitro models of oral soft tissue augmentation suggest that DKK-1 is also increased in oral soft tissue wound healing (Agis et al., 2014). Thus, Wnt inhibitors can play a major role in pathological processes and regeneration in the periodontal tissue. Although periodontitis can be treated in early stages, mostly because of the chronic entity of problem, it is diagnosed in the advanced phase of destruction of the periodontal ligament and the prognosis of maintaining the teeth are poor. Conventional regenerative approaches use biomaterials of natural or synthetic origin as filler for the defect thereby aiding the host to replace lost periodontal tissue and bone. While these interventions can stimulate tissue repair and stop the destruction of the periodontium, methods to archive full regeneration are still the focus of research (Hernández-Monjaraz et al., 2018).\n\nThe blocking of Wnt signaling impairs the periodontal ligament and alveolar bone (Lim et al., 2014), while enhancing Wnt signalling by SOST knock out stimulates alveolar bone formation and reduces the width of periodontal ligament (Kuchler et al., 2014). Expression of SOST by cementocytes suggests that these cells may regulate cell activity on the cementum surface (Bao et al., 2013; Lehnen et al., 2012). TGF-β can increase the production of SOST in fibroblasts from periodontal ligament and gingiva (Gruber et al., 2017). This mechanism seems to be involved in the impact of mechanical loading on mineralized tissue formation in the periodontal ligament (Manokawinchoke et al., 2015). Deletion of SOST leads to more cellular cementum, in parallel to more dramatically increased alveolar bone deposition (Kuchler et al., 2014). Blocking SOST by application of a SOST-specific antibody enhances healing of alveolar bone in experimental periodontitis (Chen et al., 2015; Liu et al., 2018; Taut et al., 2013). In addition, it was reported that reduced SOST in periostin knockout mice can re-establish periodontal ligament and alveolar bone (Rangiani et al., 2016; Ren et al., 2015). This evidence supports that targeting of SOST is a feasible approach for periodontal therapy.\n\nDental cementum is a mineralized hard tissue on the surface of root dentin and present either in acellular or cellular form. Defective cementum results in periodontal breakdown, tooth dysfunction, and finally leads to tooth loss. Cementogenesis is a key element in the process of periodontal tissue regeneration (Bosshardt, 2005; Kao & Fiorellini, 2012). SOST was detected only in cementocytes of cellular cementum in the late stages of cementum development (Lehnen et al., 2012). SOST levels in cementocytes increased in periodontal ligament cultures, following mineralization treatment (Jäger et al., 2010). Interestingly, in periodontal ligament cells Baicalein can promote osteoblastic differentiation involving Wnt/β-catenin signaling (Chen et al., 2017). DKK-1 significantly reversed the effects of Baicalein on human periodontal ligament cells (Chen et al., 2017). It is possible that this mechanism can be exploited in regenerative approaches.\n\nThe here presented literature supports the significant effects of SOST and DKK-1 in the periodontium system and periodontal diseases. As a result, they could be the main targets in future periodontics regenerative therapies.\n\n\nOral surgery perspective\n\nThe alveolar bone supports the tooth in the maxilla and mandible and is characterized by continuous and rapid remodeling in response to mechanical forces (Javed et al., 2010; Pagni et al., 2012). Thereby alveolar bone continuously adapts to functional load. If this mechanical stimuli is lacking the alveolar bone undergoes a resorptive process (Einhorn & Gerstenfeld, 2015; Pagni et al., 2012; Sodek & McKee, 2000). Following trauma due to overloading or surgery bone has the capacity to regenerate. While long bone healing occurs by endo-chondral ossification, alveolar bone healing typically occurs without histological cartilage formation (Devlin et al., 1997). The success of oral surgery procedures, such as implants, depends on the proper healing of alveolar bone and strategies which stimulate bone regeneration (Lin et al., 2011). Thus understanding the cell and molecular biological background of bone healing is clearly of clinical relevance.\n\nIn bone, SOST is mainly secreted by osteocytes and represents a key modulator of bone homeostasis (Brunkow et al., 2001; van Bezooijen et al., 2004). The importance of SOST in bone formation is illustrated by sclerosteosis, a rare autosomal recessive disorder with a loss-of-function mutation in SOST (Sebastian & Loots, 2018; Yavropoulou et al., 2014). Further evidence comes from Van Buchem Disease, which is characterized by a noncoding deletion which removes a SOST-specific regulator (Sebastian & Loots, 2018; Yavropoulou et al., 2014). These diseases show bone overgrowth, particularly in the craniofacial bones and the jaw bone (Balemans et al., 2002; Brunkow et al., 2001). There is also a phenotype in oral tissue; partial anodontia, malocclusion, and delayed tooth eruption is seen in subjects with Van Buchem Disease or SOST (Stephen et al., 2001; van Bezooijen et al., 2009). Animal studies on the role of SOST in loss of function or gain function models indicate SOST decreases bone formation and can stimulate bone resorption (Canalis, 2013; Li et al., 2008; O’Brien et al., 2013). SOST knockout mice show a dramatically increased basal mandibular bone and less effect on the coronal and apical part of the alveolar bone (Kuchler et al., 2014).\n\nThe response of bone to mechanical loading is highly important in implantology. Since SOST is a key player in the regulation of bone formation in the response to mechanical loading it is important to understand its role in the alveolar bone. Upon loading, osteocytes reduce the expression of SOST permitting Wnts to bind their receptors. (Burgers & Williams, 2013; Galli et al., 2010; Zhao et al., 2013). Interestingly, SOST expression is up-regulated around implants without primary stability (Shu et al., 2017). Thus, it is very likely that SOST regulates the adaptation of bone around dental implants to the mechanical forces of loading. Application of SOST specific antibodies has been shown to stimulate bone formation around dental implants (Yu et al., 2018).\n\nDKK-1 can decrease Wnt/β catenin signalling and reduced the production of type II collagen in chondrocytes under mechanical loading (Niu et al., 2016). Interestingly, DKK-1 has been found to play a distinct role in inflammation-induced bone loss (Heiland et al., 2010). However, when evaluating genetic markers for peri-implantitis clinically, no direct relation of DKK-1 and peri-implantitis is seen (Hall et al., 2011). Thus, further research is required to understand the role of DKK-1 in the alveolar bone.\n\nFrom a clinical standpoint, osteoporosis is a risk factor in implantology. Inspired by the fact that inhibition of SOST can effectively improve bone formation in osteoporotic patients strategies which harness this capacity to improve bone formation around implants have been evaluated. (Canalis, 2013; Costa & Bilezikian, 2012; Recker et al., 2015; Shu et al., 2017). Implant osseointegration is superior in the SOST knockout mice suggesting that SOST is a promising target to enhance implant osseointegration in osteoporosis. (Shu et al., 2017) Also application of antibodies specific for SOST can improve bone formation around implants (Yu et al., 2018).\n\nTaken together these data show the relevance of the Wnt signalling inhibitors of SOST and DKK-1 for implantology and bone augmentation, as well as for tumor development and cancer therapy. Future studies will show how effective targeting of SOST and DKK-1 can be applied to stimulate regeneration and for tumor therapy.\n\n\nConclusion\n\nThe here presented review highlights the relevance of Wnt signaling, the inhibitors SOST and DKK-1, and its key role in oral tissue homeostasis throughout life (Florio et al., 2016; Heiland et al., 2010; Witcher et al., 2018). Major sources of Wnt inhibitors are the mechanosensors of the bone, the osteocytes (Moustafa et al., 2012; Odagaki et al., 2018; Robling et al., 2008). Cells from oral tissue including cementocytes can modulate Wnt signaling by these inhibitors and mice deficient of SOST or DKK-1 have highlighted the important role in the oral tissue (Jäger et al., 2010; Lehnen et al., 2012). Based on this knowledge concepts to antagonize inhibitors Wnt signaling have been developed to support oral tissue regeneration (Taut et al., 2013; Yu et al., 2018). The future will reveal the capacity of these strategies which target these Wnt signaling inhibitors for regenerative therapy in dentistry. Given the interplay of SOST and DKK-1, a bidirectional approach which targets both SOST and DKK-1 locally has high potential (Florio et al., 2016).\n\n\nData availability\n\nNo data is associated with this article.",
"appendix": "Grant information\n\nWe thank the European Society of Endodontology (ESE) for financial support of our research [ESE research grant 2015]. The authors deny any conflict of interest.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nAgholme F, Isaksson H, Li X, et al.: Anti-sclerostin antibody and mechanical loading appear to influence metaphyseal bone independently in rats. Acta Orthop. 2011; 82(5): 628–632. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgis H, Collins A, Taut AD, et al.: Cell population kinetics of collagen scaffolds in ex vivo oral wound repair. PLoS One. 2014; 9(11): e112680. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAhn VE, Chu ML, Choi HJ, et al.: Structural basis of Wnt signaling inhibition by Dickkopf binding to LRP5/6. Dev Cell. 2011; 21(5): 862–873. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAndl T, Reddy ST, Gaddapara T, et al.: WNT signals are required for the initiation of hair follicle development. Dev Cell. 2002; 2(5): 643–653. PubMed Abstract | Publisher Full Text\n\nAntoun JS, Mei L, Gibbs K, et al.: Effect of orthodontic treatment on the periodontal tissues. Periodontol 2000. 2017; 74(1): 140–157. PubMed Abstract | Publisher Full Text\n\nArana-Chavez VE, Massa LF: Odontoblasts: the cells forming and maintaining dentine. Int J Biochem Cell Biol. 2004; 36(8): 1367–1373. PubMed Abstract | Publisher Full Text\n\nArmitage GC: Periodontal diagnoses and classification of periodontal diseases. Periodontol 2000. 2004; 34(1): 9–21. PubMed Abstract | Publisher Full Text\n\nBalemans W, Patel N, Ebeling M, et al.: Identification of a 52 kb deletion downstream of the SOST gene in patients with van Buchem disease. J Med Genet. 2002; 39(2): 91–97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBao X, Liu Y, Han G, et al.: The effect on proliferation and differentiation of cementoblast by using sclerostin as inhibitor. Int J Mol Sci. 2013; 14(10): 21140–21152. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBao X, Liu X, Zhang Y, et al.: Strontium promotes cementoblasts differentiation through inhibiting sclerostin expression in vitro. Biomed Res Int. 2014; 2014: 487535. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoskey AL, Coleman R: Aging and bone. J Dent Res. 2010; 89(12): 1333–1348. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBosshardt DD: Are cementoblasts a subpopulation of osteoblasts or a unique phenotype? J Dent Res. 2005; 84(5): 390–406. PubMed Abstract | Publisher Full Text\n\nBourhis E, Tam C, Franke Y, et al.: Reconstitution of a frizzled8.Wnt3a.LRP6 signaling complex reveals multiple Wnt and Dkk1 binding sites on LRP6. J Biol Chem. 2010; 285(12): 9172–9179. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBourhis E, Wang W, Tam C, et al.: Wnt antagonists bind through a short peptide to the first β-propeller domain of LRP5/6. Structure. 2011; 19(10): 1433–1442. PubMed Abstract | Publisher Full Text\n\nBrunkow ME, Gardner JC, Van Ness J, et al.: Bone dysplasia sclerosteosis results from loss of the SOST gene product, a novel cystine knot-containing protein. Am J Hum Genet. 2001; 68(3): 577–589. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBurgers TA, Williams BO: Regulation of Wnt/β-catenin signaling within and from osteocytes. Bone. 2013; 54(2): 244–249. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCanalis E: Wnt signalling in osteoporosis: mechanisms and novel therapeutic approaches. Nat Rev Endocrinol. 2013; 9(10): 575–583. PubMed Abstract | Publisher Full Text\n\nCarvalho TS, Lussi A: Age-related morphological, histological and functional changes in teeth. J Oral Rehabil. 2017; 44(4): 291–298. PubMed Abstract | Publisher Full Text\n\nChen H, Xu X, Liu M, et al.: Sclerostin antibody treatment causes greater alveolar crest height and bone mass in an ovariectomized rat model of localized periodontitis. Bone. 2015; 76: 141–148. PubMed Abstract | Publisher Full Text\n\nChen J, Lan Y, Baek JA, et al.: Wnt/beta-catenin signaling plays an essential role in activation of odontogenic mesenchyme during early tooth development. Dev Biol. 2009; 334(1): 174–185. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChen LJ, Hu BB, Shi XL, et al.: Baicalein enhances the osteogenic differentiation of human periodontal ligament cells by activating the Wnt/β-catenin signaling pathway. Arch Oral Biol. 2017; 78: 100–108. PubMed Abstract | Publisher Full Text\n\nCheng Z, Biechele T, Wei Z, et al.: Crystal structures of the extracellular domain of LRP6 and its complex with DKK1. Nat Struct Mol Biol. 2011; 18(11): 1204–1210. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChueh LH, Huang GT: Immature teeth with periradicular periodontitis or abscess undergoing apexogenesis: a paradigm shift. J Endod. 2006; 32(12): 1205–1213. PubMed Abstract | Publisher Full Text\n\nCollignon AM, Amri N, Lesieur J, et al.: Sclerostin Deficiency Promotes Reparative Dentinogenesis. J Dent Res. 2017; 96(7): 815–821. PubMed Abstract | Publisher Full Text\n\nCosta AG, Bilezikian JP: Sclerostin: therapeutic horizons based upon its actions. Curr Osteoporos Rep. 2012; 10(1): 64–72. PubMed Abstract | Publisher Full Text\n\nDevlin H, Hoyland J, Newall JF, et al.: Trabecular bone formation in the healing of the rodent molar tooth extraction socket. J Bone Miner Res. 1997; 12(12): 2061–2067. PubMed Abstract | Publisher Full Text\n\nDiogenes A, Ruparel NB, Shiloah Y, et al.: Regenerative endodontics: A way forward. J Am Dent Assoc. 2016; 147(5): 372–380. PubMed Abstract | Publisher Full Text\n\nDissanayaka WL, Zhu L, Hargreaves KM, et al.: Scaffold-free Prevascularized Microtissue Spheroids for Pulp Regeneration. J Dent Res. 2014; 93(12): 1296–1303. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEinhorn TA, Gerstenfeld LC: Fracture healing: mechanisms and interventions. Nat Rev Rheumatol. 2015; 11(1): 45–54. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFlorio M, Gunasekaran K, Stolina M, et al.: A bispecific antibody targeting sclerostin and DKK-1 promotes bone mass accrual and fracture repair. Nat Commun. 2016; 7: 11505. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFoster BL, Nagatomo KJ, Tso HW, et al.: Tooth root dentin mineralization defects in a mouse model of hypophosphatasia. J Bone Miner Res. 2013; 28(2): 271–282. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFretwurst T, Larsson L, Yu SH, et al.: Periodontal Tissue Bioengineering: Is the Future Now? Compend Contin Educ Dent. 2018; 39(4): 218–223; quiz 224. PubMed Abstract\n\nGabet Y, Bab I: Microarchitectural changes in the aging skeleton. Curr Osteoporos Rep. 2011; 9(4): 177–183. PubMed Abstract | Publisher Full Text\n\nGaller KM, Buchalla W, Hiller KA, et al.: Influence of root canal disinfectants on growth factor release from dentin. J Endod. 2015; 41(3): 363–368. PubMed Abstract | Publisher Full Text\n\nGalli C, Passeri G, Macaluso GM: Osteocytes and WNT: the mechanical control of bone formation. J Dent Res. 2010; 89(4): 331–343. PubMed Abstract | Publisher Full Text\n\nGruber M, Gruber R, Agis H: Transforming growth factor-β1 increases sclerostin in fibroblasts of the periodontal ligament and the gingiva – Matters. Matters Archive. 2017. Reference Source\n\nHall J, Britse AO, Jemt T, et al.: A controlled clinical exploratory study on genetic markers for peri-implantitis. Eur J Oral Implantol. 2011; 4(4): 371–382. PubMed Abstract\n\nHan XL, Liu M, Voisey A, et al.: Post-natal effect of overexpressed DKK1 on mandibular molar formation. J Dent Res. 2011; 90(11): 1312–1317. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHeiland GR, Zwerina K, Baum W, et al.: Neutralisation of Dkk-1 protects from systemic bone loss during inflammation and reduces sclerostin expression. Ann Rheum Dis. 2010; 69(12): 2152–2159. PubMed Abstract | Publisher Full Text\n\nHernández-Monjaraz B, Santiago-Osorio E, Monroy-García A, et al.: Mesenchymal Stem Cells of Dental Origin for Inducing Tissue Regeneration in Periodontitis: A Mini-Review. Int J Mol Sci. 2018; 19(4): pii: E944. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHighfield J: Diagnosis and classification of periodontal disease. Aust Dent J. 2009; 54 Suppl 1: S11–26. PubMed Abstract | Publisher Full Text\n\nHoriuchi K, Amizuka N, Takeshita S, et al.: Identification and characterization of a novel protein, periostin, with restricted expression to periosteum and periodontal ligament and increased expression by transforming growth factor beta. J Bone Miner Res. 1999; 14(7): 1239–1249. PubMed Abstract | Publisher Full Text\n\nHuang GT: A paradigm shift in endodontic management of immature teeth: conservation of stem cells for regeneration. J Dent. 2008; 36(6): 379–386. PubMed Abstract | Publisher Full Text\n\nHuang GT, Sonoyama W, Liu Y, et al.: The hidden treasure in apical papilla: the potential role in pulp/dentin regeneration and bioroot engineering. J Endod. 2008; 34(6): 645–651. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIshimoto K, Hayano S, Yanagita T, et al.: Topical application of lithium chloride on the pulp induces dentin regeneration. PLoS One. 2015; 10(3): e0121938. PubMed Abstract | Publisher Full Text | Free Full Text\n\nItoh Y, Sasaki JI, Hashimoto M, et al.: Pulp Regeneration by 3-dimensional Dental Pulp Stem Cell Constructs. J Dent Res. 2018; 97(10): 1137–1143. PubMed Abstract | Publisher Full Text\n\nJäger A, Götz W, Lossdörfer S, et al.: Localization of SOST/sclerostin in cementocytes in vivo and in mineralizing periodontal ligament cells in vitro. J Periodontal Res. 2010; 45(2): 246–254. PubMed Abstract | Publisher Full Text\n\nJanjić K, Cvikl B, Kurzmann C, et al.: Do hypoxia and L-mimosine modulate sclerostin and dickkopf-1 production in human dental pulp-derived cells? Insights from monolayer, spheroid and tooth slice cultures. BMC Oral Health. 2018; 18(1): 36. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJaved A, Chen H, Ghori FY: Genetic and transcriptional control of bone formation. Oral Maxillofac Surg Clin North Am. 2010; 22(3): 283–93, v. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaigler D, Avila-Ortiz G, Travan S, et al.: Bone Engineering of Maxillary Sinus Bone Deficiencies Using Enriched CD90+ Stem Cell Therapy: A Randomized Clinical Trial. J Bone Miner Res. 2015; 30(7): 1206–1216. PubMed Abstract | Publisher Full Text\n\nKao DW, Fiorellini JP: Regenerative periodontal therapy. Front Oral Biol. 2012; 15: 149–159. PubMed Abstract | Publisher Full Text\n\nKim TH, Bae CH, Lee JC, et al.: β-catenin is required in odontoblasts for tooth root formation. J Dent Res. 2013; 92(3): 215–221. PubMed Abstract | Publisher Full Text\n\nKratochwil K, Galceran J, Tontsch S, et al.: FGF4, a direct target of LEF1 and Wnt signaling, can rescue the arrest of tooth organogenesis in Lef1-/- mice. Genes Dev. 2002; 16(24): 3173–3185. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKrishnan 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\nKrishnan V, Davidovitch Z: On a path to unfolding the biological mechanisms of orthodontic tooth movement. J Dent Res. 2009; 88(7): 597–608. PubMed Abstract | Publisher Full Text\n\nKuchler U, Schwarze UY, Dobsak T, et al.: Dental and periodontal phenotype in sclerostin knockout mice. Int J Oral Sci. 2014; 6(2): 70–76. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee SI, Kim SY, Park KR, et al.: Baicalein Promotes Angiogenesis and Odontoblastic Differentiation via the BMP and Wnt Pathways in Human Dental Pulp Cells. Am J Chin Med. 2016; 44(7): 1457–1472. PubMed Abstract | Publisher Full Text\n\nLee YH, Kim GE, Cho HJ, et al.: Aging of in vitro pulp illustrates change of inflammation and dentinogenesis. J Endod. 2013; 39(3): 340–345. PubMed Abstract | Publisher Full Text\n\nLehnen SD, Götz W, Baxmann M, et al.: Immunohistochemical evidence for sclerostin during cementogenesis in mice. Ann Anat. 2012; 194(5): 415–421. PubMed Abstract | Publisher Full Text\n\nLi X, Ominsky MS, Niu QT, et al.: Targeted deletion of the sclerostin gene in mice results in increased bone formation and bone strength. J Bone Miner Res. 2008; 23(6): 860–869. PubMed Abstract | Publisher Full Text\n\nLi X, Zhang Y, Kang H, et al.: Sclerostin binds to LRP5/6 and antagonizes canonical Wnt signaling. J Biol Chem. 2005; 280(20): 19883–19887. PubMed Abstract | Publisher Full Text\n\nLim WH, Liu B, Cheng D, et al.: Wnt signaling regulates homeostasis of the periodontal ligament. J Periodontal Res. 2014; 49(6): 751–759. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLin Z, Rios HF, Volk SL, et al.: Gene expression dynamics during bone healing and osseointegration. J Periodontol. 2011; 82(7): 1007–1017. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu F, Chu EY, Watt B, et al.: Wnt/beta-catenin signaling directs multiple stages of tooth morphogenesis. Dev Biol. 2008; 313(1): 210–224. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu M, Kurimoto P, Zhang J, et al.: Sclerostin and DKK1 Inhibition Preserves and Augments Alveolar Bone Volume and Architecture in Rats with Alveolar Bone Loss. J Dent Res. 2018; 97(9): 1031–1038. PubMed Abstract | Publisher Full Text\n\nLogan CY, Nusse R: The Wnt signaling pathway in development and disease. Annu Rev Cell Dev Biol. 2004; 20: 781–810. PubMed Abstract | Publisher Full Text\n\nMacDonald BT, Joiner DM, Oyserman SM, et al.: Bone mass is inversely proportional to Dkk1 levels in mice. Bone. 2007; 41(3): 331–339. PubMed Abstract | Publisher Full Text | Free Full Text\n\nManokawinchoke J, Limjeerajarus N, Limjeerajarus C, et al.: Mechanical Force-induced TGFB1 Increases Expression of SOST/POSTN by hPDL Cells. J Dent Res. 2015; 94(7): 983–989. PubMed Abstract | Publisher Full Text\n\nMantila Roosa SM, Liu Y, Turner CH: Gene expression patterns in bone following mechanical loading. J Bone Miner Res. 2011; 26(1): 100–112. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMohammed MK, Shao C, Wang J, et al.: Wnt/β-catenin signaling plays an ever-expanding role in stem cell self-renewal, tumorigenesis and cancer chemoresistance. Genes Dis. 2016; 3(1): 11–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoustafa A, Sugiyama T, Prasad J, et al.: Mechanical loading-related changes in osteocyte sclerostin expression in mice are more closely associated with the subsequent osteogenic response than the peak strains engendered. Osteoporos Int. 2012; 23(4): 1225–1234. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNakashima M, Nagasawa H, Yamada Y, et al.: Regulatory role of transforming growth factor-beta, bone morphogenetic protein-2, and protein-4 on gene expression of extracellular matrix proteins and differentiation of dental pulp cells. Dev Biol. 1994; 162(1): 18–28. PubMed Abstract | Publisher Full Text\n\nNapimoga MH, Nametala C, da Silva FL, et al.: Involvement of the Wnt-β-catenin signalling antagonists, sclerostin and dickkopf-related protein 1, in chronic periodontitis. J Clin Periodontol. 2014; 41(6): 550–557. PubMed Abstract | Publisher Full Text\n\nNevins M, Giannobile WV, McGuire MK, et al.: Platelet-derived growth factor stimulates bone fill and rate of attachment level gain: results of a large multicenter randomized controlled trial. J Periodontol. 2005; 76(12): 2205–2215. PubMed Abstract | Publisher Full Text\n\nNishiyama Y, Matsumoto T, Lee JW, et al.: Changes in the spatial distribution of sclerostin in the osteocytic lacuno-canalicular system in alveolar bone due to orthodontic forces, as detected on multimodal confocal fluorescence imaging analyses. Arch Oral Biol. 2015; 60(1): 45–54. PubMed Abstract | Publisher Full Text\n\nNiu Q, Li F, Zhang L, et al.: Role of the Wnt/β-catenin signaling pathway in the response of chondrocytes to mechanical loading. Int J Mol Med. 2016; 37(3): 755–762. PubMed Abstract | Publisher Full Text\n\nO'Brien CA, Nakashima T, Takayanagi H: Osteocyte control of osteoclastogenesis. Bone. 2013; 54(2): 258–263. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOdagaki N, Ishihara Y, Wang Z, et al.: Role of Osteocyte-PDL Crosstalk in Tooth Movement via SOST/Sclerostin. J Dent Res. 2018; 97(12): 1374–1382. PubMed Abstract | Publisher Full Text\n\nPagni G, Pellegrini G, Giannobile WV, et al.: Postextraction alveolar ridge preservation: biological basis and treatments. Int J Dent. 2012; 2012: 151030. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPflanz D, Birkhold AI, Albiol L, et al.: Sost deficiency led to a greater cortical bone formation response to mechanical loading and altered gene expression. Sci Rep. 2017; 7(1): 9435. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPihlstrom BL, Michalowicz BS, Johnson NW: Periodontal diseases. Lancet. 2005; 366(9499): 1809–1820. PubMed Abstract | Publisher Full Text\n\nPlonka AB, Khorsand B, Yu N, et al.: Effect of sustained PDGF nonviral gene delivery on repair of tooth-supporting bone defects. Gene Ther. 2017; 24(1): 31–39. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRangiani A, Jing Y, Ren Y, et al.: Critical roles of periostin in the process of orthodontic tooth movement. Eur J Orthod. 2016; 38(4): 373–378. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRecker RR, Benson CT, Matsumoto T, et al.: A randomized, double-blind phase 2 clinical trial of blosozumab, a sclerostin antibody, in postmenopausal women with low bone mineral density. J Bone Miner Res. 2015; 30(2): 216–224. PubMed Abstract | Publisher Full Text\n\nReitan K: Clinical and histologic observations on tooth movement during and after orthodontic treatment. Am J Orthod. 1967; 53(10): 721–745. PubMed Abstract | Publisher Full Text\n\nRen Y, Han X, Ho SP, et al.: Removal of SOST or blocking its product sclerostin rescues defects in the periodontitis mouse model. FASEB J. 2015; 29(7): 2702–2711. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobling AG, Niziolek PJ, Baldridge LA, et al.: Mechanical stimulation of bone in vivo reduces osteocyte expression of Sost/sclerostin. J Biol Chem. 2008; 283(9): 5866–5875. PubMed Abstract | Publisher Full Text\n\nRutherford RB, Wahle J, Tucker M, et al.: Induction of reparative dentine formation in monkeys by recombinant human osteogenic protein-1. Arch Oral Biol. 1993; 38(7): 571–576. PubMed Abstract | Publisher Full Text\n\nSaag KG, Petersen J, Brandi ML, et al.: Romosozumab or Alendronate for Fracture Prevention in Women with Osteoporosis. N Engl J Med. 2017; 377(15): 1417–1427. PubMed Abstract | Publisher Full Text\n\nSakai VT, Zhang Z, Dong Z, et al.: SHED differentiate into functional odontoblasts and endothelium. J Dent Res. 2010; 89(8): 791–796. PubMed Abstract | Publisher Full Text\n\nSarkar L, Cobourne M, Naylor S, et al.: Wnt/Shh interactions regulate ectodermal boundary formation during mammalian tooth development. Proc Natl Acad Sci U S A. 2000; 97(9): 4520–4524. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSarkar L, Sharpe PT: Inhibition of Wnt signaling by exogenous Mfrzb1 protein affects molar tooth size. J Dent Res. 2000; 79(4): 920–925. PubMed Abstract | Publisher Full Text\n\nSebastian A, Loots GG: Genetics of Sost/SOST in sclerosteosis and van Buchem disease animal models. Metabolism. 2018; 80: 38–47. PubMed Abstract | Publisher Full Text\n\nSemënov M, Tamai K, He X: SOST is a ligand for LRP5/LRP6 and a Wnt signaling inhibitor. J Biol Chem. 2005; 280(29): 26770–26775. PubMed Abstract | Publisher Full Text\n\nSeo MS, Hwang KG, Kim H, et al.: Analysis of gene expression during odontogenic differentiation of cultured human dental pulp cells. Restor Dent Endod. 2012; 37(3): 142–148. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShu R, Ai D, Bai D, et al.: The effects of SOST on implant osseointegration in ovariectomy osteoporotic mice. Arch Oral Biol. 2017; 74: 82–91. PubMed Abstract | Publisher Full Text\n\nSloan AJ, Smith AJ: Stimulation of the dentine-pulp complex of rat incisor teeth by transforming growth factor-beta isoforms 1-3 in vitro. Arch Oral Biol. 1999; 44(2): 149–156. PubMed Abstract | Publisher Full Text\n\nSodek J, McKee MD: Molecular and cellular biology of alveolar bone. Periodontol 2000. 2000; 24(1): 99–126. PubMed Abstract | Publisher Full Text\n\nSong M, Cao Y, Shin SJ, et al.: Revascularization-associated Intracanal Calcification: Assessment of Prevalence and Contributing Factors. J Endod. 2017; 43(12): 2025–2033. PubMed Abstract | Publisher Full Text\n\nStephen LX, Hamersma H, Gardner J, et al.: Dental and oral manifestations of sclerosteosis. Int Dent J. 2001; 51(4): 287–290. PubMed Abstract | Publisher Full Text\n\nTang N, Song WX, Luo J, et al.: BMP-9-induced osteogenic differentiation of mesenchymal progenitors requires functional canonical Wnt/beta-catenin signalling. J Cell Mol Med. 2009; 13(8B): 2448–2464. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTaut AD, Jin Q, Chung JH, et al.: Sclerostin antibody stimulates bone regeneration after experimental periodontitis. J Bone Miner Res. 2013; 28(11): 2347–2356. PubMed Abstract | Publisher Full Text\n\nTziafas D, Alvanou A, Papadimitriou S, et al.: Effects of recombinant basic fibroblast growth factor, insulin-like growth factor-II and transforming growth factor-beta 1 on dog dental pulp cells in vivo. Arch Oral Biol. 1998; 43(6): 431–444. PubMed Abstract | Publisher Full Text\n\nvan Bezooijen RL, Bronckers AL, Gortzak RA, et al.: Sclerostin in mineralized matrices and van Buchem disease. J Dent Res. 2009; 88(6): 569–574. PubMed Abstract | Publisher Full Text\n\nvan Bezooijen RL, Roelen BA, Visser A, et al.: Sclerostin is an osteocyte-expressed negative regulator of bone formation, but not a classical BMP antagonist. J Exp Med. 2004; 199(6): 805–814. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang XP, O’Connell DJ, Lund JJ, et al.: Apc inhibition of Wnt signaling regulates supernumerary tooth formation during embryogenesis and throughout adulthood. Development. 2009; 136(11): 1939–1949. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWilde J, Yokozeki M, Terai K, et al.: The divergent expression of periostin mRNA in the periodontal ligament during experimental tooth movement. Cell Tissue Res. 2003; 312(3): 345–351. PubMed Abstract | Publisher Full Text\n\nWinkler DG, Sutherland MK, Geoghegan JC, et al.: Osteocyte control of bone formation via sclerostin, a novel BMP antagonist. EMBO J. 2003; 22(23): 6267–6276. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWitcher PC, Miner SE, Horan DJ, et al.: Sclerostin neutralization unleashes the osteoanabolic effects of Dkk1 inhibition. JCI Insight. 2018; 3(11): pii: 98673. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang K, Wang X, Zhang H, et al.: The evolving roles of canonical WNT signaling in stem cells and tumorigenesis: implications in targeted cancer therapies. Lab Invest. 2016; 96(2): 116–136. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYavropoulou MP, Xygonakis C, Lolou M, et al.: The sclerostin story: from human genetics to the development of novel anabolic treatment for osteoporosis. Hormones (Athens). 2014; 13(4): 323–337. PubMed Abstract | Publisher Full Text\n\nYorgan TA, Schinke T: Relevance of Wnt signaling for osteoanabolic therapy. Mol Cell Ther. 2014; 2: 22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYu SH, Hao J, Fretwurst T, et al.: Sclerostin-Neutralizing Antibody Enhances Bone Regeneration Around Oral Implants. Tissue Eng Part A. 2018; 24(21–22): 1672–1679. PubMed Abstract | Publisher Full Text\n\nZhang R, Huang S, Wang L, et al.: Histochemical localization of Dickkopf-1 in induced rat periapical lesions. J Endod. 2014; 40(9): 1394–1399. PubMed Abstract | Publisher Full Text\n\nZhang R, Yang G, Wu X, et al.: Disruption of Wnt/β-catenin signaling in odontoblasts and cementoblasts arrests tooth root development in postnatal mouse teeth. Int J Biol Sci. 2013; 9(3): 228–236. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhao L, Shim JW, Dodge TR, et al.: Inactivation of Lrp5 in osteocytes reduces young's modulus and responsiveness to the mechanical loading. Bone. 2013; 54(1): 35–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhao Y, Yuan X, Liu B, et al.: Wnt-Responsive Odontoblasts Secrete New Dentin after Superficial Tooth Injury. J Dent Res. 2018; 97(9): 1047–1054. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "43827",
"date": "04 Feb 2019",
"name": "Francisco Javier Rodríguez-Lozano",
"expertise": [
"Reviewer Expertise endodontics",
"stem cells",
"dental stem 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\nThe manuscript fits into the aim of the journal. The manuscript is well written and the results presented interesting.\nI would like to know the search strategy. Inclusion/exclusion criteria.\nMinor aspects:\nEndodontic perspective: \"Given the importance of hypoxia-induced signaling in the early phase of pulp healing we investigated the production of SOST and DKK-1 in dental pulp cells upon treatment with hypoxia or the hypoxia mimetic agent L-mimosine in monolayer, spheroid, and tooth slice cultures\". Citation?\nPeriodontic Perspective: \" Dental cementum is a mineralized hard tissue on the surface of root dentin and present either in acellular or cellular form. Defective cementum results in periodontal breakdown, tooth dysfunction, and finally leads to tooth loss.\" I recommend adding a reference.\nFinally, I recommend re-writing the conclusion without citations\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": "46090",
"date": "15 Apr 2019",
"name": "Qiming Jin",
"expertise": [
"Reviewer Expertise periodontal regeneration and dental implant researches"
],
"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 is a review paper on sclerostin and dickkopf-1 impacts on dentistry. The authors made good publication search work, and reviewed the expressions and biological functions of sclerostin and DKK-1 in each parts of tooth and surrounding tissues. Furthermore, the authors explored the correlation between SOST and DKK-1 and various pathologies. More importantly, the treatment pathways to target SOST and DKK-1 were discussed profoundly. The only flaw is the lack of SOST and DKK-1 in enamel formation, which does not affect the merits of this paper.\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/8-128
|
https://f1000research.com/articles/8-124/v1
|
30 Jan 19
|
{
"type": "Research Article",
"title": "Identification of brain regions associated with working memory deficit in schizophrenia",
"authors": [
"Indranath Chatterjee",
"Virendra Kumar",
"Sahil Sharma",
"Divyanshi Dhingra",
"Bharti Rana",
"Manoj Agarwal",
"Naveen Kumar",
"Virendra Kumar",
"Sahil Sharma",
"Divyanshi Dhingra",
"Bharti Rana",
"Manoj Agarwal",
"Naveen Kumar"
],
"abstract": "Background: Schizophrenia, a severe psychological disorder, shows symptoms such as hallucinations and delusions. In addition, patients with schizophrenia often exhibit a deficit in working memory which adversely impacts the attentiveness and the behavioral characteristics of a person. Although several clinical efforts have already been made to study working memory deficit in schizophrenia, in this paper, we investigate the applicability of a machine learning approach for identification of the brain regions that get affected by schizophrenia leading to the dysfunction of the working memory. Methods: We propose a novel scheme for identification of the affected brain regions from functional magnetic resonance imaging data by deploying group independent component analysis in conjunction with feature extraction based on statistical measures, followed by sequential forward feature selection. The features that show highest accuracy during the classification between healthy and schizophrenia subjects are selected. Results: This study reveals several brain regions like cerebellum, inferior temporal gyrus, superior temporal gyrus, superior frontal gyrus, insula, and amygdala that have been reported in the existing literature, thus validating the proposed approach. We are also able to identify some functional changes in the brain regions, such as Heschl gyrus and the vermian area, which have not been reported in the literature involving working memory studies amongst schizophrenia patients. Conclusions: As our study confirms the results obtained in earlier studies, in addition to pointing out some brain regions not reported in earlier studies, the findings are likely to serve as a cue for clinical investigation, leading to better medical intervention.",
"keywords": [
"functional Magnetic Resonance Imaging",
"fMRI",
"Schizophrenia",
"Working Memory",
"Group Independent Component Analysis",
"Classification",
"Computer-aided Diagnosis"
],
"content": "Introduction\n\nSchizophrenia is a psychological disorder that involves auditory and visual hallucinations and delusions. A schizophrenia patient often shows symptoms such as disorganized thinking, difficulty in speech, and abnormal motor behavior. Structural and functional changes occur in the brain due to various chemical alterations in the schizophrenic patient. These changes adversely impact behavioral, emotional and cognitive capabilities of a patient. Schizophrenia patients often experience deterioration or impairment in working memory (WM)1. WM is a short-term memory of a person for perceiving things that relate to immediate consciousness that helps in language processing, decision making and reasoning2,3. It is an active and readily accessible mental state that maintains information and processes the information selectively4.\n\nThe use of functional magnetic resonance imaging (fMRI) has facilitated the diagnosis and treatment of neurological and psychological disorders and enhanced our understanding of the brain. The blood oxygenation level-dependent (BOLD) technique has been widely used in fMRI studies; it relies on the effect of magnetic susceptibility of deoxyhaemoglobin. When a brain region is activated by a task, it demands an increased inflow of oxygenated blood and a net increase in signal intensity is observed. Various paradigms such as visual task and auditory oddball task have been designed to find the pattern of functioning of the brain during different cognitive processes. As the Sternberg item recognition paradigm (SIRP) task is a popular working memory task5,6, we have used the fMRI data of the subjects performing this task.\n\nAs fMRI data involves 3-D scans of the whole brain volume across time, it is inherently high dimensional. Independent component analysis (ICA)7 is a popular method that can be applied on fMRI data to produce the temporally coherent brain networks. ICA is a data-driven approach that generates independent components without making any assumptions about the characteristics of the task and time courses. Group-ICA (GICA) is an extension of ICA that helps to analyze group fMRI studies. In this study, ICA is employed to find such independent networks that have significant differences in the regions between healthy subjects and schizophrenic patients affecting the working memory of a person.\n\nIn this study, we aim to identify the brain regions, potentially responsible for the working memory dysfunction, using fMRI data involving SIRP task. Towards this end, we have developed a decision model to differentiate between schizophrenia patients and healthy subjects (controls). We have applied group ICA to find the functionally connected components. We have demarcated the brain regions based on Automated Anatomical Labeling (AAL) atlas. In order to carry out the feature extraction, statistical measures are used to evaluate the significance of different regions. Finally, classification guided feature selection is done using support vector machine (SVM) and 1-NN classifiers.\n\nSeveral psychological, neurological, and computational studies1,6,8–12 have been conducted to identify the pattern of brain activation for different mental tasks in the schizophrenia patient. Park and Holzman1 found that schizophrenia patients suffer a loss in representational processing, leading to working memory deficit. Impairment of performance in working memory tasks such as the Wisconsin Card-Sorting Test (WCST) is an important evidence of the dysfunction of frontal lobe amongst the schizophrenia patients. Gold et al.11 studied the effect of schizophrenia on working memory dysfunction by performing WCST and letter-number (LN) span test on a group of 36 patients with schizophrenia and 30 healthy controls. They found that patients with schizophrenia showed poor performance on the WCST and LN span test, indicating the failure of working memory, typically attributed to frontal lobe dysfunction. Bertilino et al.8 performed WCST on a population of 13 patients with schizophrenia and an equal number of healthy subjects to identify the relationship between neuronal pathology of the dorsolateral prefrontal cortex (DLPFC) and activation of working memory network in the cortical region. They found that the rate of N-acetylaspartate level in the DLPFC was firmly linked with the activation of the working memory cortical network during the working memory tasks in schizophrenia patients.\n\nSome researchers experimented with other visual tasks like Sternberg Item Recognition Paradigm (SIRP) to evaluate the impact of schizophrenia on the working memory. Manoach et al.6 performed the SIRP task on 12 schizophrenic and 10 healthy subjects. Using SIRP task in fMRI, they compared the activation of DLPFC between the patients and the healthy subjects. A high working memory load condition was compared with non-working memory condition as well as with low working memory load condition. They found that schizophrenia patients performed poorly in comparison to the healthy subjects under different load conditions. They also noted increased DLPFC activation in schizophrenics in comparison to healthy subjects during WM task. In another study, Manoach et al.13 examined the participation of brain regions in WM performance by analyzing region-wise brain activations in fMRI data from nine schizophrenic subjects and an equal number of healthy subjects while performing a modified version of the SIRP task, which included a cash reward for correct responses. Again, they compared the high and the low working memory load conditions to each other, keeping the non-working memory condition as a baseline. It was seen that schizophrenic patients showed weak working memory performance along with activation in basal ganglia and thalamus. These regions were found to be activated only in the schizophrenia group. In an fMRI study involving 106 schizophrenic subjects and 111 healthy matched controls, Potkin et al.12 examined the BOLD signal change in the DLPFC in a working memory study using SIRP task. They found significantly greater DLPFC activation in patients with schizophrenia. The activation was found to vary with variation in working memory load. The mean BOLD signal was also found to be higher during intermediate memory loads in schizophrenic subjects as compared to the healthy controls. Wible et al.14 examined auditory hallucinations while performing the SIRP task in a group of 74 schizophrenic patients, subdivided into non-hallucinating and hallucinating groups. They found that the patients having auditory hallucinations showed decreased functional activity during the probe condition in working memory task mainly in the inferior parietal regions and superior temporal regions in comparison to those not having hallucinations.\n\nICA treats fMRI data as a linear combination of spatially independent components. These components derived from the fMRI data suggest the functional connectivity between brain regions (also called brain networks). Some of the fMRI studies9 used general linear model (GLM) approach to convert 4D time-series data into a 3D statistical parametric map. Pearson’s correlation coefficient15 and regional homogeneity16 were also applied in fMRI study to extract information from temporal data. Kim et al.17 used ICA to trace the temporally coherent networks in fMRI activity using a working memory task. Using the fMRI dataset for 115 patients with chronic schizophrenia and 130 healthy controls performing the SIRP task, they identified six components mainly showing disease-relevant brain networks. These components showed the regions that exhibited significant differences in the functioning of WM networks between schizophrenic patients and healthy controls. Two out of the six networks showed regions covering working memory areas such as bilateral DLPFC, inferior parietal lobules and cerebellum. They observed dysfunction in default mode network (DMN) in schizophrenia which exists across multiple subnetworks in the region. Correa et al.10 also explored the role of ICA in the analysis of fMRI data. They compared the performance of different ICA algorithms and performed an analysis of fMRI data having visual-motor task and estimated activations using Infomax, FastICA, eigenvalue decomposition (EVD) and joint approximate diagonalization of eigen matrices (JADE). The authors concluded that the infomax performed quite well on the fMRI data and showed the highest t-values and successfully estimated maximally independent components.\n\n\nMethods\n\nThe fMRI data used in this study were downloaded from the Functional BIRN Data Repository (http://fbirnbdr.birncommunity.org:8080/BDR/)12. A detailed description of the data is available at the repository. In brief, all the acquisitions were carried out using 1.5T scanners keeping all other parameters same for all the subjects across the datasets. In this study, we have considered SIRP task fMRI data available at site 0009 and site 0010 of the FBIRN repository. All the three runs of each subject’s scan are used in our experiments. All subjects had regular hearing levels and sufficient eyesight to perform the SIRP task. They were able to perform the cognitive task. Healthy subjects were excluded if they had a current or past history of head injury and major medical illness. All the healthy subjects were free from any antipsychotic exposure and they had no recent history of medication effect. fMRI data from the patients with schizophrenia and schizoaffective disorder meeting the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) criteria were included in this study. FBIRN had determined the symptom scores by using the Schedule for the Assessment of Positive Symptoms (SAPS) and Negative Symptoms assessment measures5. Table 1 summarizes the database details.\n\n*Data given as mean ± standard deviation.\n\nThe functional scans were acquired using T2*-weighted gradient echo planar imaging (EPI) sequences and were parameterized by Orientation: anterior commissure-posterior commissure line; the number of slices: 27; slice thickness: 4 mm; TR: 2 seconds; time to echo: 40 ms; matrix: 64 × 64; field of view: 22 cm; and flip angle: 90°12.\n\nIn this paper, we have considered the Sternberg item recognition paradigm (SIRP) task18,19. The SIRP is a block design task that assesses the maintenance and scanning components of WM4,19. Each phase began with the presentation of a memory set composed of one, three, or five digits, constituting three levels of WM load (low 1L, medium 3L, high 5L). This encode phase was followed by the presentation of 14 probe digits. Participants responded to each probe using a button box to indicate whether the probe digit was in the memory set. Each of the three runs contained two blocks of each of the three load phases, presented in a pseudorandom order with the blocks of each phase alternating with fixation epochs (a baseline resting period). Each run lasted for 6 minutes.\n\nFor preprocessing the raw fMRI datasets taken from FBIRN repository, we have used the Statistical Parametric Mapping version 8 (SPM8, Wellcome Trust Centre for Neuroimaging, University College London, UK)20 toolbox in Matlab. The preprocessing steps are as follows. Realignment and reslicing were performed on each of the images using the default parameters. Slice timing correction was applied to correct possible errors introduced by temporal variations during the acquisition of fMRI data. Subsequently, the fMRI scans were spatially normalized into the standard Montreal Neurological Institute (MNI) space using an EPI template. Thus, the volume of each voxel in raw fMRI scans changed from 3.4 × 3.4 × 4 mm3 to 3 × 3 × 3 mm3. This resulted in a brain volume of 53 × 63 × 46 voxels. Finally, spatial smoothing was done using a 9 × 9 × 9 mm3 full width at half-maximum (FWHM) Gaussian kernel on the normalized volumes to get the smoothed volumes.\n\nThe proposed approach is divided into the following phases: (i) application of group ICA; (ii) statistical feature extraction; (iii) classification guided feature selection; and (iv) visualization. These phases are described in the following sub-sections. The proposed approach is applied to individual ICs. The stepwise description of the proposed approach is outlined in Algorithm 1. Figure 1 shows the overall workflow of the study.\n\n1. Application of group ICA:\n\n(a) Apply GICA on the pre-processed fMRI data, where, the modified MDL criteria is used to identify the number of IC.\n\n2. Feature Extraction:\n\n(a) Segment each IC for each subject in 116 regions using AAL atlas.\n\n(b) Extract five statistical features namely, mean, standard deviation, kurtosis, skewness and entropy from each region for every subject on the basis of voxel values of that particular region. Thus a subject is represented as 580 features (=116 x 5). The dataset is represented as\n\nX→68×580=[f→1f→2f→3⋯f→580], where f→1 is the ith feature.\n\n3. Feature Selection:\n\n(a) Carry out feature selection in LOOCV manner. In ith fold of LOOCV, all, but ith sample is used for training.\n\n(b) Compute FDR score for each feature using equation mentioned in Section 4.2.4.\n\n(c) Rank the features on the basis of FDR score (the feature with highest FDR score is assigned rank 1).\n\n(d) Build decision model (DM) incrementally (forward feature selection) using SVM classifier. Begin by building the DM with the first ranked feature and add the FDR ranked features, one by one, to obtain the high classification accuracy.\n\n4. Visualization:\n\n(a) Identify the set of features having maximum classification accuracy.\n\n(b) Backtrack the features to the MNI brain space to locate the affected brain regions.\n\nThe BOLD fMRI technique acquires 3-D brain volumes across time. Each voxel in the whole brain volume contains a value that corresponds to the change of signal intensity of the voxel across time. To identify the connected brain networks that are activated while performing a task, we applied group ICA (GICA) using the GIFT toolbox v.4.0b21 in MATLAB. There are three main stages in GICA: data compression (also called data reduction), ICA, and back reconstruction. In the data compression step, principal components analysis (PCA) is used to reduce the size of the data. Group PCA is applied to all subjects. Then, ICA is used to find the independent components (ICs) and the spatial maps. Although several ICA algorithms, such as Infomax, FastICA, Jade, and AMUSE, are available in the GIFT toolbox, we use the most widely used Informax algorithm to find the ICs. The Infomax algorithm22 uses a non-linear function to maximize the information transfer from the input layer to the output layer of a network. The components resulting from ICA represent the brain networks activated during the task. The back-reconstruction step produces the ICs with the most accurate spatial maps and time courses for each subject.\n\nIn our experiments, the number of ICs was estimated using the modified minimum description length (MDL)23 criteria, which generated 13 ICs. We have used the average ICs spatial map for each subject corresponding to the three runs. In the proposed approach, we have analyzed 13 ICs independently for each subject.\n\nIn the first phase, we have segmented each of the 13 ICs for each subject using Automated Anatomical Labeling (AAL)24 atlas. AAL atlas segments the whole brain volume into 116 brain regions. Thereafter, subject-wise features were extracted from each of these 116 regions for each IC.\n\nFor dimensionality reduction, we have extracted five statistical features for each brain region of each subject, namely, mean, standard deviation (std), skewness, kurtosis, and entropy. If Vr = [v1, v2, v3, ..., vN] is the voxel set having N voxels for rth region, then these statistical measures are defined as follows:\n\n\n\n\n\n\n\n\n\n\n\nThus, for each subject, we extracted 580 (= 116 × 5) features. To identify the features relevant for identifying affected brain regions in schizophrenia, we carried out feature selection.\n\nFeature selection is a process of selecting a relevant subset of the feature set. In this paper, we have incorporated the classification guided sequential forward feature selection method in a leave-one-out cross-validation (LOOCV) manner. In the sequential forward selection, one adds the best features in every iteration, until the best classification accuracy is achieved. We have used Fisher’s discriminant ratio (FDR) score for ranking each feature. FDR score was computed using the formula,\n\n\n\nwhere\n\nx = vector containing the feature x values corresponding to all subjects,\n\nmeanh = mean of feature x corresponding to healthy patients;\n\nmeans = mean of feature x corresponding to schizophrenic patients;\n\nvarh = variance of feature x corresponding to healthy patients;\n\nvars = variance of feature x corresponding to schizophrenic patients.\n\nAfter scoring all the features, the scores were arranged in descending order. The high value of FDR indicates that the within-class scatter is low, while between-class scatter/variance is high. Forward feature selection approach is employed to identify the feature set generating high classification accuracy.\n\nSeveral works9,25–27 have attempted to identify the affected brain regions using a decision model to classify schizophrenia patients and healthy controls. In this paper, classification task is performed on the combined data involving the healthy and schizophrenic subjects using linear SVM and k - nearest neighbors (k-NN) classifiers. Classification is done in LOOCV manner i.e., training is done on all the subjects excluding one subject, which is used for testing. The classification model is built incrementally. Finally, the feature subset yielding high classification accuracy for a given test sample is chosen.\n\nEach sample is an input vector Xi (i = 1, 2, 3, ..., n) having features selected from the statistical measure of a particular region (set of voxels) and is associated with one of the two classes Yi = +1 or Yi = -1 (binary class). The class labels +1 and -1 refers to the positive class (schizophrenia) and the negative class respectively. For classification using SVM, the libsvm version 3.2328 package in Matlab-2014b is used that uses C-SVC. Besides setting all the training parameters as default, we experiment with the different values of cost parameter (C), varying C in the range of 0.01 to 1000 in powers of 10. For k-NN classifier, we take the value of k as 1 and use Euclidean distance as the distance metric.\n\nThe set of selected features, obtained after all the iterations of LOOCV approach for each IC, were backtracked to brain space to identify the affected brain regions. In order to find the most relevant regions that may contribute to the dysfunction of the working memory in the schizophrenia patients, the brain regions identified by the proposed approach, marked by different independent components, were coalesced. The frequently occurring regions were plotted on a mask using WFU PickAtlas. The mask was then overlaid onto a standard T1-weighted MRI using MANGO version 4.0.1 toolbox29. These identified brain regions were overlaid onto a standard T-1 weighted image and visualized MANGO toolbox.\n\n\nResults\n\nThe first phase of the proposed method resulted in 13 spatial ICs (see Figure 2). Figure 2a–f shows the composite view of multiple independent components showing functionally connected brain regions involved during the task. The figure highlights task-related components with functional differences across healthy and schizophrenia subjects. We have used the forward feature selection method in the second phase. The average classification accuracy using LOOCV scheme for SVM and k-NN classifiers for each IC is shown in Figure 3 and Figure 4, respectively. Overall, the linear SVM classifier (C=1.09) yielded classification accuracy in the range 94% to 100%. Similarly, the use of k-NN classifier resulted in classification accuracy in the range of 96–100% (see Figure 4 for linear SVM and Figure 5 for 1-NN classifier). Finally, the affected brain regions identified from the visualization phase are mentioned in Table 2 for each spatial IC. Table 2 shows the regions marked by increased activation in case of schizophrenia patients when compared to the healthy controls. Figure 5 (a–c)) shows the identified regions such as the cerebellum, temporal and frontal gyrus, insula, amygdala, cuneus, putamen, Heschl gyrus, and vermis.\n\nThe connected brain networks identified by (A) independent component (IC) 1–IC 2, (B) IC 3–IC 4, (C) IC 5–IC 6, (D) IC 7–IC 8, (E) IC 9–IC 10, and (F) IC 11–IC 13, are shown.\n\n(A) shows the cerebellum (red), inferior and superior temporal gyrus (green), and superior frontal gyrus (blue); (B) shows the insula (red), amygdala (green), and cuneus (blue); (C) shows the Heschl gyrus (red), vermis (green) and putamen (blue).\n\n\nDiscussion\n\nThis study aimed to identify affected brain regions in the working memory of schizophrenia patients. To achieve this, we proposed a model wherein we utilized the GICA to obtain spatial ICs, extracted statistical features from 116 brain regions, selected features using a classifier-guided forward feature selection approach, and visualization of affected brain regions. Using the proposed approach, we marked the differences in the functional activation of the following brain regions in most of the ICs (the cerebellum, inferior temporal gyrus, superior temporal gyrus, superior frontal gyrus, Heschl gyrus, insula, amygdala, vermis, thalamus, calcarine, occipital lobe and hypocampus) in schizophrenia patients in comparison to healthy controls. It may be noted that the brain regions identified in this study are largely in conformity with the previous studies30–34. Further, the connected brain regions discovered by the different spatial ICs, largely confirm to each other.\n\nOur results show the functional changes in the cerebellum region. Previous studies32,33 also suggest some changes in cortical cerebellar regions and its functional connectivity in working memory performance in schizophrenia patients. We found functional changes in the inferior temporal gyrus, superior temporal gyrus and superior frontal gyrus. While earlier studies32,35,36 suggest changes in activation and abnormal functional connectivity in temporal and frontal gyri in schizophrenia patients compared to healthy subjects, our results show changes in the Heschl gyrus region. In support of our finding, we may note that in a study by Hirayasu et al.30, they found structural volume reduction in Heschl gyri in schizophrenia patients. Grey matter atrophy in Heschl gyri was also found in a study by Kasai et al.31. In this regard, we may say that the study of relationships between the functional activations in the relating to working memory dysfunction and structural brain changes in Heschl gyrus could be an important direction of research. We find significant changes in the insula region of the schizophrenic brain similar to the result of other previous studies37,38. Our results also show functional changes in the amygdala region. While performing working memory tasks, evidence of the dysfunction or abnormalities in the amygdala in schizophrenia patients were found in several previous studies39–42. We also found some changes in the functional activation in the vermian area, specifically in cerebellar vermis. Although, some literature43,44 reports structural changes in the vermis region, our study may be a cue for further research based on ROI study to trace the changes in this region associated with schizophrenia and working memory task.\n\nIn addition, we obtained a high classification accuracy (>95%) to distinguish healthy subjects and schizophrenic. The high classification accuracy obtained by applying the proposed approach proves its efficacy in comparison to the other fMRI studies17,45,46. Overall, the proposed approach found to be effective and efficient in the identification of affected brain regions responsible for working memory dysfunction in schizophrenia.\n\n\nConclusion\n\nIn this fMRI study, based on working memory task, a feature selection scheme has been proposed to identify the brain regions affected amongst the schizophrenia patients. This study helps in the identification of brain regions responsible for impairment of working memory in schizophrenia patients. While many connected brain regions identified in our study confirm the findings of the previous studies, the results reveal some new regions in the brain which have not been reported till date in the working memory literature for schizophrenia. These regions may play role in dysfunction of working memory in the patients and could be the subject of further studies.\n\n\nData availability\n\nThe SIRP task fMRI data from the FBIRN phase II repository can be downloaded from http://schizconnect.org/queries/new, querying 1.5T fMRI data for healthy and schizophrenia subjects available at site 0009 and 0010. The list of subjects chosen for this study is mentioned in the ‘dataset_SubjectID_list.txt’ file available with the codes. Users are required to sign-up to SchizConnect to download data; conditions of use are as written in the data use agreement of the FBIRN project.\n\nTo download the data used in this study, the user has to select the project as ‘Study: fBIRNPhaseII__0010’, add ‘AND’, and select MRI as ‘Field Strength: 1.5’.\n\n\nSoftware availability\n\nThe complete source code is archived at: https://doi.org/10.5281/zenodo.252877347.\n\nLicense: Creative Commons Zero v1.0 Universal.",
"appendix": "Grant information\n\nThis work was supported by the research fellowship of Indranath Chatterjee from Council of Scientific and Industrial Research (CSIR), India having grant number 09/045(1323)/2014-EMR-I.\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\nData used in this work are taken from the Functional Biomedical Informatics Research Networks (FBIRN) data repository, under the following support: for function data, U24-RR021992, Function BIRN, and U24 GM104203, Bio-Informatics Research Network Coordinating Centre (BIRN-CC). The data were obtained from the Function BIRN Data Repository, Project Accession Number 2007-BDR-6UHZ1.\n\n\nReferences\n\nPark S, Holzman PS: Schizophrenics show spatial working memory deficits. Arch Gen Psychiatry. 1992; 49(12): 975–982. PubMed Abstract | Publisher Full Text\n\nConway A, Jarrold C, Kane M: Variation in working memory. Oxford University Press, 2008. Publisher Full Text\n\nMiller GA, Galanter E, Pribram KH: Plans and the structure of behavior. Adams Bannister Cox, 1986. Reference Source\n\nBaddeley A: Working memory: looking back and looking forward. Nat Rev Neurosci. 2003; 4(10): 829–39. PubMed Abstract | Publisher Full Text\n\nKim DI, Mathalon DH, Ford JM, et al.: Auditory oddball deficits in schizophrenia: an independent component analysis of the fMRI multisite function BIRN study. Schizophr Bull. 2009; 35(1): 67–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nManoach DS, Press DZ, Thangaraj V, et al.: Schizophrenic subjects activate dorsolateral prefrontal cortex during a working memory task, as measured by fMRI. Biol Psychiatry. 1999; 45(9): 1128–1137. PubMed Abstract | Publisher Full Text\n\nCommon P: Independent component analysis, a new concept? Signal processing. 1994; 36(3): 287–314. Publisher Full Text\n\nBertolino A, Esposito G, Callicott JH, et al.: Specific relationship between prefrontal neuronal N-acetylaspartate and activation of the working memory cortical network in schizophrenia. Am J Psychiatry. 2000; 157(1): 26–33. PubMed Abstract | Publisher Full Text\n\nChatterjee I, Agarwal M, Rana B, et al.: Bi-objective approach for computer-aided diagnosis of schizophrenia patients using fMRI data. Multimed Tools Appl. 2018; 77(20): 26991–27015. Publisher Full Text\n\nCorrea N, Adali T, Calhoun VD: Performance of blind source separation algorithms for fMRI analysis using a group ICA method. Magn Reson Imaging. 2007; 25(5): 684–694. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGold JM, Carpenter C, Randolph C, et al.: Auditory working memory and Wisconsin Card Sorting Test performance in schizophrenia. Arch Gen Psychiatry. 1997; 54(2): 159–165. PubMed Abstract | Publisher Full Text\n\nPotkin SG, Turner JA, Brown GG, et al.: Working memory and DLPFC inefficiency in schizophrenia: the FBIRN study. Schizophr Bull. 2009; 35(1): 19–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nManoach DS, Gollub RL, Benson ES, et al.: Schizophrenic subjects show aberrant fMRI activation of dorsolateral prefrontal cortex and basal ganglia during working memory performance. Biol Psychiatry. 2000; 48(2): 99–109. PubMed Abstract | Publisher Full Text\n\nWible CG, Lee K, Molina I, et al.: fMRI activity correlated with auditory hallucinations during performance of a working memory task: data from the FBIRN consortium study. Schizophr Bull. 2009; 35(1): 47–57. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChyzhyk D, Savio A, Graña M: Computer aided diagnosis of schizophrenia on resting state fMRI data by ensembles of ELM. Neural Netw. 2015; 68: 23–33. PubMed Abstract | Publisher Full Text\n\nZang Y, Jiang T, Lu Y, et al.: Regional homogeneity approach to fMRI data analysis. NeuroImage. 2004; 22(1): 394–400. PubMed Abstract | Publisher Full Text\n\nKim DI, Manoach DS, Mathalon DH, et al.: Dysregulation of working memory and default-mode networks in schizophrenia using independent component analysis, an fBIRN and MCIC study. Hum Brain Mapp. 2009; 30(11): 3795–3811. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrown GG, McCarthy G, Bischoff-Grethe A, et al.: Brain-performance correlates of working memory retrieval in schizophrenia: a cognitive modeling approach. Schizophr Bull. 2008; 35(1): 32–46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSternberg S: Memory-scanning: mental processes revealed by reaction-time experiments. Am Sci. 1969; 57(4): 421–457. PubMed Abstract\n\nPenny WD, Friston KJ, Ashburner JT, et al.: Statistical parametric mapping: the analysis of functional brain images. Elsevier, 2011. Publisher Full Text\n\nRachakonda S, Egolf E, Correa N, et al.: Group ica of fmri toolbox (gift) manual. Dostupn´e z [cit 2011-11-5]; 2007. Reference Source\n\nBell AJ, Sejnowski TJ: The \"independent components\" of natural scenes are edge filters. Vision Res. 1997; 37(23): 3327–3338. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLi YO, Adalı T, Calhoun VD: Estimating the number of independent components for functional magnetic resonance imaging data. Hum Brain Mapp. 2007; 28(11): 1251–1266. PubMed Abstract | Publisher Full Text\n\nTzourio-Mazoyer N, Landeau B, Papathanassiou D, et al.: Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. NeuroImage. 2002; 15(1): 273–289. PubMed Abstract | Publisher Full Text\n\nCaprihan A, Pearlson GD, Calhoun VD: Application of principal component analysis to distinguish patients with schizophrenia from healthy controls based on fractional anisotropy measurements. NeuroImage. 2008; 42(2): 675–682. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDu W, Calhoun VD, Li H, et al.: High classification accuracy for schizophrenia with rest and task FMRI data. Front Hum Neurosci. 2012; 6: 145. PubMed Abstract | Publisher Full Text | Free Full Text\n\nViviani R, Grön G, Spitzer M: Functional principal component analysis of fMRI data. Hum Brain Mapp. 2005; 24(2): 109–29. PubMed Abstract | Publisher Full Text\n\nChang CC, Lin CJ: LIBSVM: A library for support vector machines. ACM Transactions on Intelligent Systems and Technology. 2011; 2(3): 27. Publisher Full Text\n\nLancaster JL, Laird AR, Eickhoff SB, et al.: Automated regional behavioral analysis for human brain images. Front Neuroinform. 2012; 6: 23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHirayasu Y, McCarley RW, Salisbury DF, et al.: Planum temporale and Heschl gyrus volume reduction in schizophrenia: a magnetic resonance imaging study of first-episode patients. Arch Gen Psychiatry. 2000; 57(7): 692–699. PubMed Abstract | Free Full Text\n\nKasai K, Shenton ME, Salisbury DF, et al.: Progressive decrease of left Heschl gyrus and planum temporale gray matter volume in first-episode schizophrenia: a longitudinal magnetic resonance imaging study. Arch Gen Psychiatry. 2003; 60(8): 766–775. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeyer-Lindenberg A, Poline JB, Kohn PD, et al.: Evidence for abnormal cortical functional connectivity during working memory in schizophrenia. Am J Psychiatry. 2001; 158(11): 1809–1817. PubMed Abstract | Publisher Full Text\n\nSchlösser R, Gesierich T, Kaufmann B, et al.: Altered effective connectivity during working memory performance in schizophrenia: a study with fMRI and structural equation modeling. NeuroImage. 2003; 19(3): 751-63. PubMed Abstract | Publisher Full Text\n\nChatterjee I: Mean deviation based identification of activated voxels from time-series fMRI data of schizophrenia patients [version 2; referees: 2 approved]. F1000Res. 2018, 2018; 7: 1615. Publisher Full Text\n\nGarrity AG, Pearlson GD, McKiernan K, et al.: Aberrant \"default mode\" functional connectivity in schizophrenia. Am J Psychiatry. 2007; 164(3): 450–457. PubMed Abstract | Publisher Full Text\n\nStevens AA, Goldman-Rakic PS, Gore JC, et al.: Cortical dysfunction in schizophrenia during auditory word and tone working memory demonstrated by functional magnetic resonance imaging. Arch Gen Psychiatry. 1998; 55(12): 1097–1103. PubMed Abstract | Publisher Full Text\n\nGlahn DC, Ragland JD, Abramoff A, et al.: Beyond hypofrontality: a quantitative meta-analysis of functional neuroimaging studies of working memory in schizophrenia. Hum Brain Mapp. 2005; 25(1): 60–9. PubMed Abstract | Publisher Full Text\n\nTan HY, Choo WC, Fones CS, et al.: fMRI study of maintenance and manipulation processes within working memory in first-episode schizophrenia. Am J Psychiatry. 2005; 162(10): 1849–1858. PubMed Abstract | Publisher Full Text\n\nAnticevic A, Repovs G, Corlett PR, et al.: Negative and nonemotional interference with visual working memory in schizophrenia. Biol Psychiatry. 2011; 70(12): 1159–1168. PubMed Abstract | Publisher Full Text\n\nGruber O, Tost H, Henseler I, et al.: Pathological amygdala activation during working memory performance: Evidence for a pathophysiological trait marker in bipolar affective disorder. Hum Brain Mapp. 2010; 31(1): 115–125. PubMed Abstract | Publisher Full Text\n\nMeda SA, Stevens MC, Folley BS, et al.: Evidence for anomalous network connectivity during working memory encoding in schizophrenia: an ICA based analysis. PLoS One. 2009; 4(11): e7911. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSeidman LJ, Thermenos HW, Poldrack RA, et al.: Altered brain activation in dorsolateral prefrontal cortex in adolescents and young adults at genetic risk for schizophrenia: an fMRI study of working memory. Schizophr Res. 2006; 85(1–3): 58–72. PubMed Abstract | Publisher Full Text\n\nLevitt JJ, McCarley RW, Nestor PG, et al.: Quantitative volumetric MRI study of the cerebellum and vermis in schizophrenia: clinical and cognitive correlates. Am J Psychiatry. 1999; 156(7): 1105–1107. PubMed Abstract | Free Full Text\n\nTran KD, Smutzer GS, Doty RL, et al.: Reduced Purkinje cell size in the cerebellar vermis of elderly patients with schizophrenia. Am J Psychiatry. 1998; 155(9): 1288–1290. PubMed Abstract | Publisher Full Text\n\nArbabshirani MR, Kiehl KA, Pearlson GD, et al.: Classification of schizophrenia patients based on resting-state functional network connectivity. Front Neurosci. 2013; 7: 133. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang H, Liu J, Sui J, et al.: A Hybrid Machine Learning Method for Fusing fMRI and Genetic Data: Combining both Improves Classification of Schizophrenia. Front Hum Neurosci. 2010; 4: 192. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChatterjee I: GICA supported region-based feature selection technique for fMRI data. (Version v1.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.2528773"
}
|
[
{
"id": "43805",
"date": "21 Feb 2019",
"name": "Nipa Roy",
"expertise": [
"Reviewer Expertise Brain dynamics",
"complex systems",
"plasma"
],
"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 revealed some important brain regions active for schizophrenia patients in addition to some other reported regions in previous studies. This will give some clinical advantages in this certain brain disease. However, this whole study is based on the dysfunction of working memory, and a portion of fMRI data set. The authors did not emphasized enough on the selection criteria of elimination of certain region of the same data set. Furthermore, are the authors thinking not to generalize their finding of “some functional changes in the brain regions, such as Heschl gyrus and the vermian area” at this point (in this work) due to the sample size they are taking into account?\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? Partly",
"responses": []
},
{
"id": "45564",
"date": "18 Apr 2019",
"name": "Lovekesh Vig",
"expertise": [
"Reviewer Expertise Cognitive Neuroscience",
"Neural Networks"
],
"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 presents techniques to identify regions in working memory affected by schizophrenia from fMRI data. The procedure followed is methodical but the claims to identify previously unidentified brain regions need clinical verification. The classification is yielding impressive results but the dataset is too small to be conclusive, however I understand how difficult it is to obtain fMRI data for schizophrenic patients. Overall, I am satisfied with the technical contribution in this article.\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? Partly",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-124
|
https://f1000research.com/articles/7-1880/v1
|
02 Dec 18
|
{
"type": "Opinion Article",
"title": "It’s the network, stupid: a population’s sexual network connectivity determines its STI prevalence",
"authors": [
"Chris R. Kenyon",
"Wim Delva",
"Wim Delva"
],
"abstract": "There is little consensus as to why sexually transmitted infections (STIs), including HIV and bacterial vaginosis (BV) are more prevalent in some populations than others. Using a broad definition of sexual network connectivity that includes both structural and conductivity-related factors, we argue that the available evidence suggests that high prevalence of traditional STIs, HIV and BV can be parsimoniously explained by these populations having more connected sexual networks. Positive feedback, whereby BV and various STIs enhance the spread of other STIs, then further accentuates the spread of BV, HIV and other STIs. We review evidence that support this hypothesis and end by suggesting study designs that could further evaluate the hypothesis, as well as implications of this hypothesis for the prevention and management of STIs.",
"keywords": [
"HIV",
"STI",
"bacterial vaginosis",
"sexual network connectivity",
"concurrency"
],
"content": "Introduction\n\nThere is little consensus as to why the prevalence of bacterial vaginosis (BV), HIV and other sexually transmitted infections (STIs) varies so dramatically around the world. A range of explanations have been put forward, including variation in circumcision prevalence1, STI treatment efficacy2, poverty2–4, socioeconomic inequality5, gender inequality6, migration intensity7, hormonal contraception8, vaginal microbiome9, host genetic susceptibility10 and sexual behavior11,12. We do not dispute that each of these can play a role in differential STI spread. Rather we argue that differential connectivity of sexual networks emerges as a parsimonious dominant explanation for the global variation in STI prevalence, taking a central position in the causal pathway that links all of the above-mentioned risk factors for STI infection (Figure 1).\n\nIn both populations, STI acquisition commences when ‘A’ has sex with an older man and acquires BV-associated bacteria (yellow) and HSV-2 (black border around each node). In the high connectivity population ‘A’ also acquires T. vaginalis (TV; red) from this relationship. The major determinant of the difference in network connectivity is that more relationships run concurrently in the high connectivity population. This facilitates STI spread by: i) Creating a larger reachable path for STIs13, ii) Removing the benefits of partner sequencing seen in serial monogamy (for details see 14), iii) Reducing the time between STI transmissions since infections are not trapped in dyads14 and iv) Bypassing the rapid-clearance-in-males-buffer15,16. This is the buffer that reduces STI spread in serial monogamous networks where the gap between partnerships (time points 1 and 2) is longer than the duration of colonization of TV and BV-associated bacteria in men. This gap protects the women at time point 2 in the serial monogamy/low connectivity population (represented by the partner of B at time point 2) but not in the high connectivity population from BV and TV acquisition. Various STIs, including HSV-2, BV and TV enhance the susceptibility/infectiousness of other STIs, leading to positive feedback loops. This is conveyed via the transmission probabilities being depicted as proportional to edge width. The high connectivity network also has a low prevalence of circumcision and condom usage which further increase STI transmission probabilities in this population. The combination of high network connectivity, low circumcision/condom-use lead to a rapid spread of multiple STIs in the high- but not the low-connectivity network (blue nodes, no STIs; squares, men; circles, women).\n\n\nOutline and origins of the network connectivity theory\n\nSTIs are transmitted along sexual networks and, as a result, the structural characteristics of these networks determine the speed and extent of STI spread13,17,18. These structural characteristics include summary measures of the number of partners per unit time, coital frequency, prevalence of concurrent partnering (having two or more partners at the same time), size of core groups (and their connections with non-core populations), type of sex, size of sexual network, length of gaps between partnerships and degree/type of homophily13,18–21 (reviewed in 14). These structural factors determine the forward reachable path of a network, which is defined as the cumulative set of individuals in a population that can be infected with an STI from an initial seed via a path of temporally ordered partnerships22. Two particularly important determinants of the forward reachable path are the prevalence of concurrency and the number of partners per unit time22.\n\nSTI transmission can also be enhanced through a sexual network by factors that enhance the conductivity or probability of STI transmission per sex act. These factors include a low prevalence of circumcision, pre-exposure prophylaxis (PrEP) and condom usage (Figure 1 and Figure 2). Enhanced screening/early and effective treatment of STIs could reduce spread of STIs via reducing the duration of infectivity. Because numerous STIs enhance the transmission/acquisition of other STIs, effective STI control could then also reduce the conductivity of a network. We use a broad definition of network connectivity in this paper that includes both these structural and conductivity-related factors.\n\nOur definition of network connectivity is broad: in addition to considering the sexual links between individuals, it takes into account the “conductivity” and timing of these links.\n\nThe origins of this network connectivity theory lie in the STI modeling field. Previous modelling studies from the 1970s established the importance of the rate of partner change and mixing between core and non-core groups to STI spread23,24. Seminal modelling papers by Morris et al.13 and Watts et al.18 in the 1990s built on these findings by revealing that the prevalence of sexual partner concurrency may be a particularly important determinant of network connectivity. Their analyses found that relatively small increases in concurrency could lead to dramatic increases in network connectivity and as a result, HIV spread13. The main mechanisms whereby concurrency promotes STI spread are illustrated in Figure 1. A number of empirical studies have subsequently established that markers of network connectivity such as concurrency and rate of partner change are correlated with the prevalence of all major STIs (Table 1). In this paper, we review some of the cross sectional and longitudinal evidence that two components of network connectivity (concurrency and rate of partner change) are associated with STI prevalence. We then summarize evidence that network connectivity influences the prevalence of BV and end by noting the potential for positive feedback loops between various STIs being underpinned by network connectivity.\n\nStudies that found no association are not included.\n\nNo. of sex partners refers to number of partners over lifetime or over past year. Concurrency refers to the prevalence of concurrency (male, female or combined) for the population level studies and partner concurrency in the individual level studies. NA, not available/no studies found that evaluated this association.\n\n\nMarkers of network connectivity are correlated with the prevalence of STIs: cross-sectional evidence\n\na. Ethnic group comparative analyses\n\nUSA: In the United States the prevalence of BV, HIV and most STIs for non-Hispanic blacks in the 1990s was considerably higher than in non-Hispanic whites (Figure 3). Historical data is limited but the available data demonstrates that these divergences in prevalence extend back to the 1930s for syphilis53,54 and the 1970s for HSV-255. Morris et al. used five large national behavioural surveys to investigate which possible risk factors could underpin these differences in HIV prevalence, and found that the prevalence of concurrency was on average 3.5 and 2.1 times higher in non-Hispanic black men and women, respectively. In their modelling analysis, they found that these differences in concurrency prevalence between these groups translated into 2.6-fold differences in HIV prevalence. They did not, however, model the enhanced transmission probability that is associated with acute HIV which subsequent analyses have shown to have a synergistic effect with concurrency on HIV transmission56. Subsequent studies have demonstrated that concurrency plays an important role in the spread of the other STIs and thus the differential concurrency prevalence they found could represent a parsimonious explanation for the differences in the range of STI prevalence demonstrated in Figure 3.\n\nFigure modified from 12 and 52.\n\nSouthern Africa: The HIV prevalence varies 40-fold between ethnic groups in South Africa57. Analyses from 5 nationally representative behavioural surveys revealed that the most plausible risk factors that could explain this were the 5- to 17-fold higher prevalence of male concurrency and the higher number of partners per year in the highest compared to the lowest HIV prevalence ethnic group57. A modelling study likewise demonstrated that the combination of concurrency and rate of partner change was responsible for approximately 75% of the HIV infections in the 1990s when antenatal HIV prevalence increased from 0.7% to 24.5%58. In a similar vein, modelling studies from Zimbabwe found that both the observed high prevalence of concurrency and the increased transmission probability associated with acute HIV were needed to replicate Zimbabwe's explosive HIV epidemic curve56.\n\nElsewhere: The prevalence of concurrency and/or number of partners have also been found to be associated with variations in HIV prevalence by ethnic group in Ethiopia34, Honduras59, Kenya60, Uganda61 and the United Kingdom62–64. Although no published study has assessed how generalizable these findings are globally, one study attempted to do this within sub-Saharan African countries. This study used demographic and health surveys to systematically assess the behavioural correlates of HIV prevalence by region (as a proxy for ethnic group) in 47 surveys from 27 African countries where HIV prevalence varied by at least two-fold between regions. It found that the lifetime number of partners reported by men and women was positively correlated with HIV prevalence in 23 and 18 out of 36 surveys, respectively. Likewise, reporting sex with a non-marital, non-cohabiting partner by men and women was positively correlated with HIV prevalence in 38 and 39 out of 47 surveys, respectively65.\n\nb. Country level comparisons\n\nIn the country level analysis we focus on studies that investigate the correlates of country level peak HIV prevalence. Peak HIV prevalence, which represents the maximal HIV prevalence that countries obtained prior to the widespread availability of antiretroviral therapy is a useful composite measure of the factors that enabled the rapid spread of HIV66. Peak HIV prevalence is based on generally high quality longitudinal data on HIV prevalence around the world. As such it offers a useful outcome measure to assess what the correlates of rapid and extensive HIV transmission are. Studies have shown that peak HIV prevalence is not associated with a number of risk factors widely believed to be important for HIV spread: poverty, socioeconomic inequality, gender inequality, prevalence of migration and STI treatment efficacy6,11,67,68. These findings are of considerable consequence. Many authors have claimed that STIs are diseases of poverty2–4. Studies from Africa and elsewhere suggest that this is far from universally the case. Using HIV-serolinked and nationally representative survey data from eight countries in sub-Saharan Africa, Mishra et al., for example, established that HIV prevalence increased monotonically with wealth quintile for both men and women69. This finding has since been confirmed in 19 other countries32.\n\nOnly two risk factors have been consistently found to be associated with peak HIV prevalence: circumcision and the prevalence of concurrency.\n\ni. Circumcision: There is a strong negative association between circumcision and peak HIV prevalence within sub-Saharan Africa, but not globally11,68. This is unsurprising since sub-Saharan Africa has the highest prevalence of HIV and the second highest prevalence of circumcision in the world68,70. The vast majority of the world’s population lives in countries with both low HIV and low circumcision prevalence11. Various lines of evidence suggest that something else is driving the spread of HIV in sub-Saharan Africa and that circumcision is then moderating this risk68,71.\n\nii. Concurrency prevalence: The prevalence of male concurrency has been found to be associated with peak HIV prevalence in a cross country study39. Other studies have however failed to find this association72,73 but serious methodological questions have been raised pertaining to these studies including the fact one of these studies compared 5 year cumulative concurrency rates from European countries with point prevalence of concurrency in African countries39,72,74. A further problem related to these cross-national comparisons is that national populations are frequently composed of multiple subpopulations that may have large differences in HIV prevalence. In 29 sub-Saharan countries with available data, for example, HIV prevalence was found to vary by a median of 3.7-fold (IQR 2.9-7.9) between regions within countries65. As argued above, more fine-grained studies investigating the correlates of HIV prevalence by ethnic group or region within these and other countries have found a range of markers of network connectivity (such as partner number and concurrency) and other risk factors to be associated with HIV prevalence12,34,61,62.\n\nc. Men who have sex with men (MSM) vs. heterosexual comparison\n\nA number of high-income countries are experiencing epidemics of a range of STIs that are disproportionately affecting (MSM)75. In the year 2014 in London, for example, MSM who comprised only 2% of the population, contributed a disproportionate number of diagnoses of STIs (23%/63%/69%/90% of all new chlamydia/HIV/gonorrhoea/syphilis, respectively)76. Lymphogranuloma venereum (LGV) and sexually transmitted hepatitis C have also been noted to disproportionately affect MSM in contemporary outbreaks in high-income countries75,77. A parsimonious explanation for this clustering of STIs in MSM is a combination of behavioural factors including number of sex partners and partner concurrency77,78. In the United Kingdom, MSM in 2012 reported a mean of 24 partners in the past year versus 3.8 reported by heterosexual men79; likewise, the proportions reporting concurrency in the prior 5 years were 52% and 15% for these two groups, respectively79.\n\n\nMarkers of network connectivity are correlated with the incidence of STIs: longitudinal evidence\n\nIn this section, we consider two (of many possible) examples where large changes in STI incidence are preceded by corresponding changes in network connectivity.\n\na) Incidence of primary/secondary syphilis in MSM in the USA 1963 to 2013.\n\nIn their review of syphilis epidemiology in the United States 1963 to 2013, Peterman et al., found evidence of an initial dramatic increase in primary/secondary syphilis in MSM between 1963 and 1982 followed by a steep decline to close to zero cases in 1994 and a subsequent increase to 228/100,000 in 2013 (Figure 4)80. Increases in multiple partnering were thought to underpin the initial increase80–83. The AIDS epidemic in the 1980s led to reductions in network connectivity via both behaviour change and deaths of individuals (from AIDS) centrally placed in sexual networks80,84. The arrival of effective antiretroviral therapy from 1996 onwards played an important role in the increases in rates of partner change and reductions in condom usage80,85 noted during the ongoing epidemic of syphilis in this population. Although this evidence is indirect and susceptible to confounding, it is at least suggestive that the two large increases and one precipitous decline in syphilis incidence were determined to some extent by corresponding changes in network connectivity. Of note, this epidemic trajectory of syphilis in MSM in the United States was similar to that of a range of other STIs such as LGV and gonorrhoea in this same population and in MSM in other high income countries75. In a range of European countries where MSM were similarly affected by the AIDS epidemic, the incidence of STIs such as syphilis, LGV and gonorrhoea declined to very low rates in the post AIDS period before large increases in the late 1990s onwards corresponding to increases in partner number and declines in condom usage75.\n\nAlso shown is how these were temporally associated with changes in network connectivity represented schematically.\n\nb) Southern and Eastern Africa\n\nA number of studies from general populations in Southern and Eastern Africa have concluded that reductions in partner number and concurrency played an important role in the impressive declines in HIV incidence in Uganda, Zimbabwe and other countries in the region86–91. Delayed sexual debut, increased condom usage, enhanced antiretroviral therapy coverage and AIDS mortality (via reduced network connectivity) also played an important role in this regard87,91,92.\n\n\nClustering of STIs includes incurable STIs and network connectivity is the most parsimonious way to explain this clustering\n\nWe have already noted the striking clustering of STIs within certain ethnic groups and sexual orientations in a number of countries. Strong evidence of clustering of STIs has also been found at WHO world regional93 and country levels. At a country level, the peak HIV prevalence has been found to be associated with the prevalence of a range of STIs before/early in the HIV epidemics: syphilis94, gonorrhoea95, HSV-294 and trichomoniasis95 and BV95.\n\nThis clustering of STIs is important for two reasons. Firstly, it suggests that one or more common risk factors could underpin variations in all these STIs. Secondly, the incurable STI, HSV-2 is correlated with both peak HIV prevalence94 and antenatal syphilis prevalence from the pre-HIV period62. This is relevant because differential STI treatment efficacy can explain differences in the prevalence of treatable STIs such as syphilis but not HSV-2. Differential network connectivity, which can explain the differential spread of all STIs, is thus a more parsimonious way to explain the clustering of STIs.\n\n\nNetwork connectivity is also a risk factor for BV\n\nRecent couple studies have demonstrated that the consortia of bacteria that constitute the dysbiosis, BV are sexually transmitted96–102. BV is strongly associated with number of sexual partners and reporting partner concurrency at an individual level15,25,103,104. An ecological study found associations at the level of countries between the prevalence of male concurrency and the prevalence of BV26. The same association was also found at the level of ethnic groups within countries26.\n\nPopulations with high network connectivity are thus likely to have a high prevalence of BV which in addition to the adverse clinical effects of BV105, is important because of the positive feedback cycles between BV and the traditional STIs. BV for example has been shown to enhance susceptibility to chlamydia106,107, gonorrhoea106, HIV108,109, HSV-2110,111 and TV106,107,112. HSV-2 and TV have in turn been shown to increase the risk for acquisition of BV and other STIs106–108,113. A more detailed review of the evidence linking BV prevalence to network connectivity is provided in 114.\n\nPrevious modelling studies have found that relatively small increases in parameters of network connectivity can lead to non-linear increases in HIV/STI spread13. If this applies to BV as well, then more connected sexual networks would be expected to facilitate the rapid spread of BV and the various STIs soon after sexual debut. These would then increase susceptibility and transmission of other STIs, adding a further means by which enhanced network connectivity could lead to increases in STI spread. Network connectivity would thus indirectly enhance probability of transmission per sex act for different STIs (Figure 1).\n\n\nLimitations\n\nIt should be emphasized that this paper presents a narrative, non-systematic review of evidence for network connectivity as a parsimonious explanation of variations in genital microbiomes and STI prevalence. As such, our sampling of evidence is likely biased. We acknowledge that we have picked evidence that is supportive of our hypothesis. Our definition of network connectivity could also be criticized as being impractical because it includes such a breadth of structural and conductivity variables. Consequently, in our conceptual framework of network connectivity, different combinations of these variables could yield the same STI prevalence.\n\nConsiderable further work is necessary to construct formulae of the determinants of network connectivity and then establish how these relate to empirical estimates of STI prevalence around the world. A global study that uses a standardized methodology (Table 2) to map the variations in STI prevalence and associated risk factors by ethnic group/region within all relevant countries could provide valuable further information. So too, longitudinal studies that follow up populations from high and low STI prevalence populations from the time of sexual debut would be useful. These should accurately map the timing and correlates of STI spread including alterations of vaginal and penile microbiomes and allow more precise quantitation of which risk factors are most important for STI spread. These studies should enable the construction of more accurate models of STI spread that can be used to predict STI prevalence for specific populations under various counterfactual scenarios such as reductions in the prevalence of concurrent partnering.\n\n\nImplications of network connectivity: Know Your Network, Determine Your Prevalence\n\nIf confirmed by further experimental data, the network connectivity approach would generate new opportunities for STI prevention interventions. Whilst individual level biomedical STI control interventions have delivered considerable successes, they do not address the root cause of high STI prevalence and are therefore unlikely to accomplish radical prevention115,116. HIV pre-exposure prophylaxis and treatment as prevention, for example, may reduce HIV transmission but will not reduce the transmission of other STIs. If differential network connectivity is a fundamental determinant of STI and BV prevalence then this could be communicated to affected populations as an opportunity to effect radical prevention. Along these lines, a ‘Know your Network’ intervention has been successfully piloted in Kenya117. During a community meeting, the community's sexual network was computed by fitting a dynamic network model to data from individual sexual diaries, and a graphical representation of the network was fed back to the community. Participants reported the intervention to be transformative but formal trials are required to assess the efficacy on STI incidence of this type of intervention117. Uganda’s ‘Zero Grazing’ campaign118 and similar processes elsewhere in Africa91,119 which resulted in dramatic declines in side-partners and HIV incidence, could be viewed as providing both guidance and evidence for this approach. In areas where social, economic and demographic factors are thought to be strong upstream determinants of network connectivity, interventions should be directed at these21,120.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in funding this work.\n\n\nAcknowledgements\n\nWe would like to thank Leigh Johnson for helpful comments on an earlier draft of the paper and discussants at a South African Centre for Epidemiological Modelling and Analysis seminar where this hypothesis was presented.\n\n\nReferences\n\nGray R, Kigozi G, Kong X, et al.: The effectiveness of male circumcision for HIV prevention and effects on risk behaviors in a posttrial follow-up study. AIDS. 2012; 26(5): 609–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChesson HW, Mayaud P, Aral SO, et al.: Sexually Transmitted Infections: Impact and Cost-Effectiveness of Prevention. Disease Control Priorities. 2017; 6: 203–232. Publisher Full Text\n\nDallabetta G, Wi T, Nielsen G, et al.: Prevention and control of STD and HIV infection in developing countries. Sexually transmitted diseases 4th edn New York, NY: McGraw-Hill. 2008.\n\nGewirtzman A, Bobrick L, Conner K, et al.: Epidemiology of Sexually Transmitted Infections. In: Gross G, Tyring SK, editors. Sexually transmitted infections and sexually transmitted diseases. Heidelbergh: Springer Verlag. 2011; 13–34. Publisher Full Text\n\nHolmqvist G: HIV and Income Inequality: If there is a link, what does it tell us? Working Paper, International Policy Centre for Inclusive Growth, 2009. Reference Source\n\nKenyon CR, Buyze J: No association between gender inequality and peak HIV prevalence in developing countries - an ecological study. AIDS Care. 2015; 27(2): 150–9. PubMed Abstract | Publisher Full Text\n\nVoeten HA, Vissers DC, Gregson S, et al.: Strong association between in-migration and HIV prevalence in urban sub-Saharan Africa. Sex Transm Dis. 2010; 37(4): 240–3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorrison CS, Chen PL, Kwok C, et al.: Hormonal contraception and the risk of HIV acquisition: an individual participant data meta-analysis. PLoS Med. 2015; 12(1): e1001778. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBuve A, Jespers V, Crucitti T, et al.: The vaginal microbiota and susceptibility to HIV. AIDS. 2014; 28(16): 2333–44. PubMed Abstract | Publisher Full Text\n\nKaul R, Cohen CR, Chege D, et al.: Biological factors that may contribute to regional and racial disparities in HIV prevalence. Am J Reprod Immunol. 2011; 65(3): 317–24. PubMed Abstract | Publisher Full Text\n\nDrain PK, Smith JS, Hughes JP, et al.: Correlates of national HIV seroprevalence: an ecologic analysis of 122 developing countries. J Acquir Immune Defic Syndr. 2004; 35(4): 407–20. PubMed Abstract | Publisher Full Text\n\nMorris M, Kurth AE, Hamilton DT, et al.: Concurrent partnerships and HIV prevalence disparities by race: linking science and public health practice. Am J Public Health. 2009; 99(6): 1023–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorris M, Kretzschmar M: Concurrent partnerships and the spread of HIV. AIDS. 1997; 11(5): 641–8. PubMed Abstract\n\nMorris M, Goodreau S, Moody J: Sexual networks, concurrency and STD/HIV. In: Holmes KK, editor. Sexually transmitted diseases. 4th ed. New York: McGraw-Hill Medical; 2008; 2166.\n\nKenyon CR, Buyze J, Klebanoff M, et al.: Association between bacterial vaginosis and partner concurrency: a longitudinal study. Sex Transm Infect. 2018; 94(1): 75–77. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKenyon CR, Buyze J, Klebanoff M, et al.: Incident Trichomonas vaginalis Is Associated With Partnership Concurrency: A Longitudinal Cohort Study. Sex Transm Dis. 2017; 44(11): 695–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKeeling MJ, Eames KT: Networks and epidemic models. J R Soc Interface. 2005; 2(4): 295–307. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWatts CH, May RM: The influence of concurrent partnerships on the dynamics of HIV/AIDS. Math Biosci. 1992; 108(1): 89–104. PubMed Abstract | Publisher Full Text\n\nPotterat JJ, Zimmerman-Rogers H, Muth SQ, et al.: Chlamydia transmission: concurrency, reproduction number, and the epidemic trajectory. Am J Epidemiol. 1999; 150(12): 1331–9. PubMed Abstract | Publisher Full Text\n\nLaumann EO, Youm Y: Racial/ethnic group differences in the prevalence of sexually transmitted diseases in the United States: a network explanation. Sex Transm Dis. 1999; 26(5): 250–61. PubMed Abstract | Publisher Full Text\n\nKraut-Becher J, Eisenberg M, Voytek C, et al.: Examining racial disparities in HIV: lessons from sexually transmitted infections research. J Acquir Immune Defic Syndr. 2008; 47 Suppl 1: S20–7. PubMed Abstract\n\nArmbruster B, Wang L, Morris M: Forward reachable sets: Analytically derived properties of connected components for dynamic networks. Netw Sci. 2017; 5(3): 328–54. Publisher Full Text\n\nHethcote HW, Yorke JA: Lecture notes in biomathematics: gonorrhea transmission and control. Levin S, editor. Berlin: Springer-Verlag; 1984.\n\nYorke JA, Hethcote HW, Nold A: Dynamics and control of the transmission of gonorrhea. Sex Transm Dis. 1978; 5(2): 51–6. PubMed Abstract | Publisher Full Text\n\nKoumans EH, Sternberg M, Bruce C, et al.: The prevalence of bacterial vaginosis in the United States, 2001-2004; associations with symptoms, sexual behaviors, and reproductive health. Sex Transm Dis. 2007; 34(11): 864–9. PubMed Abstract | Publisher Full Text\n\nKenyon CR, Colebunders R: Strong association between the prevalence of bacterial vaginosis and male point-concurrency. Eur J Obstet Gynecol Reprod Biol. 2014; 172: 93–6. PubMed Abstract | Publisher Full Text\n\nBuvé A, Weiss HA, Laga M, et al.: The epidemiology of gonorrhoea, chlamydial infection and syphilis in four African cities. AIDS. 2001; 15 Suppl 4: S79–88. PubMed Abstract | Publisher Full Text\n\nFenton KA, Korovessis C, Johnson AM, et al.: Sexual behaviour in Britain: reported sexually transmitted infections and prevalent genital Chlamydia trachomatis infection. Lancet. 2001; 358(9296): 1851–4. PubMed Abstract | Publisher Full Text\n\nCanchihuaman FA, Carcamo CP, Garcia PJ, et al.: Non-monogamy and risk of infection with Chlamydia trachomatis and Trichomonas vaginalis among young adults and their cohabiting partners in Peru. Sex Transm Dis. 2010; 86 Suppl 3: iii37–iii44. PubMed Abstract | Publisher Full Text\n\nGarnett GP, Mertz KJ, Finelli L, et al.: The transmission dynamics of gonorrhoea: modelling the reported behaviour of infected patients from Newark, New Jersey. Philos Trans R Soc Lond B Biol Sci. 1999; 354(1384): 787–97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGhani AC, Swinton J, Garnett GP: The role of sexual partnership networks in the epidemiology of gonorrhea. Sex Transm Dis. 1997; 24(1): 45–56. PubMed Abstract | Publisher Full Text\n\nMishra V, Medley A, Hong R, et al.: Levels and spread of HIV seroprevalence and associated factors: evidence from national household surveys. 2009. Reference Source\n\nAuvert B, Buvé A, Ferry B, et al.: Ecological and individual level analysis of risk factors for HIV infection in four urban populations in sub-Saharan Africa with different levels of HIV infection. AIDS. 2001; 15 Suppl 4: S15–30. PubMed Abstract | Publisher Full Text\n\nKenyon CR, Tsoumanis A, Schwartz IS: HIV Prevalence Correlates with High-Risk Sexual Behavior in Ethiopia's Regions. PLoS One. 2015; 10(10): e0140835. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKenyon CR, Tsoumanis A, Schwartz IS: A population's higher-risk sexual behaviour is associated with its average sexual behaviour-An ecological analysis of subpopulations in Ethiopia, Kenya, South Africa, Uganda and the United States. Epidemics. 2016; 15: 56–65. PubMed Abstract | Publisher Full Text\n\nHelleringer S, Kohler HP, Kalilani-Phiri L: The association of HIV serodiscordance and partnership concurrency in Likoma Island (Malawi). AIDS. 2009; 23(10): 1285–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKenyon CR, Tsoumanis A, Schwartz IS, et al.: Partner concurrency and HIV infection risk in South Africa. Int J Infect Dis. 2016; 45: 81–7. PubMed Abstract | Publisher Full Text\n\nAdimora AA, Schoenbach VJ, Martinson FE, et al.: Heterosexually transmitted HIV infection among African Americans in North Carolina. J Acquir Immune Defic Syndr. 2006; 41(5): 616–23. PubMed Abstract | Publisher Full Text\n\nKenyon C, Colebunders R: Strong association between point-concurrency and national peak HIV prevalence. Int J Infect Dis. 2012; 16(11): e826–7. PubMed Abstract | Publisher Full Text\n\nBohl DD, Raymond HF, Arnold M, et al.: Concurrent sexual partnerships and racial disparities in HIV infection among men who have sex with men. Sex Transm Infect. 2009; 85(5): 367–9. PubMed Abstract | Publisher Full Text\n\nAdimora AA, Schoenbach VJ, Bonas DM, et al.: Concurrent sexual partnerships among women in the United States. Epidemiology. 2002; 13(3): 320–7. PubMed Abstract | Publisher Full Text\n\nXu F, Sternberg MR, Kottiri BJ, et al.: Trends in herpes simplex virus type 1 and type 2 seroprevalence in the United States. JAMA. 2006; 296(8): 964–73. PubMed Abstract | Publisher Full Text\n\nKenyon C, Colebunders R, Buve A, et al.: Partner-concurrency associated with herpes simplex virus 2 infection in young South Africans. Int J STD AIDS. 2013; 24(10): 804–12. PubMed Abstract | Publisher Full Text\n\nKenyon C, Colebunders R, Hens N: Determinants of generalized herpes simplex virus-2 epidemics: the role of sexual partner concurrency. Int J STD AIDS. 2013; 24(5): 375–82. PubMed Abstract | Publisher Full Text\n\nShi R, Devarakonda S, Liu L, et al.: Factors associated with genital human papillomavirus infection among adult females in the United States, NHANES 2007-2010. BMC Res Notes. 2014; 7: 544. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSonnenberg P, Clifton S, Beddows S, et al.: Prevalence, risk factors, and uptake of interventions for sexually transmitted infections in Britain: findings from the National Surveys of Sexual Attitudes and Lifestyles (Natsal). Lancet. 2013; 382(9907): 1795–806. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJavanbakht M, Gorbach PM, Amani B, et al.: Concurrency, sex partner risk, and high-risk human papillomavirus infection among African American, Asian, and Hispanic women. Sex Transm Dis. 2010; 37(2): 68–74. PubMed Abstract | Publisher Full Text\n\nKoumans EH, Farley TA, Gibson JJ, et al.: Characteristics of persons with syphilis in areas of persisting syphilis in the United States: sustained transmission associated with concurrent partnerships. Sex Transm Dis. 2001; 28(9): 497–503. PubMed Abstract | Publisher Full Text\n\nKenyon CR, Osbak K: The Prevalence of Syphilis Is Associated with the Prevalence of Male Point-Concurrency: An Ecological Analysis. World J AIDS. 2015; 5(02): 57363. Publisher Full Text\n\nRogers SM, Turner CF, Hobbs M, et al.: Epidemiology of undiagnosed trichomoniasis in a probability sample of urban young adults. PLoS One. 2014; 9(3): e90548. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKenyon CR, Hamilton DT: Correlation between Trichomonas vaginalis and Concurrency: An Ecological Study. Interdiscip Perspect Infect Dis. 2016; 2016: 5052802. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKoumans EH, Sternberg M, Bruce C, et al.: The prevalence of bacterial vaginosis in the United States, 2001-2004; associations with symptoms, sexual behaviors, and reproductive health. Sex Transm Dis. 2007; 34(11): 864–9. PubMed Abstract | Publisher Full Text\n\nLeiby GM, Turner TB, Clark GE, et al.: Studies on syphilis in the Eastern Health District of Baltimore city; syphilis among parturient women as an index of the trend of syphilis in the community. Am J Hyg. 1947; 46(2): 260–7. PubMed Abstract | Publisher Full Text\n\nKenyon CR, Osbak K, Tsoumanis A: The Global Epidemiology of Syphilis in the Past Century - A Systematic Review Based on Antenatal Syphilis Prevalence. PLoS Negl Trop Dis. 2016; 10(5): e0004711. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFleming DT, McQuillan GM, Johnson RE, et al.: Herpes simplex virus type 2 in the United States, 1976 to 1994. N Engl J Med. 1997; 337(16): 1105–11. PubMed Abstract | Publisher Full Text\n\nGoodreau SM, Cassels S, Kasprzyk D, et al.: Concurrent partnerships, acute infection and HIV epidemic dynamics among young adults in Zimbabwe. AIDS Behav. 2012; 16(2): 312–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKenyon C, Buyze J, Colebunders R: HIV prevalence by race co-varies closely with concurrency and number of sex partners in South Africa. PLoS One. 2013; 8(5): e64080. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJohnson LF, Dorrington RE, Bradshaw D, et al.: Sexual behaviour patterns in South Africa and their association with the spread of HIV: insights from a mathematical model. Demogr Res. 2009; 21: 289–340. Publisher Full Text\n\nPaz-Bailey G, Morales-Miranda S, Jacobson JO, et al.: High rates of STD and sexual risk behaviors among Garífunas in Honduras. J Acquir Immune Defic Syndr. 2009; 51 Suppl 1: S26–34. PubMed Abstract | Publisher Full Text\n\nKenyon CR, Vu L, Menten J, et al.: Male circumcision and sexual risk behaviors may contribute to considerable ethnic disparities in HIV prevalence in Kenya: an ecological analysis. PLoS One. 2014; 9(8): e106230. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKenyon CR: HIV prevalence by ethnic group covaries with prevalence of herpes simplex virus-2 and high-risk sex in Uganda: An ecological study. PLoS One. 2018; 13(4): e0195431. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKenyon CR, Osbak K, Buyze J: The prevalence of HIV by ethnic group is correlated with HSV-2 and syphilis prevalence in Kenya, South Africa, the United Kingdom and the United States. Interdiscip Perspect Infect Dis 2014: 2014: 284317. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFenton KA, Mercer CH, McManus S, et al.: Ethnic variations in sexual behaviour in Great Britain and risk of sexually transmitted infections: a probability survey. Lancet. 2005; 365(9466): 1246–55. PubMed Abstract | Publisher Full Text\n\nTurner KM, Garnett GP, Ghani AC, et al.: Investigating ethnic inequalities in the incidence of sexually transmitted infections: mathematical modelling study. Sex Transm Infect. 2004; 80(5): 379–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKenyon C, Buyze J, Schwartz IS: Strong association between higher risk sex and HIV prevalence at regional level: an ecological study of 27 sub-Saharan African countries (In Submission).\n\nKenyon C, Colebunders R, Voeten H, et al.: Peak HIV prevalence: a useful outcome variable for ecological studies. Int J Infect Dis. 2013; 17(5): e286–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKenyon C, Colebunders R, Voeten H, et al.: Migration intensity has no effect on peak HIV prevalence: an ecological study. BMC Infect Dis. 2014; 14(1): 350. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKenyon C, Buyze J: Correlation between National Peak HIV Prevalence and Concurrency, STI Treatment Capability, Condom Use and Circumcision: An Ecological Study. World J AIDS. 2014; 4(2): 47249. Publisher Full Text\n\nMishra V, Assche SB, Greener R, et al.: HIV infection does not disproportionately affect the poorer in sub-Saharan Africa. AIDS. 2007; 21 Suppl 7: S17–28. PubMed Abstract | Publisher Full Text\n\nDrain PK, Halperin DT, Hughes JP, et al.: Male circumcision, religion, and infectious diseases: an ecologic analysis of 118 developing countries. BMC Infect Dis. 2006; 6: 172. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKenyon C: Ecological association between HIV and concurrency point-prevalence in South Africa's ethnic groups. Afr J AIDS Res. 2013; 12(2): 79–84. PubMed Abstract | Publisher Full Text\n\nSawers L, Stillwaggon E: Concurrent sexual partnerships do not explain the HIV epidemics in Africa: a systematic review of the evidence. J Int AIDS Soc. 2010; 13: 34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLagarde E, Auvert B, Caraël M, et al.: Concurrent sexual partnerships and HIV prevalence in five urban communities of sub-Saharan Africa. AIDS. 2001; 15(7): 877–84. PubMed Abstract | Publisher Full Text\n\nMorris M: Barking up the wrong evidence tree. Comment on Lurie & Rosenthal, \"Concurrent partnerships as a driver of the HIV epidemic in sub-Saharan Africa? The evidence is limited\". AIDS Behav. 2010; 14(1): 31–3; discussion 34-7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFenton KA, Imrie J: Increasing rates of sexually transmitted diseases in homosexual men in Western europe and the United States: why? Infect Dis Clin North Am. 2005; 19(2): 311–31. PubMed Abstract | Publisher Full Text\n\nPublic Health England: Inequalities in sexual health: Update on HIV and STIs in men who have sex with men in London. London: 2016. Reference Source\n\nFenton KA, Wasserheit JN: The courage to learn from our failures: syphilis control in men who have sex with men. Sex Transm Dis. 2007; 34(3): 162–5. PubMed Abstract | Publisher Full Text\n\nKenyon C, Osbak K: Certain attributes of the sexual ecosystem of high-risk MSM have resulted in an altered microbiome with an enhanced propensity to generate and transmit antibiotic resistance. Med Hypotheses. 2014; 83(2): 196–202. PubMed Abstract | Publisher Full Text\n\nMercer CH, Prah P, Field N, et al.: The health and well-being of men who have sex with men (MSM) in Britain: Evidence from the third National Survey of Sexual Attitudes and Lifestyles (Natsal-3). BMC Public Health. 2016; 16: 525. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPeterman TA, Su J, Bernstein KT, et al.: Syphilis in the United States: on the rise? Expert Rev Anti Infect Ther. 2015; 13(2): 161–8. PubMed Abstract | Publisher Full Text\n\nStenger MR, Baral S, Stahlman S, et al.: As through a glass, darkly: the future of sexually transmissible infections among gay, bisexual and other men who have sex with men. Sex Health. 2017; 14(1): 18–27. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAral SO, Fenton KA, Holmes KK: Sexually transmitted diseases in the USA: temporal trends. Sex Transm Infect. 2007; 83(4): 257–66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSandholzer TA: Factors Affecting the Incidence and Management of Sexually Transmitted Diseases in Homosexual Men. Sexually Transmitted Diseases in Homosexual Men: Springer; 1983; 3–12. Publisher Full Text\n\nChesson HW, Dee TS, Aral SO: AIDS mortality may have contributed to the decline in syphilis rates in the United States in the 1990s. Sex Transm Dis. 2003; 30(5): 419–24. PubMed Abstract | Publisher Full Text\n\nCrepaz N, Hart TA, Marks G: Highly active antiretroviral therapy and sexual risk behavior: a meta-analytic review. JAMA. 2004; 292(2): 224–36. PubMed Abstract | Publisher Full Text\n\nStoneburner RL, Low-Beer D: Sexual partner reductions explain human immunodeficiency virus declines in Uganda: comparative analyses of HIV and behavioural data in Uganda, Kenya, Malawi, and Zambia. Int J Epidemiol. 2004; 33(3): 624. PubMed Abstract | Publisher Full Text\n\nGregson S, Garnett GP, Nyamukapa CA, et al.: HIV decline associated with behavior change in eastern Zimbabwe. Science. 2006; 311(5761): 664–6. PubMed Abstract | Publisher Full Text\n\nStoneburner RL, Low-Beer D: Population-level HIV declines and behavioral risk avoidance in Uganda. Science. 2004; 304(5671): 714–8. PubMed Abstract | Publisher Full Text\n\nCheluget B, Baltazar G, Orege P, et al.: Evidence for population level declines in adult HIV prevalence in Kenya. Sex Transm Infect. 2006; 82 Suppl 1: i21–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEaton JW, Takavarasha FR, Schumacher CM, et al.: Trends in concurrency, polygyny, and multiple sex partnerships during a decade of declining HIV prevalence in eastern Zimbabwe. J Infect Dis. 2014; 210 Suppl 2: S562–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHalperin DT, Mugurungi O, Hallett TB, et al.: A surprising prevention success: why did the HIV epidemic decline in Zimbabwe? PLoS Med. 2011; 8(2): e1000414. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJohnson LF, Dorrington RE, Bradshaw D, et al.: The effect of syndromic management interventions on the prevalence of sexually transmitted infections in South Africa. Sex Reprod Healthc. 2011; 2(1): 13–20. PubMed Abstract | Publisher Full Text\n\nKenyon C, Buyze J, Colebunders R: Classification of incidence and prevalence of certain sexually transmitted infections by world regions. Int J of Infec Dis. 2014; 18: 73–80. PubMed Abstract | Publisher Full Text\n\nOsbak KK, Rowley JT, Kassebaum NJ, et al.: The Prevalence of Syphilis from the Early HIV Period is Correlated With Peak HIV Prevalence at a Country Level. Sex Transm Dis. 2016; 43(4): 255–7. PubMed Abstract | Publisher Full Text\n\nKenyon C: Strong associations between national prevalence of various STIs suggests sexual network connectivity is a common underpinning risk factor. BMC Infect Dis. 2017; 17(1): 682. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu CM, Hungate BA, Tobian AA, et al.: Penile Microbiota and Female Partner Bacterial Vaginosis in Rakai, Uganda. MBio. 2015; 6(3): e00589. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBalashov SV, Mordechai E, Adelson ME, et al.: Identification, quantification and subtyping of Gardnerella vaginalis in noncultured clinical vaginal samples by quantitative PCR. J Med Microbiol. 2014; 63(Pt 2): 162–75. PubMed Abstract | Publisher Full Text\n\nSwidsinski A, Doerffel Y, Loening-Baucke V, et al.: Gardnerella biofilm involves females and males and is transmitted sexually. Gynecol Obstet Invest. 2010; 70(4): 256–63. PubMed Abstract | Publisher Full Text\n\nGardner HL, Dukes CD: Haemophilus vaginalis vaginitis: a newly defined specific infection previously classified non-specific vaginitis. Am J Obstet Gynecol. 1955; 69(5): 962–76. PubMed Abstract | Publisher Full Text\n\nZozaya M, Ferris MJ, Siren JD, et al.: Bacterial communities in penile skin, male urethra, and vaginas of heterosexual couples with and without bacterial vaginosis. Microbiome. 2016; 4: 16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNelson DE, Dong Q, Van der Pol B, et al.: Bacterial communities of the coronal sulcus and distal urethra of adolescent males. PLoS One. 2012; 7(5): e36298. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNelson DE, Van Der Pol B, Dong Q, et al.: Characteristic male urine microbiomes associate with asymptomatic sexually transmitted infection. PLoS One. 2010; 5(11): e14116. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFethers KA, Fairley CK, Hocking JS, et al.: Sexual risk factors and bacterial vaginosis: a systematic review and meta-analysis. Clin Infect Dis. 2008; 47(11): 1426–35. PubMed Abstract | Publisher Full Text\n\nFethers KA, Fairley CK, Morton A, et al.: Early sexual experiences and risk factors for bacterial vaginosis. J Infect Dis. 2009; 200(11): 1662–70. PubMed Abstract | Publisher Full Text\n\nBautista CT, Wurapa E, Sateren WB, et al.: Bacterial vaginosis: a synthesis of the literature on etiology, prevalence, risk factors, and relationship with chlamydia and gonorrhea infections. Mil Med Res. 2016; 3: 4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrotman RM, Klebanoff MA, Nansel TR, et al.: Bacterial vaginosis assessed by gram stain and diminished colonization resistance to incident gonococcal, chlamydial, and trichomonal genital infection. J Infect Dis. 2010; 202(12): 1907–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaul R, Nagelkerke NJ, Kimani J, et al.: Prevalent herpes simplex virus type 2 infection is associated with altered vaginal flora and an increased susceptibility to multiple sexually transmitted infections. J Infect Dis. 2007; 196(11): 1692–7. PubMed Abstract | Publisher Full Text\n\nRottingen JA, Cameron DW, Garnett GP: A systematic review of the epidemiologic interactions between classic sexually transmitted diseases and HIV: how much really is known? Sex Transm Dis. 2001; 28(10): 579–97. PubMed Abstract | Publisher Full Text\n\nWard H, Ronn M: Contribution of sexually transmitted infections to the sexual transmission of HIV. Curr Opin HIV AIDS. 2010; 5(4): 305–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCherpes TL, Meyn LA, Krohn MA, et al.: Risk factors for infection with herpes simplex virus type 2: role of smoking, douching, uncircumcised males, and vaginal flora. Sex Transm Dis. 2003; 30(5): 405–10. PubMed Abstract | Publisher Full Text\n\nCherpes TL, Meyn LA, Krohn MA, et al.: Association between acquisition of herpes simplex virus type 2 in women and bacterial vaginosis. Clin Infect Dis. 2003; 37(3): 319–25. PubMed Abstract | Publisher Full Text\n\nRathod SD, Krupp K, Klausner JD, et al.: Bacterial vaginosis and risk for Trichomonas vaginalis infection: a longitudinal analysis. Sex Transm Dis. 2011; 38(9): 882–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEsber A, Vicetti Miguel RD, Cherpes TL, et al.: Risk of Bacterial Vaginosis Among Women With Herpes Simplex Virus Type 2 Infection: A Systematic Review and Meta-analysis. J Infect Dis. 2015; 212(1): 8–17. PubMed Abstract | Publisher Full Text\n\nKenyon CDW, Brotman R: Differential sexual network connectivity offers a parsimonious explanation for population-level variations in the prevalence of bacterial vaginosis: a hypothesis. BMC Womens Health. In Submission.\n\nRose G, Day S: The population mean predicts the number of deviant individuals. BMJ. 1990; 301(6759): 1031–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAral SO, Leichliter JS, Blanchard JF: Overview: the role of emergent properties of complex systems in the epidemiology and prevention of sexually transmitted infections including HIV infection. Sex Transm Infect. 2010; 86 Suppl 3: iii1–3. PubMed Abstract | Publisher Full Text\n\nKnopf A, Agot K, Sidle J, et al.: \"This is the medicine:\" A Kenyan community responds to a sexual concurrency reduction intervention. Soc Sci Med. 2014; 108: 175–84. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKirby D: Changes in sexual behaviour leading to the decline in the prevalence of HIV in Uganda: confirmation from multiple sources of evidence. Sex Transm Infect. 2008; 84 Suppl 2: ii35–ii41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShisana O, Rehle, T, Simbayi LC, et al.: South African National HIV Prevalence, Incidence and Behaviour Survey, 2012. Cape Town: HSRC Press; 2014. Reference Source\n\nAral SO, Lipshutz J, Blanchard J: Drivers of STD/HIV epidemiology and the timing and targets of STD/HIV prevention. Sex Transm Infect. 2007; 83 Suppl 1: i1–4. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "41827",
"date": "03 Jan 2019",
"name": "Ann M. Jolly",
"expertise": [
"Reviewer Expertise Social networks",
"infectious disease",
"epidemiology",
"sexually transmitted infections",
"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 is a well referenced, very interesting paper which attempts and in large part succeeds, in critically examining, and in many cases rejecting common “risk factors/ determinants” of high STI (sexually transmitted infection) rates. The authors have traced a long history in mathematical modelling of primarily HIV and summarised the most common and consistently found explanations of “peak HIV” rates at an ecological level.\n\nI think this article is thought provoking and even if one is not entirely convinced by it, it adds much deeper and more critical thought to assumptions of mathematical modellers and epidemiologists about the immediate causes of HIV than I have ever seen. I am glad it is published.\n\nI think the paper could be improved in general by addressing two major areas. First, some mention of the empirical study of sexual networks should be made.There are examples from all over the world which the authors may use to strengthen their ultimate hypothesis; that sexual interactions between sexual network members in which some people are infected with STI are the primary and necessary conditions for STI to transmit 1,2. This will help define for their readers very clearly the fact that because ecological findings focus on whole systems, whether national or regional, these studies are more indicative of large networks and should not be regarded as “weaker” evidence as they are in epidemiology. The authors can also argue that the ultimate risk factor on an individual level for acquiring gonorrhoea for example is having unprotected sex with another person with gonorrhea; that is it and that is all. Other factors are moderators of that plain risk; such as use of condoms, cumulative frequency of intercourse, etc. Concurrency or number of sex partners are indirect indicators of the risk itself; because transmission of STI is an interaction which occurs between at minimum two people, the other “factors” are reduced to characteristics or behaviors of individuals within the population, which again are proxies, though more proximal than age, sex, ethnic background or income.\n\nThe second thing I would be very clear on is the part played by ethnic group. This is again just a proxy for sexual interactions and cultural norms, and may not be a consistent nor accurate marker. However, it is very easy in the discussion of HIV in Africa to “racialize” the epidemic and/or behavior. For example, the effect of the labour policies under apartheid South Africa, where black labourers were forced into a pattern of migrant labor for a year at a time from their rural homes to large urban or mining compounds did much to accelerate sexual interactions which in turn exacerbated HIV spread. Likewise, sexual violence in some parts of Africa is extremely high, and while it is not mentioned here specifically, certainly does contribute to transmission. However, this also is linked to harsh colonial conditions and the authors could make it very clear that this should not be interpreted as being a intrinsically ethnic or cultural characteristic. The successful Ugandan educational intervention is a great demonstration of that. One of the best network explanations of different STI rates in people of different ethnic backgrounds can be found in 3, pages 689-697.\n\nMore minor comments follow;\n\nPage 2 - I love Figure 1.\nPage 3, 2nd paragraph - better give credit where credit is due 4.\n\nPage 3, 3rd paragraph - so concurrency may be a marker of connectivity; or a determinant, but one may equally pose that if one has many sex partners the only way to fit them all in is to have concurrent ones.. and that may have an effect also on frequency of intercourse with each partner. Lots of food for thought!\n\nPage 4 - wonderful summary! I would love to see a systematic review which includes an exhaustive list with negative findings. Again, one could look for only proximal causes of transmission rather than secondary or structural determinants.\n\nPage 5 - you state that “Peak high prevalence HIV is based generally of high quality longitudinal data”. Please justify and provide references.\n\nPage 5 - the statements on poverty and socioeconomic status, gender inequality which are not consistently associated with HIV are great; this is because these determinants are only distally linked to transmission.\n\nPage 5, paragraph 8 - to describe numbers of partners more accurately I would use the median, range or IQR as the mean implies these distributions are Gaussian, but they are not. The devil is literally in \"de tail\" Also one may have to look at the penile anal penetration risk of transmission which is not exactly comparable to penile vaginal sex. But I agree in principle.\n\nPage 7, paragraph 1 - I would be more accurate here, HSV is certainly treatable, but not as easily diagnosed or treated as the bacterial STI.\n\nPage 7, Table 2 - you may find the discussion of ethnic homophily and heterogeneity in networks interesting in 1,2.\n\nGreat paper, good thinking, and great to challenge convention!\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? Yes\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": [
{
"c_id": "4371",
"date": "22 Jan 2019",
"name": "Chris Kenyon",
"role": "Author Response",
"response": "This is a well referenced, very interesting paper which attempts and in large part succeeds, in critically examining, and in many cases rejecting common “risk factors/ determinants” of high STI (sexually transmitted infection) rates. The authors have traced a long history in mathematical modelling of primarily HIV and summarised the most common and consistently found explanations of “peak HIV” rates at an ecological level. I think this article is thought provoking and even if one is not entirely convinced by it, it adds much deeper and more critical thought to assumptions of mathematical modellers and epidemiologists about the immediate causes of HIV than I have ever seen. I am glad it is published. I think the paper could be improved in general by addressing two major areas. First, some mention of the empirical study of sexual networks should be made.There are examples from all over the world which the authors may use to strengthen their ultimate hypothesis; that sexual interactions between sexual network members in which some people are infected with STI are the primary and necessary conditions for STI to transmit 1,2. This will help define for their readers very clearly the fact that because ecological findings focus on whole systems, whether national or regional, these studies are more indicative of large networks and should not be regarded as “weaker” evidence as they are in epidemiology. The authors can also argue that the ultimate risk factor on an individual level for acquiring gonorrhoea for example is having unprotected sex with another person with gonorrhea; that is it and that is all. Other factors are moderators of that plain risk; such as use of condoms, cumulative frequency of intercourse, etc. Concurrency or number of sex partners are indirect indicators of the risk itself; because transmission of STI is an interaction which occurs between at minimum two people, the other “factors” are reduced to characteristics or behaviors of individuals within the population, which again are proxies, though more proximal than age, sex, ethnic background or income. The second thing I would be very clear on is the part played by ethnic group. This is again just a proxy for sexual interactions and cultural norms, and may not be a consistent nor accurate marker. However, it is very easy in the discussion of HIV in Africa to “racialize” the epidemic and/or behavior. For example, the effect of the labour policies under apartheid South Africa, where black labourers were forced into a pattern of migrant labor for a year at a time from their rural homes to large urban or mining compounds did much to accelerate sexual interactions which in turn exacerbated HIV spread. Likewise, sexual violence in some parts of Africa is extremely high, and while it is not mentioned here specifically, certainly does contribute to transmission. However, this also is linked to harsh colonial conditions and the authors could make it very clear that this should not be interpreted as being a intrinsically ethnic or cultural characteristic. The successful Ugandan educational intervention is a great demonstration of that. One of the best network explanations of different STI rates in people of different ethnic backgrounds can be found in 3, pages 689-697. Reply:In addition to small additions to the text along the lines suggested we have also added the two references suggested and the following new paragraphs: Page 6, L28:The relationship between ethnicity, race and STI prevalenceIt is of paramount importance to emphasize that our hypothesis makes no reference to race. The hypothesis proposes that there are differences in sexual behavior between different groups of people which translate into differences in network connectivity and as a result differential STI prevalence. These groups can be defined by sexual orientation, ethnicity, social class, caste or whatever categories meaningfully segregate sexual networks. These categories are social constructs and thus vary considerably across time and place. It is our considered opinion that investigators who conduct investigations into STI epidemiology using these categories do so with sufficient sensitivity to the concerns as to how these categories are and have been used and abused. Other authors have hypothesized that biological differences between racial groups play an important role in STI epidemiology. A recent form of this argument is that ‘black populations’ are innately more likely to have bacterial-vaginosis-type vaginal microbiomes which in turn facilitates the transmission of various STIs in this population 1. We and others have argued that the evidence does not support this, and other race-based explanations of differential STI spread 2,3. As an example, we noted that ‘black populations’ with evidence of low sexual network connectivity have a very low prevalence of BV and conversely ‘white populations’ with high connectivity had a high BV prevalence 2,3.Page 10, L22:Our theory includes mention of the wide array of upstream socioeconomic and political factors that have been shown to influence the spread of STIs 4. We argue that the pathways through which these factors facilitate STI transmission is to a large extent mediated via alterations in network connectivity 5. We have not, however, gone into any detail into reviewing the evidence on which this view is based 4,6,7. Furthermore, our focus on the more downstream factors responsible for STI transmission should not detract from efforts to target the upstream determinants of enhanced STI transmission. More minor comments follow; Page 2 - I love Figure 1.Page 3, 2nd paragraph - better give credit where credit is due 4. Reply:This reference has been addedPage 3, 3rd paragraph - so concurrency may be a marker of connectivity; or a determinant, but one may equally pose that if one has many sex partners the only way to fit them all in is to have concurrent ones.. and that may have an effect also on frequency of intercourse with each partner. Lots of food for thought! Reply:Indeed increasing concurrency tends to lead to an increase in number of partners per unit time. Interestingly a number of modelling studies have shown that, in certain scenarios, increasing concurrency whilst keeping total number of partnership per unit time unchanged can still result in increases in markers of connectivity such as the forward reachable path 8,9. In the datasets we have reviewed, the two however tend to covary at population levels 10-13 and thus in our estimation a reasonable case can be made to not consider these two variables in isolation.Page 4 - wonderful summary! I would love to see a systematic review which includes an exhaustive list with negative findings. Again, one could look for only proximal causes of transmission rather than secondary or structural determinants. Reply:Hopefully someone reading this will be tempted to do this review. Page 5 - you state that “Peak high prevalence HIV is based generally of high quality longitudinal data”. Please justify and provide references. Reply:A justification for this assertion has been provided and backed up with 3 new references (Page 7, L9-14).Page 5 - the statements on poverty and socioeconomic status, gender inequality which are not consistently associated with HIV are great; this is because these determinants are only distally linked to transmission. Reply:Thank you.Page 5, paragraph 8 - to describe numbers of partners more accurately I would use the median, range or IQR as the mean implies these distributions are Gaussian, but they are not. The devil is literally in \"de tail\" Also one may have to look at the penile anal penetration risk of transmission which is not exactly comparable to penile vaginal sex. But I agree in principle. Reply:Thank you for pointing this out. We agree and have changed the figures provided to medians and interquartile ranges.Page 7, paragraph 1 - I would be more accurate here, HSV is certainly treatable, but not as easily diagnosed or treated as the bacterial STI. Reply: This has been changed as suggested.Page 7, Table 2 - you may find the discussion of ethnic homophily and heterogeneity in networks interesting in 1,2. Reply:Thank you for these references which have been added to the paper.References1. Buve A, Jespers V, Crucitti T, Fichorova RN. The vaginal microbiota and susceptibility to HIV. AIDS. 2014;28(16):2333-2344.2. Kenyon C, Osbak K. Sexual networks, HIV, race and bacterial vaginosis. AIDS. 2015;29(5):641-642.3. Kenyon CR, Delva W, Brotman RM. Differential sexual network connectivity offers a parsimonious explanation for population-level variations in the prevalence of bacterial vaginosis: a data-driven, model-supported hypothesis. BMC Womens Health. 2019;19(1):8.4. Aral SO. Determinants of STD epidemics: implications for phase appropriate intervention strategies. Sex Transm Infect. 2002;78 Suppl 1:i3-13.5. Gorbach PM, Stoner BP, Aral SO, WL HW, Holmes KK. \"It takes a village\": understanding concurrent sexual partnerships in Seattle, Washington. Sex Transm Dis. 2002;29(8):453-462.6. Aral SO, Leichliter JS. Non-monogamy: risk factor for STI transmission and acquisition and determinant of STI spread in populations. Sex Transm Infect. 2010;86 Suppl 3:iii29-36.7. Aral SO, Leichliter JS, Blanchard JF. Overview: the role of emergent properties of complex systems in the epidemiology and prevention of sexually transmitted infections including HIV infection. Sex Transm Infect. 2010;86 Suppl 3:iii1-3.8. Morris M, Kretzschmar M. Concurrent partnerships and the spread of HIV. AIDS. 1997;11(5):641-648.9. Morris M, Goodreau S, Moody J. Sexual networks, concurrency and STD/HIV. In: Holmes KK, ed. Sexually transmitted diseases. 4th ed. New York: McGraw-Hill Medical; 2008:xxv, 2166 p.10. Kenyon C, Buyze J, Colebunders R. HIV prevalence by race co-varies closely with concurrency and number of sex partners in South Africa. PLoS One. 2013;8(5):e64080.11. Kenyon CR. HIV prevalence by ethnic group covaries with prevalence of herpes simplex virus-2 and high-risk sex in Uganda: An ecological study. PLoS One. 2018;13(4):e0195431.12. Kenyon C, Menten J, Vu L, Maughan Brown B. Male circumcision and sexual risk behaviors may contribute to considerable ethnic disparities in HIV prevalence in Kenya: an ecological analysis. PLoS One. 2014;9(8):e106230.13. Kenyon CR, Tsoumanis A, Schwartz IS. A population's higher-risk sexual behaviour is associated with its average sexual behaviour-An ecological analysis of subpopulations in Ethiopia, Kenya, South Africa, Uganda and the United States. Epidemics. 2016;15:56-65."
}
]
},
{
"id": "42433",
"date": "11 Jan 2019",
"name": "Eline L Korenromp",
"expertise": [
"Reviewer Expertise epidemiology",
"program and impact evaluation",
"disease surveillance",
"cost effectiveness modelling",
"infectious diseases",
"HIV/AIDS",
"STIs",
"malaria",
"reproductive 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\nI agree with the earlier comments by Ann Jolly, and am adding only a few minor additional comments here, meant for clarification.\nLegend of Figure 2: Please clarify or rephrase 'if one conditions on this variable.'\nPage 5, section(b), left column: 'generally high-quality longitudinal data on HIV prevalence around the world': I think these prevalence surveys are not really longitudinal (which commonly denotes cohort data, where a cohort of identified individuals is followed for incidence at individual-level), but rather periodic (in a similar population, but with some turnover of individuals). Please rephrase here - and in a similar instance later on in the article.\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": []
}
] | 1
|
https://f1000research.com/articles/7-1880
|
https://f1000research.com/articles/8-119/v1
|
29 Jan 19
|
{
"type": "Research Note",
"title": "Synthesis of adsorbent from food industry waste for purification of used cooking oil",
"authors": [
"Sulistyo Prabowo",
"Muflihah Muflihah",
"Muflihah Muflihah"
],
"abstract": "This study aimed to utilize food industry waste (i.e. banana (Musa paradisiaca L.) peel), as a raw material for making banana peel activated carbon (BPAC). The activated carbon-making process was conducted at varying temperatures (200, 400 and 600°C) and furnacing times (1, 2 and 3 hours). The purification function of the BPAC obtained from the optimization process was assessed with used cooking oil (UCO) from the food industry. The purified oil was tested for three quality parameters, the peroxide value, free fatty acid value and iodine value. The results of this testing showed that BPAC could slightly improve the quality of used cooking oil.",
"keywords": [
"Activated carbon",
"banana peel",
"used cooking oil",
"purification",
"quality"
],
"content": "Introduction\n\nFrying food with cooking oils, such as palm oil, is a common cooking method. Cooking oil functions as a heat transfer medium, improves the physical appearance and texture of the food, provides savory flavors, and adds nutritional and calorific value to foodstuffs1.\n\nThe need for cooking oil in Indonesia was estimated at 1.1 million tons and 3.5 million tons on particular days of high use, such as religious holidays2. This situation led to the rising price of palm oil, forcing some producers of small-medium sized enterprises (SMEs) to use cooking oil repeatedly to save money3.\n\nThe repeated use of cooking oil at high temperatures will change the physicochemical properties, damaging the oil, due to oxidation4. A concise review of number of studies has been carried out5, which has demonstrated the adverse effects of using UCO. Thus, UCO tends to be discarded since there is no further use for it. As a result, increased cooking oil consumption can become a major environmental issue. Based on this concern, this present research was conducted to assess the potential for UCO purification utilizing other unused materials, such as banana peels.\n\nSeveral publications have described the re-use of cooking oil so it is not wasted4–10. One method is adsorption using adsorbents. Common adsorbents include carbon compounds obtained from sources such as coconut shells, wood and coal, which have been pretreated, turning them into activated charcoal.\n\nTaking into account the abundance of banana peel waste, this study attempted to produce activated charcoal derived from this resource. The aim was to determine the optimum temperature and time for the banana peel carbonation process to produce adsorbents, which could later be used to improve the quality of UCO.\n\n\nMethods\n\nUCO samples were randomly taken from three SMEs foodstalls in Samarinda, Indonesia. All samples were then mixed together and tested in duplicate. As the control, we used the most widely oil brand used by them. Banana peels were collected from fried banana traders in the same location. Samples were collected on August 8, 2018, before the SME closed their stalls in the night.\n\nThe tools used in this study were an oven, furnace, sieve, burette, Erlenmeyer flask, stirring rod, pipette, measuring glass and hotplate. The materials used were distilled water, acetic acid, chloroform, KI, 0.01 M Na2S2O3, starch, phenolphthalein, ethanol, 0.5 M KOH, 0.5 M HCl, iodine-bromide solution, and NaOH.\n\nThe dried banana peel was furnaced for 3 hours at 200°C, 400°C or 600°C until the materials turned into charcoal. Subsequently, the temperature of the charcoal was allowed to cool, and the charcoal was immediately sieved with a 200-mesh-size sieve and then activated using 1 M NaOH solution at ambient temperature. The charcoal was stirred for 2 hours and then neutralized by washing with distilled water until the pH is neutral. The obtained charcoal was then dried in the oven at 80°C.\n\nSome impurities and foreign materials were initially removed in the precipitation and separation process. Samples were put in a glass tube (30 cm height, 5 cm diameter) and the impurities were left to settle at room temperature by gravitation. This process was performed without reducing the amount of free fatty acids in the UCO.\n\nIn the UCO refining process, 10 g of BPAC was added to 200 ml UCO and stirred with magnetic stirrer at the room temperature. After 24 hours, the UCO was filtered out using a paper filter.\n\nThe color of the oil was analyzed using a scoring sensory test by 16 untrained panelists, using a scale of 1 (very dark brown), 2 (dark brown), 3 (rather dark brown), 4 (clear yellow) and 5 (very clear yellow, (equivalent to fresh oil)). Judgement of oil color was performed using UCO filtered with charcoal heated to 600°C for 3 hours. Peroxide value (PV), free fatty acid value (AV) and iodine value (IV) were analyzed using the official methods of analysis of the Association of Official Analytical Chemists (AOAC)11. PV, AV and IV were evaluated using charcoal heated at each temperature for each length of time.\n\nPV was determined as follow. Approximately 5.0 g of UCO sample was placed into iodine flask and was dissolved in the solution mixture of 50 mL glacial acetic-acid isooctane (3: 2, v/v). The solution was added with 0.5 mL saturated solution of KI. The mixture was then shaken vigorously for 0.5 min and allowed in the dark condition for another 3 min. The solution was added with 30 mL distilled water and was titrated using sodium thiosulphate 0.01 N using 1 mL of starch indicator 0.05%. Titration was stopped if blue colour of solution just disappeared. The blank titration was also carried out under similar condition without addition of samples. PV was calculated as:\n\n\n\nwhere PV is peroxide value (in meq/kg), Vs is volume (in mL) of thiosulphate used for sample titration, Vb is volume (in mL) of thiosulphate used for blank titration, Nthio is normality of thiosulphate.\n\nAcid value (AV) was determined using a titration method. A 10.0 g of samples were accurately weighed and dissolved in 100 mL ethanol-ethyl ether mixture (1:1 v/v). This solution was then titrated using standardized KOH-ethanolic solution using phenolphthalein as indicator until pink-violet color was observed. AV was expressed as the number of mg KOH needed to neutralize free fatty acids in 1 g of sample. AV was calculated as the volume of KOH needed per gram sample multiplied by N KOH and 56.1.\n\nDetermination of iodine value. A-1.0 g of samples was added with 20 mL cyclohexane-acetic acid mixture (1:1 v/v). The solution was added with 25 mL of Wijs solution (iodine monochloride, ICl) and was kept in the dark condition for 1 h. The mixture was added with 20 mL saturated KI solution and 150 mL distilled water, shaken homogenously, and titrated with 0.1 N sodium thiosulphate 0.1 N using 1 mL of starch indicator 0.05% until the color became clear. The blank titration was also carried under similar condition without addition of FO samples. IV was calculated as:\n\n\n\nVb is volume (in mL) of thiosulphate used for blank titration and Vs is volume (in mL) of thiosulphate used for sample titration.\n\n\nResults and discussion\n\nTable 1 presents the results of the score observations of color. Initially, the color of fresh cooking oil is very clear yellow due to the presence of α and β-carotene or xanthophyll4. This color and clarity becomes darker and brown, respectively, after several uses for frying as a result of the degradation of natural dyes and oxidation12. Some foodstuffs with high protein and sugar contents will initiate browning reactions as a result of Maillard and caramelization reactions. During the refining process, significant color changes occurred. Banana peel is able to absorb dark colors in UCO because of its lutein content.\n\nFigure 1 presents the results analysis on the quality of samples before and after the treatments using BPAC which were made from different temperatures and time. Increasing the temperature and time in the charcoal production process has a positive effect on the quality of cooking oil refinery. Higher temperatures and longer casting processes make the activated carbon more refined to make the material surface wider. Surface area is one factor that affects the rate of adsorption of dirt in used cooking oil.\n\nThe level of free fatty acids is an important parameter when determining the quality of cooking oil since it could determine the age, purity and the level of oil hydrolysis. Free fatty acid content exceeding 0.30% of the oil mass result in an undesirable flavor4. Based on the results of UCO sample analysis, the level of free fatty acids was quite high (1.23%). However, after soaking with BPAC the AV levels decreased to 0.71%, but this number still does not conform to Indonesia’s standard (0.3%) of oil quality13. The high levels of free fatty acids in UCO indicate the degree of oil damage4. In this research, soaking with BPAC gives a quite efficient effect in reducing about 40–50% damage.\n\nIn this study, the saponification number of UCO cannot be improved to the level of recommended quality standard. Instead, the value tends to decrease. It was also found that UCO has a higher molecular weight than fresh oil, meaning that it has bound many other foreign materials other than oil; so that it is no longer suitable for use.\n\nThe peroxide compound concentration in oil indicates the level of cooking oil degradation due to oxidation in heating process4. The content of peroxide compounds in cooking oil can accelerate rancidity and drive the presence of undesired flavor.\n\nInterestingly, peroxide values significantly dropped during soaking with BPAC. Yet, the peroxide value still does not conform to the quality standard which is at maximum of 2 mEq/ kg. Immersion with BPAC reduced these values by about 50–60%.\n\n\nConclusion\n\nProducing activated carbon from banana peels was found to have an impact in improving the quality of used cooking oil.\n\n\nData availability\n\nRaw data for oil color assessments and effect of charcoal filtering on free fatty acids, saponification and peroxide values of used cooking oil are available on Zenodo. DOI: https://doi.org/10.5281/zenodo.167394014.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Grant information\n\nThe authors thank to Project Implementation Unit Islamic Development Bank (PIU-IDB) Mulawarman University for granting this research through grant number 2248/UN17.11/PL/2018.\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 acknowledge Awalus for helping this research, and Ma’adzatul Adawiyah and Lilik Sri Rahayu who collected the samples.\n\n\nReferences\n\nRamdja AF, Febrina L, Krisdianto D: Pemurnian minyak jelantah menggunakan ampas tebu sebagai adsorben. Jurnal Teknik Kimia. 2010; 17(1): 7–14. Reference Source\n\nFauziah Sirajuddin S, Najamuddin U: Analysis of fatty acid in fried and used oil from snack food frying results in workshop Unhas. Fakultas Kesehatan Masyarakat Universitas Hasanuddin Makassar. 2013.\n\nSumardi E: Waduh! Minyak Goreng Dipakai Penjual Gorengan Kebanyakan Beracun. Accessed 12 October 2018, 2015. Reference Source\n\nRohman A: Lipid: Sifat fisika kimia dan analisisnya. Pustaka Pelajar. Yogyakarta, 2016. Reference Source\n\nKu SK, Muhamad Ruhaifi MS, Fatin SS, et al.: The harmful effects of consumption of repeatedly heated edible oils: a short review. Clin Ter. 2014; 165(4): 217–221. PubMed Abstract | Publisher Full Text\n\nAlireza S, Tan CP, Hamed M, et al.: Effect of frying process on fatty acid composition and iodine value of selected vegetable oils and their blends. Int Food Res J. 2010; 17: 295–302. Reference Source\n\nAmalia F, Retnaningsih, Johan IR: Analysis Of The Behaviour Effects In Using Cooking Oil On The Participation Program Of Collecting The Used Cooking Oil In Bogor City. Jur Ilm Kel & Kons. 2010; 3(2): 184–189.\n\nAdinata MR: Pemanfaatan Kulit Pisang sebagai Karbon Aktif. UPN: Jawa Timur. 2013. Reference Source\n\nHidayati FC, Masturi, Yulianti I: Pemurnian Minyak Goreng Bekas Pakai (Jelantah) dengan Menggunakan Arang Bonggol Jagung. Jurnal Ilmu Pendidikan Fisika. 2016; 1(Nomor 2): 67–70. Publisher Full Text\n\nKontan: Konsumsi minyak goreng meningkat. Accessed: 9 October 2018. 2017. Reference Source\n\nNatsir NSW, Nurhaeni, Musafira: Pemanfaatan Arang Aktif Kulit Pisang Kepok (Musa Normalis) sebagai Adsorben untuk menurunkan Angka Peroksida dan Asam Lemak Bebas Minyak Goreng Bekas. Online Jurnal of Natural Science. 2014; 3(1): 18–30. Reference Source\n\nWarner K: Impact of high-temperature food processing on fats and oils. Adv Exp Med Biol. In: Jackson LS, Knize MG, Morgan JN. (eds) Impact of Processing on Food Safety. Springer, Boston, MA. 1999; 459: 67–77. PubMed Abstract | Publisher Full Text\n\nBadan Standardisasi Nasional: Standar Minyak Goreng. SNI 01- 3741-2002. Jakarta; 2002. Reference Source\n\nPrabowo S, Muflihah: Used Cooking Oil sample quality parameters [Data set]. Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1673940"
}
|
[
{
"id": "43732",
"date": "04 Mar 2019",
"name": "Nattapol Tangsuphoom",
"expertise": [
"Reviewer Expertise food lipids",
"food waste utilization and recovery",
"food material characterization"
],
"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 entitled “Synthesis of adsorbent from food industry waste for purification of used cooking oil” demonstrates the application of activated carbon prepared from banana peels in improving the quality of used cooking oil. Although the experiment was reasonably designed and executed, there are some significant points that need to be improved. I thus recommend major revision on these following points before accepting for indexing:\n\nAbstract:\nLine 8: Quantitative description of “slightly improve” should be given.\n\nMethods:\nIn general this part has a lack of statistical analysis. Means and standard deviations of triplicate samples should be reported and analyzed for the statistical difference. It would be more concrete if characterization of the BPAC is conducted such as bilk density, porosity and microscopic analysis for microstructure and compared with commercial activated carbon. Discussion of results should also be included. The quality of UOC after being treated with BPAC should be compared to the standard for cooking oil.\nResearch sample:\nWhat type of oils were collected (palm, coconut, etc.)? Which processing/cooking method were these UCO samples subjected to?\nOil quality testing:\nWhy was only the UCO sample treated with BPAC sample prepared at 600°C for 3 hours tested for colour? What about the colour of other treated UCO samples?\n\nResults and discussion:\nTable 1: The distribution of score should be presented.\n\nFigure 1: Data should be presented in the form of a Table.\n\nSaponification number: This measurement does not appear in the Methods part.\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? No\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": "4540",
"date": "04 Apr 2019",
"name": "Sulistyo Prabowo",
"role": "Author Response",
"response": "Thank you very much for this great review to improve our research. This is ongoing research and we still working on it. We acknowledge many limitations in this short communication and will do the suggestion."
}
]
},
{
"id": "47861",
"date": "06 Jun 2019",
"name": "Abdul Rohman",
"expertise": [
"Reviewer Expertise Fats and Oils 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 statistical analysis should be improved. How many replicates? Please add error bars for the results in Figure 1.\n\nPlease, compare the results with those published regarding the quality of cooking oils from other frying oils.\n\nThe Conclusion must be more elaborated.\n\nPlease add newer references from primary sources. There are some papers published on this issue.\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/8-119
|
https://f1000research.com/articles/8-118/v1
|
29 Jan 19
|
{
"type": "Case Report",
"title": "Case Report: A case of metastatic adenocarcinoma found during inguinal hernia repair",
"authors": [
"Akram Rajput",
"Aatera Haq",
"Syed Mohammad Mazhar Uddin",
"Zeeshan Zafar",
"Uzair Yaqoob",
"Akram Rajput",
"Aatera Haq",
"Syed Mohammad Mazhar Uddin",
"Zeeshan Zafar"
],
"abstract": "Tumors found during hernia repair are rare. They may be intrasaccular (most commonly), saccular or extrasaccular, and they are usually primary. In this case report we present a case of metastatic adenocarcinoma (confirmed by biopsy) found inside the inguinal hernia sac. Following further investigation, primary tumor of the ileum and ileocecal junction was found. An elderly male presented with a history of bilateral inguinal swelling for a year, that had been asymptomatic initially, but had increased in size and was painful following prolonged activity. During laparoscopic hernioplasty a right sided inguinal hernia with nodules on peritoneum and omentum were observed. After taking a biopsy, repair of the hernia was performed. Carcinoembryonic antigen (CEA) and carbohydrate antigen (CA) 19-9 levels were raised. A computed tomography (CT) scan of the abdomen and chest showed a mass involving the distal ileum and the ileocecal junction, with lymphadenopathy, peritoneal carcinomatosis, as well as hepatic and lung metastasis. Surgery was not possible in this patient. A possible relation of an inguinal hernia with an ileal neoplasm was found in this patient. An irreducible swelling not causing any problems can be suspicious, which should be properly investigated to get to an early diagnosis.",
"keywords": [
"Adenocarcinoma",
"inguinal hernia",
"hernioplasty",
"metastases",
"incidental finding"
],
"content": "Introduction\n\nInguinal hernia sac tumors are a rare occurrence, with one study reporting only 0.07% of repairs being positive for metastatic tumors1. Traditionally, hernia sac tumors are classified as intrasaccular, saccular and extrasaccular, based on the anatomical relationship of the tumor to the hernia sac2,3. Intrasaccular tumors are one the most frequent type. They generally consist of primary tumors of organs lying within the hernia sac, such as cancers of colon, bladder, and metastatic neoplasms involving the omentum. Tumors that extend into the hernia sac by way of peritoneal involvement are classified as saccular and include, among others, primary mesothelioma and peritoneal metastasis from the intra-abdominal organs. When the tumor is within the hernia defect but lies outside the hernia sac, it is classified as extrasaccular. Examples include metastatic involvement of inguinal lymph nodes2–4. In this case report we present a case of metastatic growth that was found inside the inguinal hernial sac, later confirmed by histopathology to be metastatic adenocarcinoma. Post-operative workup revealed the primary tumor to be an adenocarcinoma involving the distal ileum as well as the ileocecal junction.\n\n\nCase presentation\n\nA 57-year-old man, with no known comorbid, from Karachi, Pakistan, was admitted to the Civil Hospital, Karchi, Pakistan, in November 2017, with a history of bilateral inguinal swelling that had begun a year prior to presentation, that had been small and inconsequential initially, but had recently grown larger and would become painful during prolonged standing and walking. In last three weeks prior to admittance, the swelling would not reduce with lying down. On physical examination, he was found to have a bilateral, incomplete, non-reducible inguinal hernia. His baseline laboratory investigations were all within normal limits, and chest radiograph was unremarkable. The patient underwent a cardiac review and was declared fit for general anesthesia. He was scheduled for laparoscopic hernioplasty. Intra operative findings revealed white nodules on parietal peritoneum and the omentum, while only the right-sided inguinal hernia was observed. A biopsy was taken, and mesh repair was done on the right side. Histopathology of the biopsy showed metastatic adenocarcinoma. The patient was evaluated for an intra-abdominal neoplasm. Carcinoembryonic antigen (CEA) levels and carbohydrate/cancer antigen (CA 19-9) levels were sent and patient scheduled for an ultrasound scan (US) of the abdomen and pelvis, esophagogastroduodenoscopy (OGD), colonoscopy and computed tomography (CT) scan of the chest, abdomen and pelvis with contrast enhancement. The patient’s CEA levels were 14.04 ng/ml (normal level <3.5 ng/ml) and CA 19-9 levels were 125.2 U/ml (normal level <27 U/ml). The only findings of the abdominal and pelvic ultrasound were hypo-echoic lesions in the liver. OGD showed antral gastritis and biopsies were taken which showed no evidence of malignancy. Colonoscopy showed rounded well-demarcated lesions measuring 1.5 cm seen in the cecum, multiple biopsies were taken. However, biopsy showed no evidence of malignancy. Computed tomography (CT) of abdomen and chest showed a mass involving the distal ileum and the ileocecal junction, with lymphadenopathy, peritoneal carcinomatosis, as well as hepatic and lung metastasis (Figure 1). The patient was considered inoperable and hence, no surgery was considered. The patient left against medical advice and no follow up was obtained.\n\n\nDiscussion\n\nMalignant tumors presenting within inguinal hernias are a rare occurrence. Literature reveals less than 0.4 % of the excised hernia tissue shows microscopic evidence of neoplasia5. Among these the most common primary tumor associated with hernia sac metastasis is carcinoma of the colon2,5. This case was unique as in our patient the neoplasm involved the small bowel (distal ileum) as well as the ileocecal junction. To our knowledge, there is no literature available which mentions a case of metastatic ileal/ileocecal junction neoplasm presenting as an inguinal hernia. In most cases of colon cancer presenting as an inguinal hernia one possible explanation for the occurrence of an inguinal hernia could be the increased intra-abdominal pressure, perhaps secondary to the intra-abdominal neoplasm, occurring especially in the elderly2,6. For instance, obstructive colon cancer, massive tumors or tumors associated with ascites can lead to increased intra-abdominal pressures and result in an inguinal hernia. However, in our patient, there were no additional symptoms, indicating that other factors were involved.\n\nHere we turn to the history the patient gave us on presentation. A patient’s presenting complainant is an important factor and can raise suspicion of an underlying malignancy3. It has been suggested that a long-standing hernia, becoming acutely incarcerated has a greater chance of containing a tumor. Some authors have even gone so far as to say that any non-reducible mass in the inguinal canal, lacking an impulse, should raise concern for malignancy3. Furthermore, constitutional symptoms can also suggest the possibility of an underlying malignancy. In many such cases, abdominal pain was the most frequent symptom present pre-operatively3. Of note in our patient is the history of acute incarceration, as the patient said that up till recently the swelling would reduce on lying down. Since the peritoneum is a common site of intra-abdominal metastasis, a hernia repair gives a good opportunity for the surgeon to perform a peritoneal biopsy, thereby providing an earlier diagnosis. Furthermore, due to reports of occult malignancies from histopathological examination of hernia sac, several authors have recommended routine microscopic examination of the hernia sac2,7. However, other authors have reported histopathological examination in only selected cases2. Several authors also suggest a routine fiberoptic sigmoidoscopy in patients presenting with a hernia, because of the coexistence of an inguinal hernia and colonic cancer2,8.\n\n\nConclusion\n\nIn conclusion, our study showed that an inguinal hernia and ileum/ileocecal junction neoplasm can coexist together, especially in the elderly. In addition, we also reported that an irreducible hernia can raise suspicion of metastasis and, the patient can be asymptomatic at the time of presentation. Despite the rare co-existence of hernia and malignancy, we still recommend routine microscopic histopathological examination in all suspicious cases, as this can lead to an earlier diagnosis, and if the patient is asymptomatic this can be the only means to determine an occult malignancy.\n\n\nConsent\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": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nVeal DR, Hammill CW, Wong LL: Lymphoma diagnosed at inguinal hernia repair. Hawaii Med J. 2010; 69(2): 32–4. PubMed Abstract | Free Full Text\n\nOruç MT, Kulah B, Saylam B, et al.: An unusual presentation of metastatic gastric cancer found during inguinal hernia repair: case report and review of the literature. Hernia. 2002; 6(2): 88–90. PubMed Abstract | Publisher Full Text\n\nNicholson CP, Donohue JH, Thompson GB, et al.: A study of metastatic cancer found during inguinal hernia repair. Cancer. 1992; 69(12): 3008–11. PubMed Abstract | Publisher Full Text\n\nBrenner J, Sordillo PP, Magill GB: An unusual presentation of malignant mesothelioma: the incidental finding of tumor in the hernia sac during herniorrhaphy. J Surg Oncol. 1981; 18(2): 159–61. PubMed Abstract | Publisher Full Text\n\nMarsden M, Curtis N, McGee S, et al.: Intrasaccular caecal adenocarcinoma presenting as enlarging right inguinoscrotal hernia. Int J Surg Case Rep. 2014; 5(10): 643–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGeelhoed GW, Millar RC, Ketcham AS: Hernia presentation of cancer in the groin. Surgery. 1974; 75(3): 436–41. PubMed Abstract\n\nYoell JH: Surprises in hernial sacs; diagnosis of tumors by microscopic examination. Calif Med. 1959; 91: 146–8. PubMed Abstract | Free Full Text\n\nLovett J, Kirgan D, McGregor B: Inguinal herniation justifies sigmoidoscopy. Am J Surg. 1989; 158(6): 615–6; discussion 616–7. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "50930",
"date": "12 Jul 2019",
"name": "Abdul-Wahed N. Meshikhes",
"expertise": [
"Reviewer Expertise General and minimally invasive surgery. Gastrointestinal and colorectal 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\nThis case report summarizes the case of an elderly man presenting with inguinal hernia that became recently symptomatic. During laparoscopic repair, nodules were found in the sac. Biopsy revealed metastatic ileo-caecal adenocarcinoma. Whether to go ahead with the repair using a prosthetic mesh or no repair in such advanced metastatic disease is a matter of controversy.\n\nThe present case is interesting as the primary cancer turned up to be an advanced ileo-caecal carcinoma.\nI have the following comments:\nI presume the procedure was done trans-abdominally i.e transabdominal preperitoneal (TAPP) approach. This needs elaboration.\n\nWhy there is no comment on the laparoscopic exploration of the abdominal cavity upon the discovery of the nodules? Was the primary seen laparoscopically especially as it was a right hernia which was repaired?\n\nWhy was clinically bilateral hernia found, but laparoscopically, only the right side was obvious?\n\nIt is worth including a figure of the laparoscopic view of the nodules.\n\nWas the idea of sending the nodule for frozen section entertained?\n\nWhy was not a preoperative chest x-ray done, as the patient was elderly undergoing a laparoscopic abdominal procedure? If done, why were the metastatic lung nodules missed?\n\nThere was no mention of the abdominal examination. Also no mention whether the tumour was causing any obstructive symptoms or not?\n\nI disagree with the authors’ notion that “acutely incarcerated hernia has a great chance of containing a tumor”. This is not supported by our daily surgical practice. However, it needs to be emphasized that acute hernia incarceration in the elderly should raise suspicion.\n\nI also find the suggestion of “doing routine fibrooptic sigmoidoscopy in all patients presenting with hernia to exclude co-existence of inguinal hernia and colon cancer” is an ‘overkill’. I believe this should be selective and only if there is an indication or a clinical suspicion of co-existing colorectal cancer.\n\nI agree with routine histopathological examination of the excised hernia sac.\n\nThe authors did not discuss the surgical approach recommended in such cases of metastatic nodule in the hernia sac when discovered during the repair. Will the surgeon go ahead with the repair or not? Is the use of a prosthetic mesh advisable or not?\n\nThe discussion is deficient on the literature review of similar cases especially that of metastatic intestinal/colorectal primary. It is worth listing such cases which were reported over the past 10 years in a table and whether they were discovered electively vs. emergency, open repair vs. laparoscopic, or mesh vs. no mesh, etc.\n\nA spelling mistake in the 2nd paragraph of the Discussion: Complainant should be corrected to complaint.\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": "54648",
"date": "04 Nov 2019",
"name": "Deepti M. Reddi",
"expertise": [
"Reviewer Expertise Gastroenterology."
],
"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 good article which emphasizes further investigation and work-up of irreducible hernia before hernioplasty, especially in the elderly population for occult malignancy. It is noted that there is a biopsy specimen, which confirms metastatic adenocarcinoma. Later, it reports cecal biopsies showed no evidence of malignancy. If there are supporting immunohistochemistry (CK7, CK20, CDX2, Villin, SATB2, etc.) along with the primary biopsy, then it would be more convincing to justify the relationship between the hernia and an intestinal neoplasm.\nThere are many grammatical errors which need to be addressed before indexing. If there are any histology pictures then it will be more interesting.\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? Partly",
"responses": []
},
{
"id": "63880",
"date": "22 Jun 2020",
"name": "Ruchika K. Goel",
"expertise": [
"Reviewer Expertise Histopathologist."
],
"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 case report of an elderly patient with a diagnosis of a metastatic adenocarcinoma in an inguinal hernia sac. The case report has outlined the clinical history along with preoperative investigations in in detail. The discussion of the article clearly describes importance of history and the relevance of diagnostic biopsy. In the biopsy however, there is no mention of the immunohistochemistry markers performed (CK7,CDX2,VILLIN, CALRETININ etc). Also when small whitish nodules were seen on parietal peritoneum and omentum, a differential diagnosis of mesothelioma does arise. There is no mention of the differentials considered in the biopsy in this case which was reported as metastatic adenocarcinoma. This is of utmost importance as all the subsequent biopsies were negative. It would be nice to have more detailed hisopathological findings along with necessary histological pictures.\nThe article is brief and written well, however grammatical errors need to be corrected and submission of pictographs would further enhance the scientific evidence.\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? No\n\nIs the case presented with sufficient detail to be useful for other practitioners? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-118
|
https://f1000research.com/articles/8-117/v1
|
29 Jan 19
|
{
"type": "Method Article",
"title": "An in vitro model for studying CNS white matter: functional properties and experimental approaches",
"authors": [
"Silvia Bijland",
"Gemma Thomson",
"Matthew Euston",
"Kyriakos Michail",
"Katja Thümmler",
"Steve Mücklisch",
"Colin L. Crawford",
"Susan C. Barnett",
"Mark McLaughlin",
"T. James Anderson",
"Christopher Linington",
"Euan R. Brown",
"Eric R. Kalkman",
"Julia M. Edgar",
"Silvia Bijland",
"Gemma Thomson",
"Matthew Euston",
"Kyriakos Michail",
"Katja Thümmler",
"Steve Mücklisch",
"Colin L. Crawford",
"Susan C. Barnett",
"Mark McLaughlin",
"T. James Anderson",
"Christopher Linington",
"Euan R. Brown",
"Eric R. Kalkman"
],
"abstract": "The normal development and maintenance of CNS white matter, and its responses to disease and injury, are defined by synergies between axons, oligodendrocytes, astrocytes and microglia, and further influenced by peripheral components such as the gut microbiome and the endocrine and immune systems. Consequently, mechanistic insights, therapeutic approaches and safety tests rely ultimately on in vivo models and clinical trials. However, in vitro models that replicate the cellular complexity of the CNS can inform these approaches, reducing costs and minimising the use of human material or experimental animals; in line with the principles of the 3Rs. Using electrophysiology, pharmacology, time-lapse imaging, and immunological assays, we demonstrate that murine spinal cord-derived myelinating cell cultures recapitulate spinal-like electrical activity and innate CNS immune functions, including responses to disease-relevant myelin debris and pathogen associated molecular patterns (PAMPs). Further, we show they are (i) amenable to siRNA making them suitable for testing gene-silencing strategies; (ii) can be established on microelectrode arrays (MEAs) for electrophysiological studies; and (iii) are compatible with multi-well microplate formats for semi-high throughput screens, maximising information output whilst further reducing animal use. We provide protocols for each of these. Together, these advances increase the utility of this in vitro tool for studying normal and pathological development and function of white matter, and for screening therapeutic molecules or gene targets for diseases such as multiple sclerosis, motor neuron disease or spinal cord injury, whilst avoiding in vivo approaches on experimental animals.",
"keywords": [
"(re)myelination",
"microglia",
"microelectrode array",
"semi-high throughput",
"spinal cord",
"siRNA"
],
"content": "\n\nRecapitulates in vivo-like myelination, innate immune responses and neuronal electrical activity and is amenable to live cell imaging.\n\nCompared to single cell type or mixed PNS-CNS cell cultures, this pure CNS multi-cell type system more closely resembles the in vivo situation.\n\nUncouples CNS-intrinsic responses from those mediated by peripheral organs and systems.\n\nEase of manipulation, i.e. genetic using siRNA; neuronal electrical activity using pharmacological modulators of neural activity; small molecules using small molecule libraries; relevant PAMPs or DAMPS.\n\nCan be generated from transgenic reporter mice for live imaging and functional readouts or from mutant or transgenic models of disease.\n\nInform studies and/or acts as an adjunct to studies using in vivo models of multiple sclerosis, motor neurone disease, the leukodystrophies and other neurodegenerative diseases involving white matter, minimising the use of experimental animals and maximising information obtained.\n\nStraightforward to establish in any lab with cell culture facilities.\n\nCompared to CNS slice cultures, which also contain all major neural cell types, this cell culture system is easier to maintain (oxygen and nutrients readily reach all cells) and quicker to set up.\n\nCells can be grown on microelectrode arrays that do not require the specialist expertise needed for single cell electrophysiology.\n\nCells can be grown on multi-well plates for semi-high throughput assays.\n\nMulti-well microplate formats facilitate the testing of multiple factors on parallel cultures.\n\nSemi-high throughput screens for pro-myelinating or inhibitory factors; testing and verifying siRNA constructs and/or testing the consequences of gene knockdown; modulating neural activity to assess secondary consequences for other cells types; live imaging of cellular interactions.\n\nSemi-high throughput screens for factors that affect axonal survival or regeneration.\n\nAssessing effects of microglial ablation on other cell types, by manipulating CSF1R signalling pathway.\n\nLive imaging of organelle distribution and/or transport.\n\nInitial screen of anti-sense oligonucleotides for gene silencing in the context of in vivo-like cellular phenotypes and morphologies.\n\n\nIntroduction\n\nDiseases that affect white matter are many and varied, and include spinal cord injury, motor neuron disease, Alzheimer’s and multiple sclerosis. Typically, these diseases are modelled and studied in experimental rats and mice, and occasionally in primates, undergoing procedures such as spinal cord contusion, genetic modification or induction of experimental allergic encephalopathy (EAE); procedures that range in severity from moderate to severe under Animals (Scientific Procedures) Act 1986 licensing. Whilst such studies can and do provide important information relevant to human disease, they require careful design and large numbers of animals to provide sufficient power (Baker & Amor, 2014). For example, we estimate that to test the efficacy of a single drug, 3 cohorts each of 12 animals with EAE is required, depending on effect size and consistency in response (C. Linington, personal communication). Alternative approaches include in vitro techniques, such as human-derived induced pluripotent stem cell models or murine cell cultures.\n\nIndeed, cell and tissue culture is used widely in neuroscience to study the development and function of the major cell types of the CNS; neurons, oligodendroglia, astrocytes and microglia. The main reasons being, in vitro models are a) relatively inexpensive; b) amenable to manipulation, including pharmacological and genetic; c) accessible to live imaging and optogenetic approaches; d) reduce the reliance on experimental animal models for early-stage screening/proof-of-concept studies; and e) inform subsequent animal studies, if they are required. In particular, cell culture assays can reduce the numbers of animal used in in vivo studies by, for example (i) guiding selection of candidate therapeutic molecules, (ii) informing drug dosage concentrations, (iii) measuring toxicity (iv) defining gene targeting efficiency (v) testing gene constructs prior to the generation of transgenic animals.\n\nSingle cell type-enriched cultures have been invaluable in addressing questions relating to cell autonomous characteristics (Bechler et al., 2015; Lee et al., 2012a; Mei et al., 2014; Sanchez-Gomez et al., 2018) and simple co-cultures have shed light on bi-cellular interactions (Froger et al., 2010; Lundgaard et al., 2013). However, multi-cell type (Madhavan et al., 2018; Thomson et al., 1993) or explant cultures (Thomson et al., 2006; Zhang et al., 2011) that maintain the complex cellularity and functional properties of the CNS, more closely represent the in vivo situation. As such, they are more relevant for addressing questions where the answer relies on physical interactions and/or paracrine signaling between CNS cell types; notwithstanding they remain uncoupled from the influence of peripheral organs or systems including the gut microbiome, the adaptive immune system and the endrocrine system.\n\nMyelinated nerve fibers are the culmination of complex bidirectional molecular, structural and functional interactions between axons and oligodendrocytes. For example, neuronal electrical activity modulates myelination during development (Demerens et al., 1996; Wake et al., 2011 and reviewed in Almeida, 2018; Almeida & Lyons, 2017; Zalc & Fields, 2000) and the myelinating cell supports the myelinated axon throughout life (Edgar et al., 2009; Griffiths et al., 1998; Lappe-Siefke et al., 2003) by modulating axonal transport and the axonal cytoskeleton (Edgar et al., 2004; Kirkpatrick et al., 2001; Pan et al., 2005), maintaining white matter homeostasis (Kassmann et al., 2007; Schirmer et al., 2018) and providing energy substrates (Fünfschilling et al., 2012; Lee et al., 2012b; Meyer et al., 2018). Adjacent microglia and astrocytes shape and support the myelinated fiber during development and adulthood (reviewed in Allen & Lyons, 2018; Thion et al., 2018). For example, microglia clear apoptotic oligodendrocytes generated in excess during development (Barres & Raff, 1994; Barres et al., 1992), boost developmental myelination by expressing IGF-1 (Wlodarczyk et al., 2017) and remove debris following demyelination or Wallerian degeneration, albeit slowly (George & Griffin, 1994). Astrocytes secrete factors that enhance developmental myelination (Ishibashi et al., 2006 and reviewed in Barnett & Linington, 2013), form gap junctions with each other and with oligodendrocytes for exchange of ions and small metabolites (reviewed in Cotrina & Nedergaard, 2012) and regulate the structure of the mature myelin sheath through secretion of inhibitors of thrombin proteases (Dutta et al., 2018). Conversely, ‘activated’ microglia and astrocytes can contribute to disease pathogenesis and injury to myelinated fibers, as in mouse models of motor neuron disease (Beers et al., 2006; Boillee et al., 2006; Hall et al., 1998; Nagai et al., 2007; Yamanaka et al., 2008), the leukodystrophies (Ip et al., 2007), and cerebral vascular disease (Fowler et al., 2018). Thus, in vitro models that replicate the interdependence of cells of the intact CNS provide an informative prelude or adjunct to in vivo studies.\n\nWe previously described a murine, spinal cord-derived myelinating cell culture system (Thomson et al., 2006; Thomson et al., 2008) to explore axonal organelle distribution (Edgar et al., 2008), cell dynamics (Ioannidou et al., 2012), pro-myelination factors (Berghoff et al., 2017; Goebbels et al., 2017) and neurotropism and pathogenesis of Zika virus infection (Cumberworth et al., 2017). However, little is known about the culture’s functional properties with respect to neuronal electrical activity or innate immune responses; factors that influence normal development as well as pathology. Further, the possibility of modifying it for use as an electrophysiological or gene silencing assay has not been explored. Here we validate this system as a functional model of CNS white matter, and describe adaptations to increase its utility in the study of development and disease. This model is straightforward to establish in any laboratory with basic cell culture facilities, and its use can be extended as described here, if equipment is available for live imaging, microelectrode array and/or multi-well plate microscopy.\n\n\nMethods\n\nAll animals were bred and maintained in conventional caging, with up to 4 cage companions, in the Biological Services Facilities (BSF) at the University of Glasgow; 12 h light-dark cycle and food and water ad libitum. Bedding was non-sterile wood-chip; food was normal maintenance diet; water was normal tap water; temperature was 19–21 °C; and humidity was 55% ± 10%. Environmental enrichment took the form of mini-tubes, sizzle nest and burrowing treats (sunflower and pumpkin seeds). The mice themselves were not health-screened but the BSF was free of the major rodent pathogens but positive for some adventitious agents, namely – for the period in question – pinworms, Helicobacter spp., Pasteurella pneumotropica and Mouse Norovirus. Embryonic day 13 (E13; day of plug being E0) mice (of both sexes; sex undetermined) were obtained by time mating wild type mice; Cnp+/+ with Cnp-/- mice (Lappe-Siefke et al., 2003), both on a C57BL/6J (Charles River) background; or wild type females with hemizygous Thy1-CFP males (Feng et al., 2000), both on a C57BL/6N (Charles River) background. Pregnant dams were killed on the morning of E13 by cervical dislocation followed by decapitation and the uterine horns containing the developing embryos were removed after laparotomy, and immediately placed on ice in a sterile 10 cm Petri dish. Adult (postnatal day [P] 60 – P120, male and female homozygous Plp1 transgenic (line #72; (Readhead et al., 1994) and wild type mice (littermates or closely related mice from the same colony) were killed in gradually increasing levels of CO2, followed by decapitation, and the spinal cord was extracted rapidly for preparation of a myelin-enriched tissue fraction. All animal use was approved by the Ethical Committee of the University of Glasgow and licensed under the Animal [Scientific Procedures] Act 1986 project licence PPL60/3656. Experiments on animal-derived cell cultures were conducted according to ARRIVE guidelines, including randomisation of samples in multi-well microplates, blinding of the experimenter and/or automation of quantification, as indicated. Numbers of technical repeats and independent biological repeats are indicated in the Figure Legends. Technical repeat: independent wells/dishes from a single cell culture. Biological repeat (or experimental unit): independent cell culture, generated by pooling all embryos from a single pregnant mouse.\n\nPlp1 tg mice ≥3 months of age mice are prone to seizures. Mice appearing dull or apathetic (suggesting a post-ictal problem) or which has an observed seizure lasting more than 2 minutes were killed by a humane method. In the vast majority of instances, the mice were used prior to the development of seizures.\n\nA step-by-step protocol is provided in the supplementary protocol document (Supplementary File 1). The procedure for preparing murine myelinating cultures described (Thomson et al., 2008) was modified slightly. Briefly, E13.5 mouse spinal cords were dissected, then incubated in 1 ml per 6 cords of 0.25% trypsin in HBSS minus calcium and magnesium for 15 minutes at 37°C. The digestion was stopped using 1 ml per 6 cords SD solution (Thomson et al., 2008) or Plating Medium (PM; 50% DMEM, 25% HBSS and 25% horse serum) plus 0.04 mg/ml DNase (or more, if required). The cells were triturated, resuspended in PM and plated on coverslips (3 x 13 mm diameter per 35 mm Petri dish), imaging dishes or microelectrode arrays (MEAs) at ~150,000 cells per 100 μl PM, per 133 mm2; or in 96 or 384-well dishes in 50 μl PM, at various concentrations (see Results). From day in vitro (DIV) 0, cells were grown in 50% PM and 50% differentiation medium (DMEM supplemented with 30% D-glucose (4500 mg/l glucose final), 10 ng/ml biotin, 50 nM hydrocortisone, 10 μg/ml insulin, 0.5% hormone mix (stock concentration 1 mg/ml apo-transferrin, 20 mM putrescine, 4 μM progesterone, and 6 μM selenium; based on (Bottenstein & Sato, 1979), at 37°C in 5 or 7% CO2, then fed three times a week by replacing half the medium with serum-free differentiation media. All DMEM contained 100 U/ml penicillin and 100 μg/ml streptomycin. From DIV 12 onwards, feeding was done with insulin-free differentiation media. The effect of Activin-A (AA; R&D Systems, Cat. 338-AC-010; Lot BNV3313053) and recombinant human fibroblast growth factor 9 (FGF9; R&D Systems, Cat. 273-F9-025; Lot ON1413121 or ON1413041) on myelinating cultures was investigated between 15 (AA) or 18 (FGF9) and 28 DIV, on cells grown on multi-well microplates. For treating with ‘myelin debris’, coverslips were transferred from 35 mm Petri dishes the day before, into 24 well dishes in maximum 500 μl differentiation media.\n\nTo minimise ‘edge-effects’ due to increased rate of evaporation or warming of media, cells were not plated in the wells on the outer edges of the microplate. Instead, these were filled with Hank’s Balanced Salts solution. DMSO (1% v/v), AA (1–100 ng ml-1) or FGF9 (100 ng ml-1 were added to wells following a pre-generated random pattern that varied from one experimental repeat to the next, to avoid potential effects related to the location of the wells on the microplates.\n\nA protocol for custom made imaging dishes is described in the supplementary protocol document (Supplementary File 2). Briefly, 3 x 11 mm diameter holes were burred in the bottom of 35 mm Petri-dishes (Falcon Ref. 353001) and 25 mm diameter glass coverslips (Menzel-Glaser 0,17 +/- 0,01 mm; Starke 1,5; Lot #004710182; Thermo Scientific) were stuck to the base of the dishes using non-toxic glue. Dishes are cleaned and sterilised then coated with poly-L-lysine. Live imaging dishes were alternatively purchased from MatTek Corporation, Ashland, USA or custom made (Kline, 2009).\n\nPoly-L-lysine (PLL; 13.3 μg/ml) in water or boric acid buffer (50 mM boric acid, 23.5 mM sodium tetraborate, pH 8.5) was used to coat 13 mm diameter glass coverslips (Fisher Scientific, Leicestershire; 631-0150), imaging dishes, microelectrode arrays or microplates (96 [Greiner 655891] or 384-well [Greiner 781856] Sensoplate Plus, black, 175 µM glass bottom, Greiner Bio-One) for 1–12 hours at 37°C, after which the PLL was aspirated, the glass washed three times in dH20 and air dried in the laminar flow hood. Empirically, we found the cultures were far less reproducible if PLL was prepared in water compared to boric acid buffer.\n\nA step-by-step protocol is provided in the supplementary protocol document (Supplementary File 3). Adult (P45-P120) wild type or Plp1 transgenic mice were killed in gradually increasing levels of CO2 and the spinal column was severed at the lumbar and cervical regions. The spinal cord was rapidly removed by introducing sterile saline under pressure to the lumbar region of the spinal canal, using an 18-gauge hypodermic needle attached to a 5 ml syringe. Cords were immediately processed or snap-frozen and stored in liquid nitrogen. Myelin was harvested following the method of (Norton & Poduslo, 1973), with slight modification. All steps were performed at 4°C, using filter sterilised solutions prepared from cell culture grade diluents or MilliQ water. Cords were homogenised in sterile 0.85 M sucrose solution in 10 mM HEPES using a polytron homogeniser at full speed for 12 strokes. Three ml of a 0.25 M sterile sucrose solution in 10 mM HEPES was slowly added on top of 7.5 ml of the homogenate. The samples were spun at 70,000 x g for 90 minutes at 4°C in a Beckman SW41 rotor. The interface between the sucrose layers, containing the membrane fractions, was gently aspirated and washed by vortexing in 6 ml chilled MilliQ water then spun at 23,000 x g for 30 minutes in a Beckman J21 rotor, to remove the excess sucrose. This osmotic shock was repeated twice more. Following a final 19,000 x g spin, the resultant myelin-enriched pellet was resuspended in cell culture grade phosphate buffered saline (PBS; Sigma-Aldrich, catalogue number 806552). The protein concentration was measured using Pierce BCA protein assay kit (Thermo Fisher Scientific, catalogue number 23225) and this myelin fraction was tested for sterility (by incubating a sample in cell culture media at 37°C for 7 days), and labeled with NHS-Rhodamine Antibody Labeling Kit (Thermo Fisher Scientific, product #53031), according to the manufacturer’s instructions.\n\nA rhodamine-labeled myelin enriched tissue fraction (2 mg protein ml-1 myelin homogenate) was added to the myelinating cultures at DIV 21, DIV 25 or DIV 27 (all ± 1 day), to a final concentration of 0.075 or 0.1 mg protein ml-1 (unless otherwise indicated) for respectively, 7, 3 or 1 DIV. Cell culture grade PBS alone (to the same volume as the myelin emulsion), or 3 × 106 one μm diameter Flurobite fluorescently labeled latex beads (Polyscience, Park scientific, Northampton, UK) were added in parallel. Two coverslips, each plated at the start with 150,000 spinal cord cells, were treated with wild type myelin, Plp1 tg myelin, PBS or latex beads. Cells were fixed and stained with rat anti-CD45 (Serotec, catalogue number MCA 1388, monoclonal; 1 in 600) to label microglia at DIV 28 (± 1 day) for analysis. Independent myelin preparations were used for each experimental repeat.\n\nA supplementary protocol document is attached (Supplementary File 4). In general, cell cultures were fixed with 4% paraformaldehyde for 10–20 min at room temperature (RT), washed in PBS, permeabilized in 0.5% Triton X-100/PBS for 15 min at RT or in methanol for 10 minutes -20°C, washed with PBS, and blocked with 1% BSA/10% horse serum/PBS or with 10% goat serum/PBS for 1 hour at RT. Primary antibodies used were: rat anti-MBP (MCA4095; 1:500); rat anti-CD45 (MCA 1388; 1:300); rat anti-mouse CD68 (MCA1957T; 1:200; all Serotec Ltd, Oxford, UK), mouse anti-CNP (SMI191; 1:1000) and rat anti CD11b (Biolegend 101205 1:50; both Cambridge Bioscience, Cambridge, UK), SMI31 mouse antibody to phosphorylated neurofilament (1:1500; Affiniti Research Products Ltd., Derbyshire, UK), rabbit anti-Caspr (1:1000; kindly provided by Professor E. Peles), rabbit anti-green fluorescent protein (ab6556; 1:1000; Abcam, Cambridge, UK), rat anti-lysosome-associated membrane glycoprotein 1 (1D4B; 1:2 Developmental Studies Hybridoma Bank, Iowa, U.S.A.), mouse anti-myelin oligodendrocyte glycoprotein, (1:200, clone Z2 provided by C. Linington), rabbit anti-ionized calcium binding adapter molecule 1 (Iba1; 019-19741; 1:800, Wako, Neuss, Germany). These were diluted in blocking buffer and incubations performed at RT for 1 hour or overnight at 4°C. Anti-CD45 worked best on acetone fixed cells (10 mins, -20°C). Bound antibodies were visualised using appropriate combinations of species/isotype-specific fluorochrome-conjugated secondary antibodies (1:400 [or in later experiments, 1:1000] 488 goat-anti mouse IgG1, catalogue number A2112; 568 goat anti-mouse IgG1, catalogue number A21124; 488 goat-anti-rat IgG, catalogue number A11006; 568 goat anti-rat IgG, catalogue number A11077; 488 goat-anti rabbit IgG, catalogue number A11008; 568 goat anti-rabbit IgG, catalogue number A11036; all Alexa Fluor, Life Technologies) after incubation at RT for 15–60 minutes. Nuclei were stained with DAPI (2 µg/ml) for 5 minutes, then wells of the multi-well microplates were filled with 100 μl PBS (although Mowiol also works well, and solidifies). Coverslips were mounted on glass slides in Citifluor AF1 mounting medium (Agar Scientific, Essex, UK, catalogue number AGR1320) or Mowiol 4-88 prepared as described in the supplementary protocol document, by dissolving 2.4 g Mowiol 4-88 (#81381, Sigma-Aldrich) in 6.0 g analytical grade glycerol, 6 ml distilled water, 12 ml 0.2 M Tris pH 8.5. Plates were imaged on an IN Cell Analyzer 2000 (GE Healthcare) and coverslips were imaged using wide-field fluorescence microscopy.\n\nThe IN Cell Analyzer 2000 (GE Healthcare) was used to acquire multiplexed wide-field fluorescent images in 96-well microplate format with a Nikon 10X 0.45NA Plan Apo objective, with the following channels 490/20 nm & 525/36 nm (green), 579/24 nm & 624/40 nm (red) and 350/50 nm & 455/50 nm (DAPI). Using these three channels, six fields of view were imaged per well. Flat Field Correction (illumination correction) was used as an image pre-processing step during image acquisition. Axonal density (area stained with antibody SMI31, anti-neurofilament), myelin (area stained with anti-MOG or anti-MBP) and cell counts (number DAPI +ve nuclei) were quantified using CellProfiler image analysis software version 2.1.0 (Carpenter et al., 2006). The pipelines developed for this study are available from GitHub. A supplementary protocol document is attached (Supplementary File 5). Briefly, images were coded with parental metadata (imageID, wellID) and also with row/column metadata. Quality control of the images was based on cell numbers, axon density and artefacts. Nuclei were identified in the DAPI image as primary objects after image thresholding using the Otsu global method. Shape was used to segment closely spaced cell. Images with less than 1500 nuclei/image were not used for subsequent analyses. Axonal density was measured in the red image (SMI31) after global thresholding using the Otsu, three classes weighted variance method. Total area occupied by phosphorylated neurofilament was measured and image sets with less than 40% ‘axonal density’ were not used for subsequent analyses. In the green channel (MBP), after applying an Otsu global threshold, myelin-like sheaths were identified using compactness ≥ 2.5 and form factor ≤ 0.2. Nuclei count, total area, axon area and myelin area measurements were exported to a comma-delimited spreadsheet for data analysis. The average value obtained from all images from each of the treatment conditions, from a single independent cell culture, were considered one independent experimental unit, unless otherwise stated in the Figure legends.\n\nTo quantify cellular parameters in myelin treated cultures, 10 images (selected at random on the blue [DAPI] channel) each of the green (CD45), red (rhodamine-labelled myelin) and blue channels were captured using a x20 objective over two 13 mm cover slips per condition, for each independent experiment, using a CCD camera system (Photonic Science Colour Coolview) and ImagePro 6.0 software (Media Cybernatics, Silver Spring MD). To calculate CD45 +ve cell density, manual counts of CD45 positive cells containing a DAPI-labelled nucleus were made from each image within an AOI of 124384 μm2 and the sum of the values were converted to cells per mm2. To quantify myelin uptake, the number of CD45 +ve cells containing rhodamine-labelled myelin per AOI was expressed as a percentage of all CD45 +ve cells per AOI. The experimenter was blinded to the ‘genotype’ of the myelin during cell quantification. Immunostained cultures were also imaged on a Zeiss Axioimager M2 wide-field microscope with Zen 2012 (Blue Edition) version 1.1.2.0 software.\n\nCytokine arrays were used to provide semi-quantitative data on cytokines and chemokines produced in response to damage and pathogen associated molecular patterns. Cultures were incubated with cell culture grade PBS, myelin of either genotype in PBS (0.1 mg myelin-protein ml-1 final concentration), or LPS (100 ng ml-1; E-coli mutant O111:B4; VWR). One ml conditioned media (CM) was collected at DIV 27, spun at 12470 g for 1 minute and the upper 800 μl was stored at -80°C until required. Semi-quantitative analysis of cytokine/chemokine levels was performed using Proteome ProfilerTM Array, Mouse Cytokine Array Panel A (ARY006; R&D Systems Europe Ltd., Abingdon, UK), according to manufacturer’s instructions. Briefly, nitrocellulose membranes, containing 40 anti-cytokine antibodies in duplicate, were blocked in 2 ml of blocking buffer (array buffer 6) on a rotating platform for one hour. CM from matched samples were probed in parallel. Eight hundred μl CM was added to 0.5 ml of array buffer 4 and then adjusted to 1.5 ml with array buffer 5. Fifteen µl of reconstituted Cytokine Array Panel A detection antibody cocktail, containing biotinylated antibodies, was added to each CM sample, for 1 h RT. After removal of array buffer 6, the 4 nitrocellulose membranes (1 per condition) were incubated in the sample/antibody mixture overnight at 4°C. Membranes were incubated in 1.5 ml of streptavidin-HRP (in Array Buffer 5) for 30 m at RT then treated with Pierce enhanced chemiluminescence western blotting substrate catalogue number 32106 (Perbio Science UK Ltd., Cramblington, UK) and subsequently exposed to x-ray film for between 1 and 20 minutes. X-ray films were scanned and made into digital images and the volume of each spot (minus the average edge volume) was quantified using array analysis software (TotalLab TL100 Array v2008; Newcastle upon Tyne, UK). For each membrane, the signal volume of each of two paired spots (representing a single cytokine) was normalised to the average spot volume of 6 positive control spots on the same membrane, and the average value of the paired spots, as a percentage of the control spot value, was derived. This method corrected for different incubations and/or exposures across different independent experiments. The experimenter was not blinded during this experiment or its analysis, but the automated quantification prevented experimenter bias. Medium from each of the four conditions, from a single independent cell culture, was considered one experimental unit.\n\nDissociated E13 spinal cord cells were established on MEAs (60MEA200/30iR-Ti-gr; Multi Channel Systems, Reutlingen, Germany) as described in ‘Myelinating Cell Cultures’. Electrophysiological recordings were performed from DIV 20-30. Sixteen channels were read simultaneously at 37°C in differentiation medium, in a custom-built laboratory MEA holder connected to a 16-channel amplifier (A-M Systems; Washington, USA). The array was subsequently rotated in the holder in order to read from all channels. Three minutes of extracellular recording were collected per channel (High Pass: 3 Hz; Low Pass: 500 Hz, Gain 20 k; Notch On). A semi-permeable membrane (fluorinated ethylene-propylene) was placed on top of the array, allowing cultures to be returned to the incubator and recordings to be made on subsequent days. Acquired data were analysed off-line. Modulators of neural activity were added directly to the bath: tetrodotoxin (TTX; 1 µM; Tocris Cat. 1069), cyanquixaline (CNQX; 5 µM; Tocris Cat. 1090), picrotoxin (100 µM, Sigma-Aldrich, Cat. P1675).\n\nFor single cell recording, whole-cell current clamp recording was performed using an Axopatch 200B amplifier with a Digidata 1440A digital acquisition system and pClamp 10 software. Experiments were performed at 37°C in atmospheric CO2 using an extracellular solution containing identical ionic concentrations to the cell culture media (in mM): 110.3 NaCl, 5.3 KCl, 1.8 CaCl2, 0.8 MgCl2, 10 HEPES, 25 glucose, pH 7.4. The pipette solution contained (in mM): 135 K-gluconate, 2 MgCl2, 2 Na-ATP, 0.5 Na-GTP, 10 HEPES, 0.5 EGTA, pH 7.2. Borosilicate glass pipettes were pulled to a resistance of 3-8 MΩ. TTX and other modulators (see above) were added directly to the tissue chamber.\n\nSpinal cord cells were plated on 35 mm diameter glass bottom Petri dishes (mentioned previously) and cultured for 20–22 days. Immediately prior to imaging, rhodamine-labelled myelin was vortexed and added to the Petri dish at final concentration 0.05 mg myelin-protein ml-1. The dish was set in a Nikon TE 2000 time-lapse microscope inside a temperature/CO2-controlled chamber. Using Metamorph 7.5.2 imaging software set for multi-stage positions and multiple wavelengths, positions of interest were selected on the bright field (phase), cherry (rhodamine) and CFP channels, and the co-ordinates recorded. Focus was set and maintained using a PSF perfect focus control. Images were taken every 15 minutes over a 15 h period. AVI videos of the stills collected were then generated using ImageJ 1.44 software.\n\nCell lysates from siRNA studies or spinal cord homogenates were prepared to 1 mg protein ml-1 in RIPA lysis buffer system with protease inhibitors (sc24948; Insight Biotechnology), plus 5x loading buffer (Sodium dodecyl sulphate/Dithiothreitol denaturing buffer [SDS/DTT]), and heated to 65°C for ten minutes. Lysates (see Table 1 for protein amounts) were run on a 4–12% gradient NuPAGE bis-tris acrylamide gel (Invitrogen, Paisley, UK) at 200 volts for 40–50 minutes and transferred using a semi-dry system to a PVDF membrane (Millipore, Watford, UK) at 225 mA for 1 h per 2 gels. PVDF membranes were blocked in 5% milk in Tris-buffered saline/0.01% tween (TBS/T; pH 7.4) and incubated overnight with primary antibody (Table 1) in blocking solution at 4°C, with gentle agitation. Following thorough washing, membranes were incubated in horseradish peroxidase (HRP)-conjugated goat anti-mouse or goat anti-rabbit secondary antibodies (New England Biolabs, Dundee, UK) in blocking solution (Table 1) for 1 h RT. Following thorough washing and incubation in Pierce enhanced chemiluminescence western blotting substrate (Thermo Fisher Scientific, catalogue number 32106)), the PVDF membrane was exposed to x-ray film (AGFA) for 30 sec to 20 min to obtain optimal exposures.\n\nDharmacon’s Accell siRNA passive delivery system including ‘smart pool’ siRNA (Perbio Scientific) for “difficult to transfect cells” (Mir & Le Breton, 2008; Shen et al., 2008), was used. Qualitative assessment indicated that cultures incubated in Accell DM supplemented with biotin (10 ng/ml final) and D-glucose (4500 μg/ml final) (referred to as Accell DM+), fared better than cells incubated in Accell DM alone, therefore siRNA was routinely added to the cultures in DM+. At DIV 20, cultures were treated with (i) Accell DM+ (ii) 1 μM non-targeting (NT) siRNA, (iii) 1 μM CyB siRNA or (iv) 1 μM Cnp1 siRNA (all ‘smart pool’) in Accell DM+ for 72 hours between DIV 20-23 and for a further 72 hours, between DIV 28-31. Between incubations with siRNA, cultures were fed as usual with differentiation media minus insulin. Cell lysate from each of the five conditions, from a single independent cell culture, was considered one experimental unit.\n\nStatistical analyses were made using GraphPad Prism versions 5.0, 6.0 or 7.0 for Windows (GraphPad Prism Software, San Diego, USA). Significance levels were set to p < 0.05. One-way ANOVA using Bonferroni’s Multiple Comparison Test was used to determine effects of siRNA, comparing all conditions to NT siRNA. For analysis of myelin uptake by CD45+ve cells between genotypes and across time points, a two-way ANOVA followed by Tukey’s multiple comparisons test was used. For analysis of steady state levels of 40 cytokines on proteome arrays, ‘multiple Student’s t tests’ were used to compare between myelin genotypes or between myelin genotype and PBS. Unless otherwise indicated, each n value derives from a single independent cell culture, comprising multiple wells or coverslips, representing technical replicates of the various treatments. In general, statistically tested experiments were carried out on 3–5 independent cell cultures, usually comprising 2 or more technical repeats (as indicated in Figure Legends or text). Up to 7 cultures per protein were examined by western blotting following treatment with siRNA, due to variation in the arbitrary units’ values from experiment to experiment.\n\n\nResults\n\nThe function and structure of CNS white matter is dependent on a complex interplay between axons, oligodendrocytes, astrocytes and microglia; including physical, metabolic, receptor-dependent signaling and gap junction-mediated cell-to-cell communication. Myelinated cultures derived from E13 mouse spinal cord, recapitulate this cellular complexity (Figure 1A–E), and by day in vitro (DIV) 24-28, comprise consecutive myelinated internodes separated by nodes of Ranvier (Figure 1F). To define this system in terms of function, we began by examining its electrical properties using microelectrode arrays (MEAs) (Figure 2A), each comprising 59 recording electrodes. The recordings obtained at a single electrode represent the combined extracellular responses of neurons and glia. Recordings taken daily between DIV 20 and 24, when myelination is occurring, demonstrated the gradual changes in electrical activity, starting with single isolated spikes and progressing to burst-like activity (Figure 2B). Additive pharmacology (sequential administration of drugs without washout) demonstrated distinct ‘network-like’ aspects of the neuronal activity (Figure 2C), not unlike the fictive bursting activity seen in isolated intact spinal cord segments (Grillner et al., 1981). Picrotoxin, a GABAergic inhibitor, evoked enhanced firing (second versus first trace) and subsequent addition of TTX (a voltage-gated Na+ channel blocker) blocked spike-like (presumably axonal) activity, whilst CNQX (blocker of AMPA/kainate receptors) further blocked (presumably) excitatory post-synaptic activity. The remaining slow potential changes (final trace) probably represent local spontaneous depolarizations at neuronal cell bodies and dendrites. To provide support for these conclusions we used single cell patch clamp recording. Spontaneous action potential generation was observed in 68 out of 75 cells tested; average spiking frequency being 2.13 ± 0.32 Hz (mean ± s.e.m.). The firing frequency was highly variable between cells, ranging from 0.07 to 15.27 Hz. This activity could be enhanced with 100 μM picrotoxin (Figure 2D) or completely blocked by 1 μM TTX (Figure 2D). In summary, spinal cord myelinating cell cultures are electrically active and activity can be manipulated bi-directionally using pharmacology.\n\nA. The morphology of CNP positive cells changes with time in culture. At 6 DIV, multiple cell process extend from the cell body and the cells have a ‘lacy’ appearance, as shown in the inset. By 12 DIV, a small number of myelin-like sheaths are present; often a single single sheath can be observed running in a line through one axis of an otherwise ‘lacy–appearing’ cell. In the inset, two such sheaths can be observed along 3 adjacent ‘lacy-appearing’ cells. By 21 DIV, many cells have extended multiple myelin-like sheaths and by 28 DIV few ‘lacy-appearing’ cells remain and a dense network of myelin sheaths are visible. When stained with antibodies to PLP/DM20, Caspr and phosphorylated neurofilament (NF), many axons are covered by consecutive sheaths with nodes of Ranvier straddled by Caspr positive paranodes (bottom right image). B. Iba1 +ve microglia are present at all stages examined. At 7DIV most microglia appear amoeboid but become ramified and extend multiple processes over time (insets). C. Antibody to NeuN labels neuronal nuclei and sometimes staining extends into the cytoplasm, but rarely enters the cell processes. D. GFAP positive astrocytes are found throughout the culture, (E) as are NG2 +ve OPCs. F. Combined staining with antibodies to PLP/DM20, Caspr and phosphorylated neurofilament (NF), reveals that many axons are covered by consecutive sheaths with nodes of Ranvier straddled by Caspr positive paranodes. Images A–F were contrast enhanced to ease viewing.\n\nA. Low power view of the whole MEA (left); all 59 recording electrodes and the ground electrode are concentrated in the center of the array. In the high-power image (right), four electrodes (black circles) can be observed in relation to the cell bodies (white arrows) and processes (including axons; blue arrowheads) of neurons and glia, on an MEA. To facilitate viewing, we selected a region in which cell density is lower than normal. Extracellular activity can be assessed from each of 59 recording electrodes. B. Extracellular recordings from a single electrode between 20 and 24. DIV. Each trace represents the sum of the activity detected at that electrode, which is in direct contact with multiple neuronal and glial cell bodies and processes. ‘Fast’ spike-like activity is evident at DIV 21 and ‘burst-like’ activity develops by DIV 24. C. Additive pharmacology on a DIV 24 culture (sequential administration of drugs without washout) demonstrates in vivo-like neuronal network activity of spinal cord myelinating cultures. The recording shows that addition of picrotoxin, a GABAergic inhibitor, evokes a massive increase in ‘fast’ neuronal activity (second versus first trace). Subsequent addition of TTX (a Na+ channel blocker) inhibits much of the ‘fast’ spike-like (presumably axonal) activity. CNQX (a blocker of AMPA/kainate receptors) blocks excitatory synaptic activity. The remaining slow potential changes probably represent local spontaneous depolarizations (at neuronal and glial cell bodies). D. Examples of whole-cell current clamp recordings from single cells, presumably neurons, firing spontaneously at a resting potential of around -50 mV. Top, trace before and 5 minutes after the addition of 1 µM TTX (top). Bottom, trace from a second cell before and 5 minutes after the addition of 100 μM picrotoxin. TTX consistently blocked activity (n = 6 cells), as expected, whilst picrotoxin caused regular burst spiking in every cell tested (n = 18 cells), but the effect on overall spike rate was inconsistent, presumably reflecting the phenotype of the cell.\n\nNext, we examined functional responses of microglia/macrophages, which were evident as IBA1 (Figure 1B) or CD45 +ve cells, at all stages examined from DIV 7-28. To assess the phagocytic capability of microglia/macrophages for a relevant target, we added increasing concentrations of rhodamine-labelled myelin (subsequently referred to as ‘myelin debris’) on DIV 26 for 24 h. The proportion of CD45 +ve cells incorporating myelin reached ~66% of cells at 0.075 or 0.1 mg myelin-protein ml-1. Internalized myelin appeared within lysosomal-associated membrane glycoprotein 1 +ve structures, indicating its incorporation into late endosomes/lysosomes (Figure 3A). To assess consequences of longer-term exposure to myelin debris, we incubated cultures for 1 or 7 days with myelin (0.1 mg myelin-protein ml-1) from wild type or Plp1 transgenic mice (Plp1 tg line #72; Readhead et al., 1994); the latter is a spontaneously demyelinating model of Pelizaeus Merzbacher disease (Anderson et al., 1999; Edgar et al., 2010) that like line #66 (Ip et al., 2006) is characterized by low-level T cell infiltration into the CNS (JE unpublished observations). The proportion of CD45 +ve cells containing myelin debris did not change significantly over time, or with the ‘genotype’ of the myelin (Figure 3B). To compare receptor-dependent versus independent phagocytosis (Sierra et al., 2013), we quantified the proportion of cells incorporating myelin versus latex beads; however the proportions were similar (Figure 3B).\n\n(A) ID4B antibody staining of cultures treated with rhodamine-labelled myelin debris at 27 DIV shows myelin debris (red) within lysosomal associated membrane glycoprotein 1 (green) late endosomes/lysosomes. All cells containing myelin were found to be CD45 +ve (not shown). (B) After incubation with saturating amounts of myelin debris (data obtained with 0.075 or 0.1 mg myelin protein ml-1 were similar, so values from each were included providing two technical replicates) over three different time periods, the percentage of CD45 positive cells that contained ‘wild type’ (pale grey bars) or Plp1-tg (dark grey bars) myelin debris, which is taken up in a receptor-dependent manner, was between 40–70 % (n = 6; 2 technical replicates for each of 3 independent experiments, per time point); not significantly different from the proportion that contained latex beads (black bar), which are taken up in a receptor-independent manner (n = 3 independent experiments). Bars represent mean +/- s.e.m.. (C) Time lapse stills over 1130 minutes show myelin debris (red) at time of addition (25 mins – 275mins) and then being phagocytosed by microglia/macrophages (440 mins – 1130 mins). Blue and green arrows highlight the locations of individual cells over time.\n\nIn the resting brain in vivo, microglial soma are rather stationery but their processes are highly motile (Davalos et al., 2005; Nimmerjahn et al., 2005). In response to focal injury, or simulation of injury by local application of ATP, cell soma (Honda et al., 2001) or processes independent of the cell soma (Davalos et al., 2005), converge on the injury site. To assess the dynamic behavior of phagocytic microglia/macrophages in spinal cord-derived cultures, we used live imaging immediately following the addition of rhodamine-labeled myelin debris. Cells identified as microglia/macrophages on the basis of their incorporation of myelin debris, were motile with respect to both the soma and processes, appearing to ‘gather’ myelin debris for ingestion. Qualitatively, the vast majority of exogenous myelin was internalized within the 15 h imaging period (Figure 3C and Supplementary Video 1).\n\nIn vivo, microglia not only express receptors that mediate phagocytosis of cellular debris, but also initiate responses to pathogen-associated molecular patterns (PAMPs) such as endotoxins and viral nucleic acids, and damage-associated molecular patterns (DAMPS) such as high mobility group box 1 (HMGB1) and ATP. To examine the secretory profile of spinal cord-derived cultures challenged with (i) lipopolysaccharide (LPS; from Gram-negative bacteria; 100 ng ml-1 final) or (ii) myelin debris (0.05 mg myelin-protein ml-1 final), we used a commercially available proteome array for the unbiased detection of 40 cytokines/chemokines. In preparing myelin debris, we took care to minimize the risk of contamination by endotoxin, by using cell-culture grade plastics and media wherever possible. As expected, LPS led to a significant increase in steady state levels of pro-inflammatory cytokines, TNFα, IL1-α, IL-6 (all p < 0.001), IL1β (p < 0.01); anti-inflammatory cytokine IL1-ra (p < 0.01); granulocyte colony stimulating factor (G-CSF; p < 0.001); chemokines CCL1, CCL3, CCL4, CCL5, CCL12, CXCL1, CXCL2, CXCL10 (all p < 0.001), CXCL9 (p < 0.05); soluble intracellular adhesion molecule 1 (sICAM1; p < 0.001) and TIMP Metallopeptidase Inhibitor 1 (TIMP1; p < 0.001), compared to PBS (n = 3-4 replicates from 2-3 independent cultures). A representative proteome array is shown in Figure 4A. In contrast, exposure of the cultures to myelin debris resulted in a much more restricted response, lacking upregulation of pro-inflammatory cytokines. However, chemokines CCL3 and CCL4 were significantly upregulated in cultures treated with either wild type or Plp1 tg myelin, compared to vehicle-treated controls (PBS; Figure 4B and C), as were TIMP1 and M-CSF (p < 0.001–0.05; n = 3 independent cell cultures and 6 independent myelin preparations [3 WT and 3 Plp1 tg]). Notably however, compared to wild type myelin debris, Plp1 tg myelin debris was associated with significantly increased steady state levels of CXCL1, CXCL2, CXCL10 and CCL5 (Figure 4B and C), the last two of which are strongly linked to T cell infiltration into the CNS (reviewed Hughes & Nibbs, 2018).\n\nA. Map of proteome arrays shows six positive control (A1,2; A23,24; F1,2) and two negative control (F23,24) spots, alongside 40 paired cytokine-specific spots (B1,2-E7,8). The array on the right was incubated with medium from cultures treated with LPS. A number of cytokines and chemokines (highlighted with white boxes), including pro-inflammatory cytokines, were upregulated in LPS-treated cultures compared to PBS-treated controls. Quantification is provided in the text. B. Proteome arrays incubated with medium from cultures treated with a myelin-enriched tissue fraction from wild type or a spontaneously demyelinating model of Pelizaeus Merzbacher disease (Plp1 tg myelin) for 1 or 7 days. After 7 days, two chemokines were upregulated in cultures treated with wild type or Plp1 tg myelin compared to PBS-treated cultures (highlighted with boxes with broken lines). An additional four chemokines were upregulated specifically in Plp1 tg myelin-treated cultures compared to wild type myelin-treated cultures (highlighted with boxes with solid lines). C. Graph of chemokines upregulated in cultures treated with myelin-enriched tissue preparations for 7 days. Bars represent mean +/- s.e.m. n = 3 independent cultures to which 6 independent myelin preparations were added (1 preparation per independent experiment from wild type mice and 1 preparation per independent experiment from Plp1 transgenic mice). Each cell culture (considered n = 1 independent biological repeat) comprises up to 24 × 35 mm dishes each containing 3 coverslips coated with cells. Eight 35 mm dishes are contained within one 25 mm diameter dish. Wild type or Plp1 transgenic myelin were added to two 35 mm dishes, from within one larger 25 mm diameter dish. There was no systematic selection process for deciding which dishes would receive wild type versus Plp1 transgenic myelin, as all dishes were identical. Raw data files attached. * p < 0.05, ** p < 0.01, *** p < 0.001.\n\nTogether, these data demonstrate that innate immune cell properties, including phagocytosis of tissue debris and appropriate cytokine secretion profiles are reproduced in pure spinal cord-derived myelinated cultures.\n\nHaving established these cultures recapitulate relevant in vivo functions, we next asked if they are amenable to manipulation using siRNA and thus capable of being used as a screen for gene silencing approaches. We selected 2’, 3’-cyclic nucleotide 3’-phosphodiesterase (CNP), which is expressed early in the oligodendroglial lineage and eventually localized to the cell soma, processes and non-compact myelin (Trapp et al., 1988). We began by defining the temporal sequence of CNP expression using immunocytochemistry. We observed a small number of CNP +ve cell bodies as early as DIV 2, and by DIV 4, CNP was also observed in short cellular processes (data not provided). By DIV 6, multiple ‘lacy-appearing’ CNP +ve cells were visible (Figure 1A) and by DIV 12, occasional myelin-like profiles were present; often as a single myelin-like sheath extending from an otherwise ‘lacy-appearing’ cell (Figure 1A). Myelin-like profiles increased in number and extent with time, and the number of ‘lacy-appearing’ cells correspondingly diminished (Figure 1A).\n\nTo deplete Cnp1 mRNA, cultures were incubated intermittently with siRNA from DIV 20, when CNP was already expressed (Figure 1A and Figure 5A), until DIV 31. CyB siRNA was used as a positive control. By western blot, steady-state levels of CNP or cyclophilin B were significantly reduced by their respective siRNAs, compared to non-targeting (NT) siRNA-treated controls (Figure 5A and B). In contrast, steady-state levels of proteolipid protein (PLP), myelin associated glycoprotein (MAG), actin, cyclophilin B or phosphorylated neurofilament heavy chain (Ph NF-H), were not altered in CyB or Cnp1 siRNA-treated cultures compared to NT siRNA controls. With the exception of MAG, levels of all proteins were similar in NT siRNA-treated cultures compared to cultures treated only with Accell DM+ incubation medium (Figure 5A and B). Qualitatively, neurofilament medium chain (NF-M), acetylated tubulin, tyrosinated tubulin (Figure 5A) and mitochondrial-related proteins, OPA1, mitofusin and parkin (data not in the manuscript, but attached in raw data files, Exps N10, 12, 13, 23, 24 and 30, Dataset 1, (Bijland et al., 2019)) were also not obviously changed under any treatment conditions, suggesting siRNA is not toxic to these cell cultures, even after relatively long incubation times. Using immunocytochemistry with antibodies to CNP (in non-compact myelin), GFP (to stain CFP +ve neurones from the Thy1-CFP mouse embryos) and myelin basic protein (MBP; in compact myelin), CNP specifically appeared reduced in cell bodies and myelin sheaths in Cnp1 siRNA-treated cultures compared to DM+ only or CyB siRNA-treated controls (Figure 5C and D).\n\nA. Western blots of (from L to R) spinal cord lysates from wild type and Cnp1 knockout (ko) mice and cell lysates from DIV 30 spinal cord cultures following treatment with supplemented incubation medium only (DM+), Cnp1 siRNA, CyB siRNA or non-targeting (NT) siRNA. Lysates from time 0 (DIV 20) cultures, were run alongside. B. Quantification myelin proteins, housekeeping proteins and axonal proteins in siRNA treated cultures. Bars represent mean +/- s.e.m. Values were obtained from n = 4–7 independent cultures. Seven independent experiments were undertaken but some proteins were only examined in 4, 5 or 6 of these as shown in the attached raw data files. All 7 experiments were probed for CNP and 6 of 7 were probed for cyclophilin B, against which the siRNA was designed. * p < 0.05, ** p < 0.01. C. Micrographs of cultures prepared from Thy1-CFP mouse embryos, treated with DM+, CyB siRNA or Cnp1 siRNA, and stained with anti-GFP (binds CFP which is expressed in a subset of neurons) and anti CNP. White arrows point to CNP +ve oligodendrocyte cell bodies or CFP +ve neuronal cell bodies. The punctate ‘staining’ indicated with broken yellow arrows represents autofluorescent lipofuscin granules, which are prevalent in these cultures. CNP is markedly reduced in Cnp1 siRNA-treated cultures. D. Micrographs of cultures treated with DM+, CyB siRNA or Cnp1 siRNA, and stained with anti-MBP (MBP is located in compact myelin) and anti-CNP (CNP is located in non-compact myelin). CNP is markedly reduced in Cnp1 siRNA-treated cultures, but MBP appears unchanged.\n\nIn summary, this myelinating cell culture system is amenable to siRNA gene expression knockdown, in the absence of overt off-target effects.\n\nWe have previously used this culture system to identify factors that influence myelination, however to increase its utility in this respect we asked if it could be adapted it to a format that would allow semi-high throughput screening of multiple factors, simultaneously.\n\nEmpirically, we have observed that poly-L-lysine in boric acid buffer is a highly effective substrate for cell adhesion, therefore 96-well microplates were prepared accordingly, and E13 mouse spinal cord cells were plated at a density of 75,000, 85,000 or 100,000 cells/well. Using immunocytochemisty to label myelin sheaths (anti-MBP), axons (SMI31, anti-phosphorylated NF-H) and oligodendroglia (anti-Olig 2), we found that a starting density of 100,000 cells/well led to a reproducible generation of myelinated axons after 24 DIV (data not shown). Qualitatively, plating densities of <100,000, led to lower levels of neuritic outgrowth and myelination.\n\nAs most small molecule libraries are diluted in DMSO, we checked if DMSO adversely affected the cultures. Growing the cells in the presence of 1% DMSO had no significant effect on axon density or myelination (mean difference 2.4 % for axon density, mean difference 0.4 % for myelin, (see raw data files pertaining to Figure 6 and Figure 7, Dataset 1, (Bijland et al., 2019)).\n\nA. Representative immunofluorescence images of 96-well microplate myelinating cultures (28 DIV) stained for MBP, phosphorylated neurofilament (SMI31) and nuclei (DAPI). Cells were untreated or treated with Activin-A or FGF9 as described. B. Percentage change in myelination, axon density and nuclei count after treatment with Activin-A or FGF9. Bars represent average values from 5 independent cultures, each with 5 wells of technical replicates (wells), +/- s.e.m.. Raw data attached. Cells were allocated to different treatments as described in Materials and Methods. * p < 0.05, **p < 0.001.\n\nSpnal cord cultures were treated with increasing doses of Activin-A. A. Representative immunofluorescence images of 96-well microplate myelinating cultures (28 DIV) stained for MBP and phosphorylated neurofilament (SMI31). B. Dose-response curve for myelination in response to Activin-A treatment. Average values from 2 independent cultures, 5 technical repeats per culture, +/- s.e.m.\n\nNext, we established a protocol for automated image capture using the IN Cell Analyzer 2000 High Content Screening microscope (GE Healthcare), and an automated image quantification method using CellProfiler (Carpenter et al., 2006). Our analysis protocol quantified nuclear number, phosphorylated neurofilament (axonal) area and PLLS (power log-log slope, as an indicator of amount of blur in the image) and excluded images with nuclei numbers < 1500, axon area < 40% or PLLS > -2.0. For each image that passed quality control, the area occupied by myelin-like sheaths (stained with anti-MBP) and the area occupied by phosphorylated neurofilament (stained with antibody SMI31) were quantified, and MBP area expressed as a proportion of neurofilament area.\n\nTo test if our automated image capture and quantification protocols were sufficiently robust to detect changes in myelination, we treated myelinating cultures from day 19 with FGF9, a known mitogen and inhibitor of myelination (Lindner et al., 2015), or from day 15 with activin A, a known enhancer of myelination (Goebbels et al., 2017; Miron et al., 2013) until DIV 28. FGF9 induced a significant (-31.7%, p < 0.05) decrease in myelination while activin A (100 ng ml-1) produced a significant (+56.5 %, p < 0.001) increase in myelination, relative to vehicle-treated cultures (Figure 6). Furthermore, FGF9 (100 ng ml-1) led to a significant (+53,8 %, p < 0.001) increase in nuclear density (average of 5 independent experiments; each with 5 technical repeats), as expected for a pro-proliferative factor (Lindner et al., 2015). To test the sensitivity of our assay, we treated cultures with 1, 5, 10, 50 or 100 ng ml-1 activin A and found that increasing concentrations (1-100 ng/ml) of activin A led to a dose-dependent increase in the degree of myelination (Figure 7; average of 2 independent experiments; each with 5 technical repeats).\n\nThus, the methods are robust and sufficiently sensitive to detect small changes in myelination and cell densities.\n\nAs libraries of small molecules are costly, we proceeded to test whether myelinating cultures could be grown in 384-well microplates, which would allow markedly smaller amounts of compounds to be used in each assay. We plated spinal cord cells at densities of 30,000, 40,000 or 50,000 cells/well and stained the cultures with antibodies to phosphorylated neurofilament and MBP. Qualitatively, we found that the optimal plating density was 50,000 cells/well but the cultures tended to be less reproducible in this format.\n\n\nDiscussion\n\nHere we demonstrated that a murine, pure spinal cord-derived cell culture system (Thomson et al., 2006; Thomson et al., 2008) recapitulates several in vivo cellular functions, namely, spontaneous spinal-like electrical activity; phagocytosis of tissue debris; and appropriate secretory responses to diverse stimuli. Further, the system, which is straightforward to establish, can be used to assay gene-silencing approaches; adapted to a 96-well format for semi-high throughput assays of multiple factors simultaneously; or for growth on MEAs to monitor electrical properties. These advances increase the utility of this system for studies to address white matter development, function and pathology, providing the research community with a robust experimental tool that reduces costs and minimises the use of experimental animals. In particular, the multi-well format allows large numbers of factors to be assayed in parallel, whilst simultaneously reducing the numbers of mice required to provide tissue.\n\nWhilst this in vitro model of CNS white matter is not intended to replace all animal work, it can inform, or augment such experiments. For example, we previously identified pro-myelinating factors from a list of candidates generated from transcriptomic analyses of laser-captured CNS tissue from Pten conditional knockout and control mice, using 11 independent cell cultures (Supplementary Figure 8 h; Goebbels et al., 2017) from 11 pregnant dams. A similar in vivo experiment would first require invasive surgery to administer a demyelinating agent into the CNS, followed by subsequent administration of the pro-myelinating compounds by the same route. Given that 8 compounds were analysed (Goebbels et al., 2017), a similar in vivo study would require the same 8 compounds to be administered to a least 11 mice per treatment group, assuming variation similar to that observed in cell culture; requiring at least 88 experimental animals. In this instance, our in vitro model affords an 87.5% reduction in the number of animals required compared to a similar in vivo experiment.\n\nIn CNS white matter, a supra-threshold axon caliber is necessary for initiation of myelination (Mayoral et al., 2018) and, when experimentally induced, is associated with de novo myelination of axons that ordinarily, are not myelinated (Goebbels et al., 2017). Myelin formation is further modulated by neuronal signals, including activity-dependent signals (Gibson et al., 2014; Hines et al., 2015; Mensch et al., 2015; Mitew et al., 2018). In neurons themselves, action potential firing patterns modulate the abundance of neuronal mRNAs across many functional gene categories affecting, for example, neurite outgrowth, synaptic connections and neural network formation (Lee et al., 2017), whilst conversely, protein levels can modulate neuronal activity (Lopez de Armentia et al., 2007). Thus, the circular relationship between fast electrical responses and longer-term changes in gene expression is important for nervous system development and plasticity (Lee et al., 2017). Here we demonstrated that spinal cord-derived myelinating cultures are spontaneously electrically active, and likely subject to similar bidirectional regulation of gene expression in neurons and in oligodendrocytes.\n\nNeural electrical activity readouts provide information on factors such as cell phenotype, metabolic status and health. As few laboratories are equipped with the necessary equipment and expertise to perform intracellular recordings, we reasoned that commercial MEAs might provide an accessible method for assaying global extracellular electrical responses. Using MEAs we identified fast spike-like, low amplitude extracellular activity that increased in frequency in the presence of the GABA receptor antagonist, picrotoxin, suggesting that inhibitory neurons suppress this spontaneous activity. Conversely, spike-like activity was reduced by the voltage-gated sodium channel blocker, TTX, indicating it represents axonal action potentials. The AMPA receptor antagonist, CNQX, blocked some of the TTX-resistant activity, suggesting it represents changes at dendrites brought about by the synaptic release of glutamate. This is compatible with ultrastructural observations in these cultures (Figures 6F and G; Thomson et al., 2008). To validate these MEA data, we used single cell current clamp recording which confirmed that high frequency action potential activity was present and could be enhanced by picrotoxin, or blocked by TTX (Figure 2) or CNQX (raw data provided separately, Dataset 1, (Bijland et al., 2019)); the last demonstrating it is glutamate-dependent. However, it is important to bear in mind that the extracellular activity recorded on the MEAs represents the sum of all neuronal and glial electrical responses at the recording electrode.\n\nIn terms of ease of use, spinal cord cells adhered well to MEAs and appeared, under phase contrast microscopy, indistinguishable from cells grown on glass. MEAs do not require the same specialist expertise or equipment needed for intracellular recordings and have the advantage that electrical activity can be assessed sequentially over days or even weeks. Further, they provide an overview of activity that cannot be achieved easily using intracellular recordings. Consequently, MEAs increase the utility of this multi-cell type culture for studies such as ion channel screening, drug testing and safety pharmacology.\n\nIn response to infectious, autoimmune or physical insults, microglia mount neuroinflammatory responses to combat pathogens and enhance repair and restoration of function (reviewed in Song & Colonna, 2018b; Song & Colonna, 2018a). However, this response must be finely balanced as excessive or chronic neuroinflammation can, in some circumstances, contribute to disease pathogenesis (Song & Colonna, 2018b; Song & Colonna, 2018a; Song et al., 2017). Pattern recognition receptors (PRRs) on microglia play a major role in the initiation of these responses. Originally discovered for their role in recognizing pathogen-derived molecular signatures, PRRs also respond to damage associated molecular patterns (DAMPs) released by dying or damaged cells, such as ATP, high mobility group box 1 protein (HMGB1), beta amyloid and α-synuclein (Kigerl et al., 2014). Chronic generation of DAMPs within the CNS can therefore run the risk of causing secondary, collateral tissue damage by sustaining long-term activation of pro-inflammatory responses (reviewed in Song et al., 2017).\n\nTo determine if spinal cord-derived cultures respond appropriately to a classical PAMP, we challenged them with bacterial lipopolysaccharide (LPS) and compared their response to that induced by myelin debris; sterile clearance of which is essential to promote lesion repair in demyelinating diseases such as MS and experimental models (Kotter et al., 2005; Napoli & Neumann, 2009; Neumann et al., 2009) in a CX3CR1-and TREM2-dependent manner (Cantoni et al., 2015; Lampron et al., 2015; Poliani et al., 2015). LPS is a well-characterized PAMP that activates toll-like receptor 4 (TLR4) to up-regulate expression of pro-inflammatory cytokines and chemokines by a wide variety of immune cells (Chow et al., 1999).\n\nAs the resident immunocompetent population of the CNS, microglia express multiple toll-like receptors, including TLR4 (Bsibsi et al., 2002; Zhang et al., 2014) and initiate inflammation in response to PAMPs by secreting pro-inflammatory factors. These include cytokines such as tumor necrosis factor (TNF)-α, interleukin (IL)-1β, and IL-6 and free radicals such as nitric oxide (NO) (Hanisch & Kettenmann, 2007; Hines et al., 2013; Tambuyzer et al., 2009). This was replicated in spinal cord myelinated cultures, in which LPS increased secretion of pro-inflammatory cytokines (IL-1α, Il1β, IL-6 and TNFα, as well as CCL and CXCL chemokines, mimicking the CNS response induced by viruses, bacteria or parasites (Forrester et al., 2018; Lobo-Silva et al., 2016). In striking contrast, clearance of ‘wild type’ myelin debris in these cultures did not induce an overt pro-inflammatory response, but selectively increased secretion of CCL and CXCL chemokines. This is in agreement with the homeostatic roles of microglia (and astrocytes) associated with clearance of myelin and other cellular debris injury (Perry et al., 1985).\n\nConspicuously however, challenging cultures with Plp1-tg myelin resulted in additional up regulation of CCL5 and CXCL10. Whilst chemokines facilitate context-dependent migration of all immune cells, CCL5 and CXCL10 and their receptors are particularly relevant with respect to T cell recruitment into the brain (Hughes & Nibbs, 2018). These observations are especially interesting in light of the fact that increased numbers of T lymphocytes are present in the CNS of spontaneously demyelinating Plp1 tg mouse models of Pelizaeus Merzbacher disease (PMD), caused by gene duplication (JE, unpublished observations and Ip et al., 2006). CXCL10 expression is induced by type I interferons to recruit T lymphocytes into the CNS to protect against neurotropic viral infections (Liu et al., 2000) and it and its receptor, CXCR3, are necessary for induction of experimental cerebral malaria (ECM); probably due to the critical role CXCL10 plays in T cell-endothelial cell adhesion, and injury of the brain endothelium (Sorensen et al., 2018). Further, the increased susceptibility to encephalitic symptoms after infection with West Nile virus, of individuals harbouring mutated CCR5, which encode the receptor for CCL5, is thought to be due to failure of trafficking of T cells into the brain (Glass et al., 2005) and reviewed in (Hughes & Nibbs, 2018). The molecular basis for the difference in chemokine induction between the two myelin ‘genotypes’ is not known; although our working hypothesis is that myelin isolated from Plp1 tg mice is subtly altered allowing it to cross-link or activate additional receptors at the microglial surface (Lennartz & Drake, 2018).\n\nThe importance of replicating the multi-cell type environment of the CNS when studying microglial functions such as these is highlighted by the fact that microglial phenotypes are modulated by neighboring cells and vice versa. For example, LPS-induced secretion of IL-1β and TNF by microglia can result in potent induction of pro-inflammatory gene expression by astrocytes in vitro (reviewed in Saijo & Glass, 2011), whilst signaling induced by microglial CD172, CD200R and CD45 interacting with CD47, CD200 and CD22, respectively at the neuronal cell surface, inhibits microglial activation (reviewed in Saijo & Glass, 2011).\n\nNeurodegenerative diseases, including those that predominantly affect white matter, such as the leukodystrophies and amyotrophic lateral sclerosis (ALS) remain largely untreatable or relatively refractory to currently available therapies. In both cases, a dominant gain-of-function effect of the protein products of mutated genes contribute to pathogenesis in some forms of these diseases, for example in familial amyotrophic lateral sclerosis (fALS) due to mutation in SOD1 gene and in Pelizaeus Merzbacher disease (PMD) due to point mutation or duplication of the PLP1 gene (reviewed in van Zundert & Brown Jr., 2017; Nave & Griffiths, 2004). Thus, gene silencing using siRNA or antisense oligonucleotides represents a rational approach to treatment that is potentially devoid of some of the problems caused by the broad modes of actions of small molecule-based drugs (Zheng et al., 2018). A number of issues related to the systemic administration of siRNA or antisense oligonucleotides for CNS disorders remain to be overcome, such as delivery in the blood circulation, passage across the blood-brain barrier, and targeting to the appropriate cell type (Zheng et al., 2018). Nonetheless, it is important to have relevant in vitro models to carry out preliminary tests of the efficiency of gene knockdown and to screen for off-target effects before proceeding to in vivo studies. This is particularly important for genes expressed in post-myelination oligodendrocytes; cells whose phenotype is altered when they wrap axons.\n\nUnlike PMD and fALS, which are genetically determined, MS is an acquired inflammatory demyelinating disease, affecting more than 2 million people worldwide (World Health Organization, 2008). MS involves a neurodegenerative process driven in part by the failure of MS lesions to remyelinate (Franklin & Ffrench-Constant, 2008; Franklin et al., 2012; Trapp & Nave, 2008). Consequently, a potential therapeutic approach is to stimulate remyelination; an effect that enhances functional recovery (Duncan et al., 2009) and is predicted to reduce ongoing axonal loss (Irvine & Blakemore, 2008; Mei et al., 2016). Spontaneous remyelination in MS is incomplete, leaving many axons chronically devoid of myelin (Patani et al., 2007). This is thought to render them vulnerable to inflammatory insult and devoid of oligodendroglial-mediated support (Franklin et al., 2012; Nave & Trapp, 2008; Trapp & Nave, 2008). In the adult brain, new myelin sheaths are formed by OPCs (Crawford et al., 2016; Keirstead & Blakemore, 1997; Zawadzka et al., 2010), which are often abundant in MS lesions but fail to myelinate the naked axons (Chang et al., 2000; Chang et al., 2002; Scolding et al., 1998). Recent studies are beginning to identify drugs, small molecules and pathways that can be modulated to overcome remyelination failure. These include Wnt/β-catenin (Fancy et al., 2009), retinoid X receptor gamma (RXRγ) signalling (Huang et al., 2011), muscarinic receptor antagonists such as benztropine and clemastine (Deshmukh et al., 2013; Mei et al., 2014) activin A (Goebbels et al., 2017; Miron et al., 2013) and drugs including miconazole and clobetasol (Najm et al., 2015). However, there is still a need for a cost-effective screening strategy to identify molecules with therapeutic potential for the treatment of MS and other multifactorial neurodegenerative disorders.\n\nA valuable pre-in vivo screen of candidate therapeutics should fulfil certain criteria. First, it should contain all major cell types of the CNS, to resolve off-target effects of the test agent. Second, it should be simple to establish, reproducible, sensitive and unbiased. Third, it should be unaffected by carriers such as DMSO. Fourth, it should be capable of screening a substantial number of agents while using the minimum number of experimental animals. Finally, it will provide added-value if it can be generated from mouse models of disease and/or transgenic reporter mice for live imaging or expression analysis. Here we demonstrated murine spinal cord-derived cultures meet these criteria for semi-high throughput screens and recapitulate relevant functional properties of CNS white matter. Given the ease with which they can be established, this system provides laboratories with a simple, functional and relatively inexpensive method to explore normal and pathological processes relevant to white matter in general, and spinal cord in particular.\n\n\nData availability\n\nUnderlying data is available from F1000Research\n\nF1000Research: Dataset 1. An in vitro model for studying CNS white matter: functional properties and experimental approaches, https://doi.org/10.5256/f1000research.16802.d233269 (Bijland et al., 2019)\n\nData is available as a zip file contain the corresponding data for the following figures:\n\nFigure 1- Raw microscope image files\n\nFigure 2B, C and D and for CNQX treated cell (not shown in Figure 2) – electrophysiology recording values.\n\nFigure 3B – images and quantification of densities of CD45 +ve cells and the proportions that contain wild type, Plp1 transgenic myelin or latex beads.\n\nFigure 4C – scans of X-ray films of proteome arrays following treatment for 7 days with PBS, latex beads, wild type myelin or Plp1 tg myelin, plus raw values in arbitrary units and normalised values (to positive control spots) for spot intensity for each of the cytokines assayed, plus a template for cytokine array.\n\nFigure 5 - scans of X-ray films and raw arbitrary unit values of band sizes and intensities.\n\nFigure 6 and 7 – images from IN Cell Analyzer 2000 and CellProfiler values for myelin, axons and DAPI +ve nuclei.\n\nDue the size of files for the microplate images these can not be provided but are available at request from the corresponding author (JE) julia.edgar@glasgow.ac.uk\n\n\nSoftware availability\n\nAnalysis pipelines are available from Github: https://github.com/muecs/cp/tree/v1.0\n\nAn archived version available of the pipelines are available from Zenodo: http://doi.org/10.5281/zenodo.2533339 (Mücklisch & Mücklisch, 2019)\n\nAvailable under a 'Creative Commons Attribution 3.0 Unported License'",
"appendix": "Grant information\n\nThis work was supported by the National Centre for the Replacement, Refinement and Reduction of Animals in Research (NC3Rs) [grant Ref. B004-13.1 to JE, ERK and CL] and the Multiple Sclerosis Society, UK [grant ref. 853 to JE and JTA; grant ref. 918 to JE, MM and JTA; and grant ref. 991 to JE and ERK].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgement\n\nA special mention goes to Duncan and Yvonne Booth and their 10in10 event which raises fund for the Multiple Sclerosis Society, UK. We are grateful to Profs. Nigel Groome and Elior Peles for gifts of antibodies; to Professor Klaus-Armin Nave for transgenic animal lines Plp1 transgenic (#72) and Cnp1 knockout; to Professor Robert Nibbs for helpful scientific discussion; and to Professor Kurt Anderson for access and training in live-imaging.\n\n\nSupplementary material\n\nDetailed protocols are provided in the following Supplementary Files:\n\nSupplementary File 1. Preparing mouse spinal cord cultures\n\nClick here to access the data\n\nSupplementary File 2. Preparing custom-made dishes for live imaging\n\nClick here to access the data\n\nSupplementary File 3. Preparing and labelling a myelin-enriched spinal cord fraction\n\nClick here to access the data\n\nSupplementary File 4. Immunostaining myelinating cell cultures\n\nClick here to access the data\n\nSupplementary File 5. Using CellProfiler analysis software\n\nClick here to access the data\n\nSupplementary Video 1: Live imaging of myelin phagocytosis\n\nClick here to access the data\n\n\nReferences\n\nAllen NJ, Lyons DA: Glia as architects of central nervous system formation and function. Science. 2018; 362(6411): 181–185. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlmeida RG: The Rules of Attraction in Central Nervous System Myelination. Front Cell Neurosci. 2018; 12: 367. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlmeida RG, Lyons DA: On Myelinated Axon Plasticity and Neuronal Circuit Formation and Function. J Neurosci. 2017; 37(42): 10023–10034. PubMed Abstract | Publisher Full Text\n\nAnderson TJ, Klugmann M, Thomson CE, et al.: Distinct phenotypes associated with increasing dosage of the PLP gene: implications for CMT1A due to PMP22 gene duplication. Ann N Y Acad Sci. 1999; 883: 234–246. PubMed Abstract | Publisher Full Text\n\nBaker D, Amor S: Experimental autoimmune encephalomyelitis is a good model of multiple sclerosis if used wisely. Mult Scler Relat Disord. 2014; 3(5): 555–564. PubMed Abstract | Publisher Full Text\n\nBarnett SC, Linington C: Myelination: do astrocytes play a role? Neuroscientist. 2013; 19(5): 442–50. PubMed Abstract | Publisher Full Text\n\nBarres BA, Hart IK, Coles HS, et al.: Cell death and control of cell survival in the oligodendrocyte lineage. Cell. 1992; 70(1): 31–46. PubMed Abstract | Publisher Full Text\n\nBarres BA, Raff MC: Control of oligodendrocyte number in the developing rat optic nerve. Neuron. 1994; 12(5): 935–942. PubMed Abstract | Publisher Full Text\n\nBechler ME, Byrne L, Ffrench-Constant C: CNS Myelin Sheath Lengths Are an Intrinsic Property of Oligodendrocytes. Curr Biol. 2015; 25(18): 2411–2416. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBeers DR, Henkel JS, Xiao Q, et al.: Wild-type microglia extend survival in PU.1 knockout mice with familial amyotrophic lateral sclerosis. Proc Natl Acad Sci U S A. 2006; 103(43): 16021–16026. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBerghoff SA, Gerndt N, Winchenbach J, et al.: Dietary cholesterol promotes repair of demyelinated lesions in the adult brain. Nat Commun. 2017; 8: 14241. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBijland S, Thomson G, Euston M, et al.: Dataset 1 in: An in vitro model for studying CNS white matter: functional properties and experimental approaches. F1000Research. 2019. http://www.doi.org/10.5256/f1000research.16802.d233269\n\nBoillée S, Yamanaka K, Lobsiger CS, et al.: Onset and progression in inherited ALS determined by motor neurons and microglia. Science. 2006; 312(5778): 1389–1392. PubMed Abstract | Publisher Full Text\n\nBottenstein JE, Sato GH: Growth of a rat neuroblastoma cell line in serum-free supplemented medium. Proc Natl Acad Sci U S A. 1979; 76(1): 514–517. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBsibsi M, Ravid R, Gveric D, et al.: Broad expression of Toll-like receptors in the human central nervous system. J Neuropathol Exp Neurol. 2002; 61(11): 1013–1021. PubMed Abstract | Publisher Full Text\n\nCantoni C, Bollman B, Licastro D, et al.: TREM2 regulates microglial cell activation in response to demyelination in vivo. Acta Neuropathol. 2015; 129(3): 429–447. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarpenter AE, Jones TR, Lamprecht MR, et al.: CellProfiler: image analysis software for identifying and quantifying cell phenotypes. Genome Biol. 2006; 7(10): R100. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChang A, Nishiyama A, Peterson J, et al.: NG2-positive oligodendrocyte progenitor cells in adult human brain and multiple sclerosis lesions. J Neurosci. 2000; 20(17): 6404–6412. PubMed Abstract | Publisher Full Text\n\nChang A, Tourtellotte WW, Rudick R, et al.: Premyelinating oligodendrocytes in chronic lesions of multiple sclerosis. N Engl J Med. 2002; 346(3): 165–173. PubMed Abstract | Publisher Full Text\n\nChow JC, Young DW, Golenbock DT, et al.: Toll-like receptor-4 mediates lipopolysaccharide-induced signal transduction. J Biol Chem. 1999; 274(16): 10689–10692. PubMed Abstract | Publisher Full Text\n\nCotrina ML, Nedergaard M: Brain connexins in demyelinating diseases: therapeutic potential of glial targets. Brain Res. 2012; 1487: 61–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCrawford AH, Tripathi RB, Foerster S, et al.: Pre-Existing Mature Oligodendrocytes Do Not Contribute to Remyelination following Toxin-Induced Spinal Cord Demyelination. Am J Pathol. 2016; 186(3): 511–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCumberworth SL, Barrie JA, Cunningham ME, et al.: Zika virus tropism and interactions in myelinating neural cell cultures: CNS cells and myelin are preferentially affected. Acta Neuropathol Commun. 2017; 5(1): 50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDavalos D, Grutzendler J, Yang G, et al.: ATP mediates rapid microglial response to local brain injury in vivo. Nat Neurosci. 2005; 8(6): 752–758. PubMed Abstract | Publisher Full Text\n\nDemerens C, Stankoff B, Logak M, et al.: Induction of myelination in the central nervous system by electrical activity. Proc Natl Acad Sci U S A. 1996; 93(18): 9887–9892. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeshmukh VA, Tardif V, Lyssiotis CA, et al.: A regenerative approach to the treatment of multiple sclerosis. Nature. 2013; 502(7471): 327–332. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDuncan ID, Brower A, Kondo Y, et al.: Extensive remyelination of the CNS leads to functional recovery. Proc Natl Acad Sci U S A. 2009; 106(16): 6832–6836. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDutta DJ, Woo DH, Lee PR, et al.: Regulation of myelin structure and conduction velocity by perinodal astrocytes. Proc Natl Acad Sci U S A. 2018; 115(46): 11832–11837. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdgar JM, McCulloch MC, Montague P, et al.: Demyelination and axonal preservation in a transgenic mouse model of Pelizaeus-Merzbacher disease. EMBO Mol Med. 2010; 2(2): 42–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdgar JM, McCulloch MC, Thomson CE, et al.: Distribution of mitochondria along small-diameter myelinated central nervous system axons. J Neurosci Res. 2008; 86(10): 2250–2257. PubMed Abstract | Publisher Full Text\n\nEdgar JM, McLaughlin M, Werner HB, et al.: Early ultrastructural defects of axons and axon-glia junctions in mice lacking expression of Cnp1. Glia. 2009; 57(16): 1815–1824. PubMed Abstract | Publisher Full Text\n\nEdgar JM, McLaughlin M, Yool D, et al.: Oligodendroglial modulation of fast axonal transport in a mouse model of hereditary spastic paraplegia. J Cell Biol. 2004; 166(1): 121–131. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFancy SP, Baranzini SE, Zhao C, et al.: Dysregulation of the Wnt pathway inhibits timely myelination and remyelination in the mammalian CNS. Genes Dev. 2009; 23(13): 1571–1585. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFeng GP, Mellor RH, Bernstein M, et al.: Imaging neuronal subsets in transgenic mice expressing multiple spectral variants of GFP. Neuron. 2000; 28(1): 41–51. PubMed Abstract | Publisher Full Text\n\nForrester JV, McMenamin PG, Dando SJ: CNS infection and immune privilege. Nat Rev Neurosci. 2018; 19(11): 655–671. PubMed Abstract | Publisher Full Text\n\nFowler JH, McQueen J, Holland PR, et al.: Dimethyl fumarate improves white matter function following severe hypoperfusion: Involvement of microglia/macrophages and inflammatory mediators. J Cereb Blood Flow Metab. 2018; 38(8): 1354–1370. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFranklin RJ, Ffrench-Constant C: Remyelination in the CNS: from biology to therapy. Nat Rev Neurosci. 2008; 9(11): 839–855. PubMed Abstract | Publisher Full Text\n\nFranklin RJ, ffrench-Constant C, Edgar JM, et al.: Neuroprotection and repair in multiple sclerosis. Nat Rev Neurol. 2012; 8(11): 624–634. PubMed Abstract | Publisher Full Text\n\nFroger N, Orellana JA, Calvo CF, et al.: Inhibition of cytokine-induced connexin43 hemichannel activity in astrocytes is neuroprotective. Mol Cell Neurosci. 2010; 45(1): 37–46. PubMed Abstract | Publisher Full Text\n\nFünfschilling U, Supplie LM, Mahad D, et al.: Glycolytic oligodendrocytes maintain myelin and long-term axonal integrity. Nature. 2012; 485(7399): 517–521. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGeorge R, Griffin JW: Delayed macrophage responses and myelin clearance during Wallerian degeneration in the central nervous system: the dorsal radiculotomy model. Exp Neurol. 1994; 129(2): 225–236. PubMed Abstract | Publisher Full Text\n\nGibson EM, Purger D, Mount CW, et al.: Neuronal activity promotes oligodendrogenesis and adaptive myelination in the mammalian brain. Science. 2014; 344(6183): 1252304. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGlass WG, Lim JK, Cholera R, et al.: Chemokine receptor CCR5 promotes leukocyte trafficking to the brain and survival in West Nile virus infection. J Exp Med. 2005; 202(8): 1087–1098. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoebbels S, Wieser GL, Pieper A, et al.: A neuronal PI(3,4,5)P3-dependent program of oligodendrocyte precursor recruitment and myelination. Nat Neurosci. 2017; 20(1): 10–15. PubMed Abstract | Publisher Full Text\n\nGriffiths IR, Klugmann M, Anderson TJ, et al.: Axonal swellings and degeneration in mice lacking the major proteolipid of myelin. Science. 1998; 280(5369): 1610–1613. PubMed Abstract | Publisher Full Text\n\nGrillner S, McClellan A, Sigvardt K, et al.: Activation of NMDA-receptors elicits \"fictive locomotion\" in lamprey spinal cord in vitro. Acta Physiol Scand. 1981; 113(4): 549–551. PubMed Abstract | Publisher Full Text\n\nHall ED, Oostveen JA, Gurney ME: Relationship of microglial and astrocytic activation to disease onset and progression in a transgenic model of familial ALS. Glia. 1998; 23(3): 249–256. PubMed Abstract | Publisher Full Text\n\nHanisch UK, Kettenmann H: Microglia: active sensor and versatile effector cells in the normal and pathologic brain. Nat Neurosci. 2007; 10(11): 1387–1394. PubMed Abstract | Publisher Full Text\n\nHines DJ, Choi HB, Hines RM, et al.: Prevention of LPS-induced microglia activation, cytokine production and sickness behavior with TLR4 receptor interfering peptides. PLoS One. 2013; 8(3): e60388. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHines JH, Ravanelli AM, Schwindt R, et al.: Neuronal activity biases axon selection for myelination in vivo. Nat Neurosci. 2015; 18(5): 683–689. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHonda S, Sasaki Y, Ohsawa K, et al.: Extracellular ATP or ADP induce chemotaxis of cultured microglia through Gi/o-coupled P2Y receptors. J Neurosci. 2001; 21(6): 1975–1982. PubMed Abstract | Publisher Full Text\n\nHuang JK, Jarjour AA, Nait Oumesmar B, et al.: Retinoid X receptor gamma signaling accelerates CNS remyelination. Nat Neurosci. 2011; 14(1): 45–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHughes CE, Nibbs RJB: A guide to chemokines and their receptors. FEBS J. 2018; 285(16): 2944–2971. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIoannidou K, Anderson KI, Strachan D, et al.: Time-lapse imaging of the dynamics of CNS glial-axonal interactions in vitro and ex vivo. PLoS One. 2012; 7(1): e30775. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIp CW, Kroner A, Bendszus M, et al.: Immune cells contribute to myelin degeneration and axonopathic changes in mice overexpressing proteolipid protein in oligodendrocytes. J Neurosci. 2006; 26(31): 8206–8216. PubMed Abstract | Publisher Full Text\n\nIp CW, Kroner A, Crocker PR, et al.: Sialoadhesin deficiency ameliorates myelin degeneration and axonopathic changes in the CNS of PLP overexpressing mice. Neurobiol Dis. 2007; 25(1): 105–111. PubMed Abstract | Publisher Full Text\n\nIrvine KA, Blakemore WF: Remyelination protects axons from demyelination-associated axon degeneration. Brain. 2008; 131(Pt 6): 1464–1477. PubMed Abstract | Publisher Full Text\n\nIshibashi T, Dakin KA, Stevens B, et al.: Astrocytes promote myelination in response to electrical impulses. Neuron. 2006; 49(6): 823–832. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKassmann CM, Lappe-Siefke C, Baes M, et al.: Axonal loss and neuroinflammation caused by peroxisome-deficient oligodendrocytes. Nat Genet. 2007; 39(8): 969–976. PubMed Abstract | Publisher Full Text\n\nKeirstead HS, Blakemore WF: Identification of post-mitotic oligodendrocytes incapable of remyelination within the demyelinated adult spinal cord. J Neuropath Exp Neurol. 1997; 56(11): 1191–1201. PubMed Abstract | Publisher Full Text\n\nKigerl KA, de Rivero Vaccari JP, Dietrich WD, et al.: Pattern recognition receptors and central nervous system repair. Exp Neurol. 2014; 258: 5–16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKirkpatrick LL, Witt AS, Payne HR, et al.: Changes in microtubule stability and density in myelin-deficient shiverer mouse CNS axons. J Neurosci. 2001; 21(7): 2288–2297. PubMed Abstract | Publisher Full Text\n\nKline D: Quantitative microinjection of mouse oocytes and eggs. Microinjections: Methods and application. (Carroll DJ, ed), Humana Press. 2009; 518: 135–156. PubMed Abstract | Publisher Full Text\n\nKotter MR, Zhao C, van Rooijen N, et al.: Macrophage-depletion induced impairment of experimental CNS remyelination is associated with a reduced oligodendrocyte progenitor cell response and altered growth factor expression. Neurobiol Dis. 2005; 18(1): 166–175. PubMed Abstract | Publisher Full Text\n\nLampron A, Larochelle A, Laflamme N, et al.: Inefficient clearance of myelin debris by microglia impairs remyelinating processes. J Exp Med. 2015; 212(4): 481–495. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLappe-Siefke C, Goebbels S, Gravel M, et al.: Disruption of Cnp1 uncouples oligodendroglial functions in axonal support and myelination. Nat Genet. 2003; 33(3): 366–374. PubMed Abstract\n\nLee PR, Cohen JE, Iacobas DA, et al.: Gene networks activated by specific patterns of action potentials in dorsal root ganglia neurons. Sci Rep. 2017; 7: 43765. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee S, Leach MK, Redmond SA, et al.: A culture system to study oligodendrocyte myelination processes using engineered nanofibers. Nat Methods. 2012a; 9(9): 917–922. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLee Y, Morrison BM, Li Y, et al.: Oligodendroglia metabolically support axons and contribute to neurodegeneration. Nature. 2012b; 487(7408): 443–448. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLennartz M, Drake J: Molecular mechanisms of macrophage Toll-like receptor–Fc receptor synergy [version 1; referees: 2 approved]. F1000Res. 2018; 7(F1000 Faculty Rev): 21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLindner M, Thümmler K, Arthur A, et al.: Fibroblast growth factor signalling in multiple sclerosis: inhibition of myelination and induction of pro-inflammatory environment by FGF9. Brain. 2015; 138(Pt 7): 1875–1893. PubMed Abstract | Publisher Full Text\n\nLiu MT, Chen BP, Oertel P, et al.: The T cell chemoattractant IFN-inducible protein 10 is essential in host defense against viral-induced neurologic disease. J Immunol. 2000; 165(5): 2327–2330. PubMed Abstract | Publisher Full Text\n\nLobo-Silva D, Carriche GM, Castro AG, et al.: Balancing the immune response in the brain: IL-10 and its regulation. J Neuroinflammation. 2016; 13(1): 297. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLopez de Armentia M, Jancic D, Olivares R, et al.: cAMP response element-binding protein-mediated gene expression increases the intrinsic excitability of CA1 pyramidal neurons. J Neurosci. 2007; 27(50): 13909–13918. PubMed Abstract | Publisher Full Text\n\nLundgaard I, Luzhynskaya A, Stockley JH, et al.: Neuregulin and BDNF induce a switch to NMDA receptor-dependent myelination by oligodendrocytes. PLoS Biol. 2013; 11(12): e1001743. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMadhavan M, Nevin ZS, Shick HE, et al.: Induction of myelinating oligodendrocytes in human cortical spheroids. Nat Methods. 2018; 15(9): 700–706. PubMed Abstract | Publisher Full Text\n\nMayoral SR, Etxeberria A, Shen YA, et al.: Initiation of CNS Myelination in the Optic Nerve Is Dependent on Axon Caliber. Cell Rep. 2018; 25(3): 544–550.e3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMei F, Fancy SPJ, Shen YA, et al.: Micropillar arrays as a high-throughput screening platform for therapeutics in multiple sclerosis. Nat Med. 2014; 20(8): 954–960. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMei F, Lehmann-Horn K, Shen YA, et al.: Accelerated remyelination during inflammatory demyelination prevents axonal loss and improves functional recovery. eLife. 2016; 5: pii: e18246. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMensch S, Baraban M, Almeida R, et al.: Synaptic vesicle release regulates myelin sheath number of individual oligodendrocytes in vivo. Nat Neurosci. 2015; 18(5): 628–630. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMeyer N, Richter N, Fan Z, et al.: Oligodendrocytes in the Mouse Corpus Callosum Maintain Axonal Function by Delivery of Glucose. Cell Rep. 2018; 22(9): 2383–2394. PubMed Abstract | Publisher Full Text\n\nMir F, Le Breton GC: A novel nuclear signaling pathway for thromboxane A2 receptors in oligodendrocytes: evidence for signaling compartmentalization during differentiation. Mol Cell Biol. 2008; 28(20): 6329–6341. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMiron VE, Boyd A, Zhao JW, et al.: M2 microglia and macrophages drive oligodendrocyte differentiation during CNS remyelination. Nat Neurosci. 2013; 16(9): 1211–1218. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMitew S, Gobius I, Fenlon LR, et al.: Pharmacogenetic stimulation of neuronal activity increases myelination in an axon-specific manner. Nat Commun. 2018; 9(1): 306. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMücklisch S, Mücklisch K (nee Thümmler): muecs/cp v1.0. 2019. http://www.doi.org/10.5281/zenodo.2533339\n\nNagai M, Re DB, Nagata T, et al.: Astrocytes expressing ALS-linked mutated SOD1 release factors selectively toxic to motor neurons. Nat Neurosci. 2007; 10(5): 615–622. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNajm FJ, Madhavan M, Zaremba A, et al.: Drug-based modulation of endogenous stem cells promotes functional remyelination in vivo. Nature. 2015; 522(7555): 216–20. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNapoli I, Neumann H: Microglial clearance function in health and disease. Neuroscience. 2009; 158(3): 1030–1038. PubMed Abstract | Publisher Full Text\n\nNave KA, Trapp BD: Axon-glial signaling and the glial support of axon function. Annu Rev Neurosci. 2008; 31: 535–561. PubMed Abstract | Publisher Full Text\n\nNave KA, Griffiths IR: Models of Pelizaeus-Merzbacher disease. In: Myelin biology and disorders (Lazzarini RA, Griffin JW, Lassmann H, Nave K-A, Miller RH, Trapp BD, eds), Amsterdam: Elsevier. Amsterdam: Elsevier, 2004; 2: 1125–1142. Publisher Full Text\n\nNeumann H, Kotter MR, Franklin RJ: Debris clearance by microglia: an essential link between degeneration and regeneration. Brain. 2009; 132(Pt 2): 288–295. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNimmerjahn A, Kirchhoff F, Helmchen F: Resting microglial cells are highly dynamic surveillants of brain parenchyma in vivo. Science. 2005; 308(5726): 1314–1318. PubMed Abstract | Publisher Full Text\n\nNorton WT, Poduslo SE: Myelination in rat brain: method of myelin isolation. J Neurochem. 1973; 21(4): 749–757. PubMed Abstract | Publisher Full Text\n\nPan B, Fromholt SE, Hess EJ, et al.: Myelin-associated glycoprotein and complementary axonal ligands, gangliosides, mediate axon stability in the CNS and PNS: neuropathology and behavioral deficits in single- and double-null mice. Exp Neurol. 2005; 195(1): 208–217. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatani R, Balaratnam M, Vora A, et al.: Remyelination can be extensive in multiple sclerosis despite a long disease course. Neuropathol Appl Neurobiol. 2007; 33(3): 277–287. PubMed Abstract | Publisher Full Text\n\nPerry VH, Hume DA, Gordon S: Immunohistochemical localization of macrophages and microglia in the adult and developing mouse brain. Neuroscience. 1985; 15(2): 313–326. PubMed Abstract | Publisher Full Text\n\nPoliani PL, Wang Y, Fontana E, et al.: TREM2 sustains microglial expansion during aging and response to demyelination. J Clin Invest. 2015; 125(5): 2161–2170. PubMed Abstract | Publisher Full Text | Free Full Text\n\nReadhead C, Schneider A, Griffiths IR, et al.: Premature arrest of myelin formation in transgenic mice with increased proteolipid protein gene dosage. Neuron. 1994; 12(3): 583–595. PubMed Abstract | Publisher Full Text\n\nSaijo K, Glass CK: Microglial cell origin and phenotypes in health and disease. Nat Rev Immunol. 2011; 11(11): 775–787. PubMed Abstract | Publisher Full Text\n\nSánchez-Gómez MV, Serrano MP, Alberdi E, et al.: Isolation, Expansion, and Maturation of Oligodendrocyte Lineage Cells Obtained from Rat Neonatal Brain and Optic Nerve. Methods Mol Biol. 2018; 1791: 95–113. PubMed Abstract | Publisher Full Text\n\nSchirmer L, Möbius W, Zhao C, et al.: Oligodendrocyte-encoded Kir4.1 function is required for axonal integrity. eLife. 2018; 7: pii: e36428. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScolding N, Franklin R, Stevens S, et al.: Oligodendrocyte progenitors are present in the normal adult human CNS and in the lesions of multiple sclerosis. Brain. 1998; 121(Pt 12): 2221–2228. PubMed Abstract\n\nShen S, Sandoval J, Swiss VA, et al.: Age-dependent epigenetic control of differentiation inhibitors is critical for remyelination efficiency. Nat Neurosci. 2008; 11(9): 1024–1034. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSierra A, Abiega O, Shahraz A, et al.: Janus-faced microglia: beneficial and detrimental consequences of microglial phagocytosis. Front Cell Neurosci. 2013; 7: 6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSong L, Pei L, Yao S, et al.: NLRP3 Inflammasome in Neurological Diseases, from Functions to Therapies. Front Cell Neurosci. 2017; 11: 63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSong WM, Colonna M: The identity and function of microglia in neurodegeneration. Nat Immunol. 2018a; 19(10): 1048–1058. PubMed Abstract | Publisher Full Text\n\nSong WM, Colonna M: The Microglial Response to Neurodegenerative Disease. Adv Immunol. 2018b; 139: 1–50. PubMed Abstract | Publisher Full Text\n\nSorensen EW, Lian J, Ozga AJ, et al.: CXCL10 stabilizes T cell-brain endothelial cell adhesion leading to the induction of cerebral malaria. JCI Insight. 2018; 3(8): pii: 98911. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTambuyzer BR, Ponsaerts P, Nouwen EJ: Microglia: gatekeepers of central nervous system immunology. J Leukoc Biol. 2009; 85(3): 352–370. PubMed Abstract | Publisher Full Text\n\nThion MS, Ginhoux F, Garel S: Microglia and early brain development: An intimate journey. Science. 2018; 362(6411): 185–189. PubMed Abstract | Publisher Full Text\n\nThomson CE, Griffiths IR, McCulloch MC, et al.: In vitro studies of axonally-regulated Schwann cell genes during Wallerian degeneration. J Neurocytol. 1993; 22(8): 590–602. PubMed Abstract | Publisher Full Text\n\nThomson CE, Hunter AM, Griffiths IR, et al.: Murine spinal cord explants: a model for evaluating axonal growth and myelination in vitro. J Neurosci Res. 2006; 84(8): 1703–1715. PubMed Abstract | Publisher Full Text\n\nThomson CE, McCulloch M, Sorenson A, et al.: Myelinated, synapsing cultures of murine spinal cord--validation as an in vitro model of the central nervous system. Eur J Neurosci. 2008; 28(8): 1518–1535. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTrapp BD, Bernier L, Andrews SB, et al.: Cellular and subcellular distribution of 2',3'-cyclic nucleotide 3'-phosphodiesterase and its mRNA in the rat central nervous system. J Neurochem. 1988; 51(3): 859–868. PubMed Abstract | Publisher Full Text\n\nTrapp BD, Nave KA: Multiple sclerosis: an immune or neurodegenerative disorder? Annu Rev Neurosci. 2008; 31: 247–269. PubMed Abstract | Publisher Full Text\n\nvan Zundert B, Brown RH Jr: Silencing strategies for therapy of SOD1-mediated ALS. Neurosci Lett. 2017; 636: 32–39. PubMed Abstract | Publisher Full Text\n\nWake H, Lee PR, Fields RD: Control of local protein synthesis and initial events in myelination by action potentials. Science. 2011; 333(6049): 1647–1651. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWlodarczyk A, Holtman IR, Krueger M, et al.: A novel microglial subset plays a key role in myelinogenesis in developing brain. EMBO J. 2017; 36(22): 3292–3308. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorld Health Organization: Atlas multiple sclerosis resources in the world 2008. Geneva: WHO Press; 2008. Reference Source\n\nYamanaka K, Chun SJ, Boillee S, et al.: Astrocytes as determinants of disease progression in inherited amyotrophic lateral sclerosis. Nat Neurosci. 2008; 11(3): 251–253. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZalc B, Fields RD: Do Action Potentials Regulate Myelination? Neuroscientist. 2000; 6(1): 5–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZawadzka M, Rivers LE, Fancy SP, et al.: CNS-resident glial progenitor/stem cells produce Schwann cells as well as oligodendrocytes during repair of CNS demyelination. Cell Stem Cell. 2010; 6(6): 578–590. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang Y, Chen K, Sloan SA, et al.: An RNA-sequencing transcriptome and splicing database of glia, neurons, and vascular cells of the cerebral cortex. J Neurosci. 2014; 34(36): 11929–11947. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang H, Jarjour AA, Boyd A, et al.: Central nervous system remyelination in culture--a tool for multiple sclerosis research. Exp Neurol. 2011; 230(1): 138–148. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZheng M, Tao W, Zou Y, et al.: Nanotechnology-Based Strategies for siRNA Brain Delivery for Disease Therapy. Trends Biotechnol. 2018; 36(5): 562–575. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "43736",
"date": "14 Feb 2019",
"name": "Tracey A. Newman",
"expertise": [
"Reviewer Expertise Referee suggested by the NC3Rs for their scientific expertise and experience in assessing 3Rs impact"
],
"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 disease burden associated with nervous system white matter (myelin) changes is significant. Newer imaging modalities/image analyses are enabling the detection of white matter changes in the clinic. The upshot of this is that there is a need for robust systems for the study of the mechanisms underlying demyelination/dysmyelination and systems in which to study remyelination. A significant amount of the research to address these issues relies on the use of in-vivo rodent experiments. Many of which fall into the moderate category of in-vivo procedure at a minimum. This work describes a relatively straight-forward model system which could replace, or be used to inform, some of the in-vivo work.\n\nThe rationale is clear and the methods are thoroughly described. The description and interpretation of the data is robust, and the suggested utility of the method is proportionate. The manuscript would be improved further by considering the following:\n\nA short section describing similar/or equivalent in-vitro systems ideally describing the advantage of the system described. Clarification regarding the value of having a system that is not part of a whole system (in the research highlights this is presented as a strength, but why is not made explicit). The cultures have been exploited for a number of features, and this is supported by the data that is presented – however some analysis of the cellular organisation (2D/3D) of the cultures and the variability in this would be very useful as other researchers consider whether the system is appropriate for the experimental questions that want to address using the culture. It would also enable the myelination to be seen in a different orientation, and support the data described. One approach to this would be to section the cultures, possibly using an en-face approach as the cultures are grown on coverslips. This approach has been used in other organoid/mini-brain culture analyses. A key issue with the development of new methods is the dissemination to other groups, the authors report that the model could be readily adopted in other labs equipped for cell culture. Has this happened? Irrespective of whether the model has been used to test a hypothesis elsewhere, it would be useful to know that a second group working from the protocols has reported similar findings, this will increase confidence in the model. Re the methods, describe briefly the sterilisation step for the live imaging dishes. Minor points a careful final proof-read is needed e.g. flurobite p5 needs to be corrected.\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\nAre the 3Rs implications of the work described accurately? Yes\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": [
{
"c_id": "4433",
"date": "19 Feb 2019",
"name": "Julia Edgar",
"role": "Author Response",
"response": "We are very grateful to the reviewer for taking the time to carefully read and consider the manuscript and will respond more fully to her comments shortly."
}
]
},
{
"id": "43885",
"date": "19 Feb 2019",
"name": "Jacqueline Trotter",
"expertise": [
"Reviewer Expertise Myelination",
"primary cell culture",
"siRNA"
],
"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 expands in substantial detail on a previously published method of myelinating cultures generated from murine spinal cord. The authors clearly demonstrate that establishing the cultures on MEA chips facilitates the monitoring of electrophysiological properties of the system. They show clearly the manipulation of gene expression in the culture using siRNA and also the ability to study developmental and pathological processes, such as innate immune responses, using imaging. They also demonstrate the use of multi-well plates facilitating drug screening.\n\nThe article thus describes an excellent straightforward method to complement, and in some instances replace, experiments in live animals. The method will be very useful for screening reagents with the potential to influence myelination or remyelination. It therefore provides a timely and important contribution to the experimental toolbox of myelin researchers.\nAre a suitable application and appropriate end-users identified? Yes\nIf applicable, is the statistical analysis and its interpretation appropriate? Yes\nAre the 3Rs implications of the work described accurately? Yes\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": "4432",
"date": "19 Feb 2019",
"name": "Julia Edgar",
"role": "Author Response",
"response": "We are very grateful to reviewer for taking the time to carefully read and consider our manuscript, which became rather long in the end."
}
]
}
] | 1
|
https://f1000research.com/articles/8-117
|
https://f1000research.com/articles/7-1725/v1
|
31 Oct 18
|
{
"type": "Systematic Review",
"title": "Oral hormone pregnancy tests and the risks of congenital malformations: a systematic review and meta-analysis",
"authors": [
"Carl J. Heneghan",
"Jeffrey K. Aronson",
"Elizabeth Spencer",
"Bennett Holman",
"Kamal R. Mahtani",
"Rafael Perera",
"Igho Onakpoya",
"Jeffrey K. Aronson",
"Elizabeth Spencer",
"Bennett Holman",
"Kamal R. Mahtani",
"Rafael Perera",
"Igho Onakpoya"
],
"abstract": "Background: Oral hormone pregnancy tests (HPTs), such as Primodos, containing ethinylestradiol and high doses of norethisterone, were given to over a million women from 1958 to 1978, when Primodos was withdrawn from the market because of concerns about possible teratogenicity. We aimed to study the association between maternal exposure to oral HPTs and congenital malformations. Methods: We have performed a systematic review and meta-analysis of case-control and cohort studies that included data from pregnant women and were exposed to oral HPTs within the estimated first three months of pregnancy, if compared with a relevant control group. We used random-effects meta-analysis and assessed the quality of each study using the Newcastle–Ottawa Scale for non-randomized studies. Results: We found 16 case control studies and 10 prospective cohort studies, together including 71 330 women, of whom 4209 were exposed to HPTs. Exposure to oral HPTs was associated with a 40% increased risk of all congenital malformations: pooled odds ratio (OR) = 1.40 (95% CI 1.18 to 1.66; P<0.0001; I2 = 0%). Exposure to HPTs was associated with an increased risk of congenital heart malformations: pooled OR = 1.89 (95% CI 1.32 to 2.72; P = 0.0006; I2=0%); nervous system malformations OR = 2.98 (95% CI 1.32 to 6.76; P = 0.0109 I2 = 78%); gastrointestinal malformations, OR = 4.50 (95% CI 0.63 to 32.20; P = 0.13; I2 = 54%); musculoskeletal malformations, OR = 2.24 (95% CI 1.23 to 4.08; P= 0.009; I2 = 0%); the VACTERL syndrome (Vertebral defects, Anal atresia, Cardiovascular anomalies, Tracheoesophageal fistula, Esophageal atresia, Renal anomalies, and Limb defects), OR = 7.47 (95% CI 2.92 to 19.07; P < 0.0001; I2 = 0%). Conclusions: This systematic review and meta-analysis shows that use of oral HPTs in pregnancy is associated with increased risks of congenital malformations.",
"keywords": [
"pregnancy",
"congenital malformations",
"hormones"
],
"content": "Introduction\n\nOral hormone pregnancy tests (HPTs), such as Primodos (known as Duogynon in Germany), were used from 1958 to 1978, before urine pregnancy tests were available1. Oral HPTs contained ethinylestradiol and large doses of norethisterone (synthetic forms of estrogen and progesterone respectively), the latter in much larger amounts than those included in current combined oral contraceptives (see Table 1). The test principle was that menstruation would be induced in those who were not pregnant.\n\n*Unbranded\n\nIn the UK more than a million women took HPTs2. However, evidence that they should not be used in pregnant women because of a risk of fetal malformations3 led the then Committee on Safety of Medicines in 1975 to conclude that a warning should be added to the Data Sheets, stating that HPTs should not be taken during pregnancy. (Supplementary File 1) In 1978, the manufacturers of Primodos, Schering AG (taken over by Bayer AG in 2008), voluntarily stopped marketing the product.\n\nSince Primodos was withdrawn, the discovery of previously confidential documents has led to renewed concerns about its potential to cause harm. In 2014, therefore, the Medicines and Healthcare products Regulatory Agency (MHRA) initiated a review, which was published in 2017 and reported that the evidence was insufficient, mixed, and too heterogeneous to support an association between oral HPTs and congenital malformations3.\n\nTo date, there has been no systematic review and meta-analysis of oral HPTs, using all the available data, to assess the likelihood of an association. We have therefore performed a systematic review to obtain all relevant data on hormone pregnancy tests and congenital malformations, used meta-analytical tools to obtain summary estimates of the likelihood of an association, and assessed the potential biases in these estimates.\n\n\nMethods\n\nFull details of our search strategy are provided in Supplementary File 2. We searched Medline, Embase, and Web of Science (which yielded German papers and conference abstracts) and searched for regulatory documents online, including the UK Government’s “Report of the Commission on Human Medicines’ Expert Working Group on Hormone Pregnancy Tests”, which includes the original Landesarchiv Berlin Files1, and reference lists of retrieved studies from the start of the databases in 1946 to 20 February 2018.\n\nWe used the following search terms without date limits or language restrictions: (Primodos OR Duogynon OR \"hormone pregnancy test\" OR \"sex hormones\" OR \"hormone administration\" OR “norethisterone” OR “ethinylestradiol”) AND pregnancy AND (congenital OR malformations OR anomalies). Several comparable high-dose HPTs were available at the same time as Primodos; we performed additional searches for evidence relating to these (See Supplementary File 3 for List of HPTs included in evidence search).\n\nWe included observational studies of women who were or became pregnant during the study and were exposed to oral HPTs within the estimated first three months of pregnancy and compared them with a relevant control group. When a study was described in more than one publication, we chose the publication that contained the most comprehensive data as the primary publication. We excluded studies where the intervention was oral hormones taken for other reasons (e.g., oral contraception) and it was not possible to extract data on hormone pregnancy tests. We did not restrict the language of publication. We checked additional relevant data and extracted them from the secondary publications when necessary.\n\nTwo reviewers (CH and ES) applied inclusion and quality assessment criteria, compared results, and resolved discrepancies through discussion with the other authors. We used a review template to extract data on study type, numbers of pregnancies exposed and not exposed to oral HPTs, and types and numbers of outcomes. Where available, we extracted data about the women studied, including ascertainment of cases, age, parity, setting, exposure to other medications, and confounding variables. In case-control studies, if data were reported on more than one control group, we extracted data where possible for non-disease/non-abnormality controls, and combined control groups if necessary.\n\nThe primary outcome of interest was all major congenital malformations. We also categorized outcomes for the congenital anomaly in the offspring at any time into congenital cardiac, gastrointestinal, musculoskeletal, nervous system, and urogenital defects, and the VACTERL syndrome (Vertebral defects, Anal atresia, Cardiovascular anomalies, Tracheoesophageal fistula, Esophageal atresia, Renal anomalies, and Limb defects).\n\nWe assessed quality using the Newcastle–Ottawa Scale (NOS) for non-randomized studies included in systematic reviews6. The scale assesses the selection of study groups (cases and controls), comparability of study groups, including cases and controls, and ascertainment of the outcome/exposure. Each positive criterion scores 1 point, except comparability, which scores up to 2 points. The maximum NOS score is 9, and we interpreted a score of 1 to 3 points as indicating a high risk of bias7. To determine whether the study had controlled for the most important factors, we selected the items reported in the original paper and resolved disagreements through consensus, using a third author (IO). We examined whether there was a linear relation between methodological quality and study results, by plotting the odds ratios against the NOS scores, using Excel, and assessed the correlations of NOS scores with several confounding variables we collected8.\n\nWe calculated study-specific odds ratios for outcomes and associated confidence intervals. We meta-analysed the data using a random-effects model. We assessed heterogeneity across studies using the I2 statistic and publication bias using funnel plots9. We performed a sensitivity analysis by removing single studies to judge the stability of the effect and to explore the effect on heterogeneity10, and we described any sources of variation. We also judged robustness by removing studies of low quality from the analysis. To examine whether the observed heterogeneity could be explained by differences in the NOS score, we also performed meta-regression using the NOS score as the covariate against the log OR as weights for traditional meta-regression using Stata version 14.\n\nWe planned subgroup analyses for the timing of administration of HPTs in relation to pregnancy and organogenesis and study design (case-control versus cohort) using Cochran’s Q test. We used RevMan v.5.3 for all analyses, except for meta-regression, for which we used Stata version 14. RevMan and Stata estimate the effects of trials with zero events in one arm by adding a correction factor of 0.5 to each arm (trials with zero events in both arms are omitted). We performed a sensitivity analysis by removing studies with zero events from the analyses.\n\nWe followed the reporting guidelines of the Meta-Analysis of Observational Studies in Epidemiology (MOOSE). A completed checklist is available as Supplementary File 411\n\nMembers of the Association for Children Damaged by HPTs were involved in the original discussions of this review and provided input to the outcome choices, the search, the location of study articles, and translations. We plan to present the study findings to relevant patient groups and make available lay interpretations.\n\n\nResults\n\nWe retrieved 409 items for screening. After title and abstract screening and removal of duplicates (n = 18), we excluded 354 records as not being relevant to the aim of the review. We assessed the full texts of 37 articles and identified 24 articles for inclusion. Figure 1 shows the PRISMA flow diagram for the inclusion of studies.\n\nThe 24 included articles reported on 26 studies (16 case-control studies and ten prospective cohort studies); one article [Nora 78] included two case-control studies and one prospective study. We found no randomized controlled trials. Of these articles, two were unpublished reports (see Supplementary File 5 for full references). The studies included 71 330 women. The case-control studies included 28 761 mothers, 594 of whom were exposed to HPTs; the cohort studies included 42 569 mothers and 3615 exposures to HPTs. The studies were published between 1972 and 2014, and all were performed either in Europe or the USA. They mostly recruited women and their infants at maternity centres or hospital paediatrics wards.\n\nThe choices of controls in the case-control studies varied; they included, at one extreme, healthy infants born on a date close to the case infants and, at the other extreme, infants with malformations other than those under investigation. Among the prospective cohort studies, the populations tended to be women recruited at antenatal clinics or birth centres (See Table 2. Characteristics of included studies).\n\nOf the 26 included studies, three were assigned a NOS score of 3 or below and were therefore judged as being at high risk of bias. One was a case-control study (Laurence 1971, a published abstract as a letter) and two were cohort studies (Fleming 1978 and Haller 1974, both unpublished). The NOS scores ranged from 2 to 9 (median 5). Twelve of the 26 included studies scored 7 to 9 and were judged to be at low risk of bias (see Table 3 of NOS scores in the data files). Item 5 of the NOS score addresses comparability of cases and controls based on design or analysis. Of the 16 case control studies, 12 controlled for the most important factor (item 5a) and nine controlled for important additional factors (item 5b). Of the ten cohort studies, six controlled for the most important factor (item 5a) and four controlled for important additional factors (item 5b). The mean Newcastle–Ottawa scale score was 6.1, indicating an overall moderate risk of bias. Table 2 also shows that seven studies did not report the confounding variables collected (Laurence 1971; Levy 1973; Tummler 2014; Fleming 1978; Haller 1974; Moire 1978; Rousel 1968). NOS scores correlated with the increasing number of confounding variables collected (r = 0.83). Supplementary File 6 shows the funnel plots for all congenital malformations and congenital heart disease; because of inadequate numbers of included studies, we did not use more advanced statistical methods to assess publication bias.\n\nNine studies, including 61 642 mothers of infants and 3274 exposed to HPTs, examined the association in pregnancy with all congenital malformations. Two were case-control studies (Greenberg 1977; Sainz 1987) and seven were cohort studies (Fleming 1987; Goujard 1979; Haller 1974; Kullander 1976; Michaelis 1983; Rumeau-Rouquette 1978; Torfs 1981) (Figure 2). Exposure to oral HPTs was associated with a 37% increased risk of all congenital malformations: pooled odds ratio (OR) = 1.40 (95% CI 1.18 to 1.66; P < 0.0001; I2 = 0%). For the two case-control studies only, pooled OR = 1.70 (95% CI 1.01 to 2.86; P = 0.04; I2 = 63%) and for the seven cohort studies, pooled OR = 1.28 (95% CI 1.05 to 1.56; P = 0.02; I2 = 0%). The test for subgroup differences was not significant (P = 0.32). In a post-hoc sensitivity analysis, removing the studies that collected no confounding variables (Haller 74 and Fleming 78, both of low quality) did not affect the significance of the result (OR 1.44; 95% CI 1.18 to 1.75; P = 0.0004, I2 = 11%). The meta-regression showed no association between total NOS score and increased risk (P = 0.51).\n\nSeven studies, including 19 267 mothers of infants and 218 exposed to oral HPTs, analysed congenital heart malformations. Five were case-control studies (Ferencz 1980; Janerich 1977; Levy 1973; Nora 1978-2/3) and two were cohort studies (Hadjigeorgiou 1982; Torfs 1981) (Figure 3). The pooled relative OR = 1.89 (95% CI 1.32 to 2.72; P = 0.0006; I2 = 0%).\n\nIn a post-hoc sensitivity analysis, removing one study that collected no confounding variables (Levy 73, a low-quality study) did not affect the significance of the result (OR = 1.88; 95% CI 1.25 to 2.85; P = 0.003, I2 = 12%) For the five case-control studies only, the pooled OR = 1.87 (95% CI 1.23 to 2.85; P = 0.004; I2 = 9%); for the two cohort studies the pooled OR = 1.95 (95% CI 0.44 to 8.69; P = 0.38; I2 = 32%). The meta-regression was not significant (P = 0.94).\n\nFor the association between exposure to oral HPTs and nervous system malformations in the offspring, five studies provided data: three case-control studies (Gal 1972; Laurence 1971; Sainz 1987) and two cohort studies (Roussel 1968; Torfs 1981), including 12 486 mothers of infants and 127 exposed (Figure 4). The pooled OR = 2.98 (95% CI 1.32 to 6.76; P = 0.009; I2 = 78%). In a post-hoc sensitivity analysis, removing the two studies that collected no confounding variables (Laurence 71; Roussel 68) did not affect the significance of the result and removed the heterogeneity (OR 6.04; 95% CI 3.33 to 10.78; P < 0.00001, I2 = 0%).\n\nGastrointestinal malformations and exposure to oral HPTs were reported in three studies: a case-control study (Lammer 1986) and two cohort studies (Meire 1978 and Torfs 1981), providing data on 2722 mothers of infants, including 79 exposed to HPTs (Figure 5). The pooled OR = 4.50 (95% CI 0.63 to 32.20; P = 0.13; I2 = 54%). One case-control study (Polednak 1983) and one cohort study (Torfs 1981) examined the relationship between exposure to oral HPTs in pregnancy and urogenital malformations: pooled OR = 2.63 (95% CI 0.84 to 8.28; P = 0.10; I2 = 0%) (Figure 6).\n\nA relation between the exposure to oral HPTs and musculoskeletal malformations was reported in three studies: three case-control studies (Hellstrom 1976; Janerich 1977; Lammer 1986) and one cohort study (Torfs 1981) (Figure 7), based on 2464 women, with 79 exposed to HPTs. The pooled OR = 2.24 (95% CI 1.23 to 4.08; P = 0.009; I2 = 0%). Removal of the zero study events (Torfs 1981) did not affect this result. The association of VACTERL with HPT exposure was reported in two case-control studies (Nora 1978-1 and Nora 1975), based on 135 women and infants and 27 exposed to HPTs; the OR was 7.57 (95% CI 2.92 to 19.07; P < 0.0001; I2 = 0%) (Figure 8).\n\n\nDiscussion\n\nWe found 24 articles containing 26 studies that reported the association between exposure to oral hormone pregnancy tests in mothers and malformations in their infants: 16 were case-control studies and ten were prospective cohort studies. The overall quality of the evidence, assessed by the Newcastle–Ottawa Scale, was moderate.\n\nWe found significant associations for all congenital malformations pooled and separately for congenital heart malformations, nervous system malformations, musculoskeletal malformations, and the VACTERL syndrome. Many of these pooled analyses had zero heterogeneity, and the direction of effect favoured the controls in 30 of the 32 analyses undertaken (Torfs 81 provided the only effect estimate favouring HPT exposure). The analyses were also robust to sensitivity analyses, and there was no relation between NOS score and increasing risk.\n\nBased on the assumptions that a teratogenic effect of HPTs would be mediated by actions on estrogen and progestogen receptors, and that concentrations of ethinylestradiol and norethisterone in the fetus would be too low to have a significant effect on those receptors, it has been suggested that there is no mechanistic argument for teratogenicity1. However, other unknown mechanisms might be at play. For example, Isabel Gal first reported concerns of malformations in the children of mothers exposed to HPTs in 196714, pointing out that bleeding often occurred in pregnant women soon after exposure and suggesting that that would affect the “equilibrium” of the uterus. Between 5 and 11% of exposed women had bleeding, and the RCGP survey reported induced abortions in about 10% of women15.\n\nThe drugs in Primodos were not tested for animal toxicity and teratogenicity at the time, which, although not unusual, meant that there was a gap in mechanistic understanding. A 2018 study showed that the components in Primodos are associated with dose-dependent and time-related damage in zebrafish embryos, and affect nerve outgrowth and blood vessel patterning in zebrafish14,16. Although it is difficult to compare drug actions between species, and evidence from animal studies is limited, the drugs accumulated in the zebrafish embryos, persisted for some time, and led to rapid embryonic damage14,16. In contrast, other animal studies have shown minimal effects on embryo development17. There is also evidence that estradiol and progestogens increase the expression of mRNA for isoforms of vascular endothelial growth factor (VEGF) in Ishikawa cells from human endometrial adenocarcinoma18.\n\nEstablishing causal associations in the absence of randomization can be difficult. However, the lack of randomized trials in our analysis should not be seen as a hindrance. It would have been unethical to randomize individuals to drugs with known concerns, and randomization, like systematic reviews, was not the norm at the time. Furthermore, for questions about harms, the Oxford CEBM levels of evidence puts systematic reviews of case-control studies on a par with systematic reviews of randomized trials19.\n\nHowever, observational methods have limitations20. First, interpretation can be affected by confounding factors. Although most of the studies in this review used matched controls, our analysis was based on raw data from the publications and did not adjust for confounders. Secondly, susceptibility bias can occur, as women with threatened abortions might be more likely to present and take the medication. Both of these problems can be mitigated by careful matching; 13 of the 16 studies controlled for the most important factor, item 5a on the NOS scale. Thirdly, the severity of malformations studied will have led to differing risk estimates across studies. Fourthly, inappropriate methods of ascertainment of the malformations and exposures could have introduced bias. Finally, incomplete and uneven reporting, along with publication bias (since it is likely that unreported studies exist) could introduce bias and alter the effect estimates.\n\nThe use of scoring systems to assess quality has been criticized. However, the NOS scale has been used widely in assessing the quality of non-randomized studies21–26. A NOS score between 0 and 9 has previously been used as a potential moderator in meta-regression27, and has been recommended by the Cochrane Collaboration28. A weakness of the NOS scale is the possible low agreement between assessors29. This was particularly the case when authors had limited experience in doing systematic reviews, but training, even of novices, improves agreement21.\n\nThe effects were also stable to sensitivity analyses, and changes in NOS score did not affect the risk estimates. The absence of subgroup differences between study designs for the risk estimates supports the robustness of the findings. We also tried to overcome publication bias by translation and assessment of unpublished data. The sample sizes in the studies for all congenital malformations, congenital heart disease, and nervous system malformations were sufficiently large to suggest that small unpublished studies would have little effect on the estimates unless they were highly heterogeneous. The analyses of gastrointestinal, urogenital, musculoskeletal, and VACTERL malformations were limited by their small sample sizes and low number of events: the interpretation of these effects should therefore be treated more cautiously. The significant effect observed for VACTERL should also be treated cautiously, as the confidence intervals for this effect were wide.\n\nA significant strength of this current study is its use of standard systematic review methods. By asking a focused question solely on exposure to HPTs, and excluding exposure to other hormones, we have been able to assess the heterogeneity of the effect estimates. However, as with any observational studies, there is always the possibility that an unknown confounder could be the cause of the observed difference. While such a possibility cannot be ruled out, the lack of heterogeneity means that such a confounder would potentially have to act in the same direction, despite many different confounders being collected and controlled for. Confounding factors with variable effects on the effect estimates would have probably led to a high degree of heterogeneity, which would have prevented pooling; this was not the case.\n\n\nConclusion\n\nRegulators were first made aware of the link between exposure to HPTs and congenital malformations in 1967. After 1975, the Primodos label was changed to state that the medication should not be used in pregnancy because of a risk of malformations (see Figure 9). The evidence of an association has previously been deemed weak, and previous litigation and reviews have been inconclusive. However, we believe that this systematic review shows an association of oral HPTs with congenital malformations.\n\nOur results show the benefit of undertaking systematic reviews, a study type not in routine use when most of these studies were done. For example, only one study (Greenberg 1997) out of nine reported a significant effect for all congenital malformations; the pooled estimate was significant. Much of the discussion over the associations of HPTs with congenital malformations at the time these studies were published focused on the lack of significance of individual studies12, although it was also recognized that the numbers involved were insufficient to reject the hypotheses13.\n\n\nDeclarations\n\nDataset 1: Study extraction sheet 10.5256/f1000research.16758.d22293730",
"appendix": "Grant information\n\nThe Evidence Synthesis Working Group is funded by the National Institute for Health Research School for Primary Care Research (NIHR SPCR) [ProjectNumber 390]. The views expressed are those of the author(s) and not necessarily those of the NIHR, the NHS or the Department of Health\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1. Committee on Safety of Medicines report 1975-colour pdf.\n\nClick here to access the data\n\nSupplementary File 2. Search Strategy.\n\nClick here to access the data\n\nSupplementary File 3. List of hormone pregnancy tests (HPTs) included in evidence search.\n\nClick here to access the data\n\nSupplementary File 4. Completed Meta-analysis Of Observational Studies in Epidemiology (MOOSE) checklist.\n\nClick here to access the data\n\nSupplementary File 5. Full references for included studies.\n\nClick here to access the data\n\nSupplementary File 6. Funnel Plots.\n\nClick here to access the data\n\n\nReferences\n\nWebsite 1: Medicines and Healthcare Products Regulatory Agency. (accessed 1 Jun 2018). Reference Source\n\nWebsite: Medicines and Healthcare products Regulatory Agency. 2015 Call for Evidence. (accessed 1 Jun 2018). Reference Source\n\nReport of the Commission on Human Medicines’ Expert Working Group on Hormone Pregnancy Tests. GOV.UK. (accessed 24 May 2018). Reference Source\n\nAnger after report finds birth defects not caused by hormone pregnancy tests. accessed 2/8/18. Reference Source\n\nReview launched to respond to patient concerns about NHS treatments. (accessed 2/08/2018). Reference Source\n\nOttawa Hospital Research Institute. (accessed 1 Jun 2018). Reference Source\n\nLunny C, Knopp-Sihota JA, Fraser SN: Surgery and risk for multiple sclerosis: a systematic review and meta-analysis of case-control studies. BMC Neurol. 2013; 13: 41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nErnst E, Pittler MH: Re-analysis of previous meta-analysis of clinical trials of homeopathy. J Clin Epidemiol. 2000; 53(11): 1188. PubMed Abstract | Publisher Full Text\n\nEgger M, Smith GD: meta-analysis bias in location and selection of studies. BMJ. 1998; 316: 61–6. Publisher Full Text\n\nHiggins JP, Thompson SG: Quantifying heterogeneity in a meta-analysis. Stat Med. 2002; 21(11): 1539–58. PubMed Abstract | Publisher Full Text\n\nStroup DF, Berlin JA, Morton SC, et al.: Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000; 283(15): 2008–12. PubMed Abstract | Publisher Full Text\n\nTorfs CP, Milkovich L, van den Berg BJ: The relationship between hormonal pregnancy tests and congenital anomalies: a prospective study. Am J Epidemiol. 1981; 113(5): 563–74. PubMed Abstract | Publisher Full Text\n\nGal I, Kirman B, Stern J: Hormonal pregnancy tests and congenital malformation. Nature. 1967; 216(5110): 83. PubMed Abstract | Publisher Full Text\n\nGal I: Risks and benefits of the use of hormonal pregnancy test tablets. Nature. 1972; 240(5378): 241–2. PubMed Abstract | Publisher Full Text\n\nBrown S, Fraga LR, Cameron G, et al.: The Primodos components Norethisterone acetate and Ethinyl estradiol induce developmental abnormalities in zebrafish embryos. Sci Rep. 2018; 8(1): 2917. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaier WE, Herman JR: Pharmacology and toxicology of ethinyl estradiol and norethindrone acetate in experimental animals. Regul Toxicol Pharmacol. 2001; 34(1): 53–61. PubMed Abstract | Publisher Full Text\n\nArcher DF, Navarro FJ, Leslie S, et al.: Effects of levonorgestrel, medroxyprogesterone acetate, norethindrone, and 17beta-estradiol on vascular endothelial growth factor isomers 121 and 165 in Ishikawa cells. Fertil Steril. 2004; 81(1): 165–70. PubMed Abstract | Publisher Full Text\n\nOCEBM Levels of Evidence. CEBM. 2016; (accessed 29 May 2018). Reference Source\n\nSackett DL: Bias in analytic research. J Chronic Dis. 1979; 32(1–2): 51–63. PubMed Abstract | Publisher Full Text\n\nOremus M, Oremus C, Hall GB, et al.: Inter-rater and test-retest reliability of quality assessments by novice student raters using the Jadad and Newcastle-Ottawa Scales. BMJ Open. 2012; 2(4): pii: e001368. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWallis CJ, Mahar AL, Choo R, et al.: Second malignancies after radiotherapy for prostate cancer: systematic review and meta-analysis. BMJ. 2016; 352: i851. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCroles FN, Nasserinejad K, Duvekot JJ, et al.: Pregnancy, thrombophilia, and the risk of a first venous thrombosis: systematic review and bayesian meta-analysis. BMJ. 2017; 359: j4452. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHackshaw A, Morris JK, Boniface S, et al.: Low cigarette consumption and risk of coronary heart disease and stroke: meta-analysis of 141 cohort studies in 55 study reports. BMJ. 2018; 360: j5855. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoland MV, Ervin AM, Friedman DS, et al.: Comparative effectiveness of treatments for open-angle glaucoma: a systematic review for the U.S. Preventive Services Task Force. Ann Intern Med. 2013; 158(4): 271–9. PubMed Abstract | Publisher Full Text\n\nHarausz EP, Garcia-Prats AJ, Law S, et al.: Treatment and outcomes in children with multidrug-resistant tuberculosis: A systematic review and individual patient data meta-analysis. PLoS Med. 2018; 15(7): e1002591. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVeronese N, Cereda E, Solmi M, et al.: Inverse relationship between body mass index and mortality in older nursing home residents: a meta-analysis of 19,538 elderly subjects. Obes Rev. 2015; 16(11): 1001–1015. PubMed Abstract | Publisher Full Text\n\nBae JM: A suggestion for quality assessment in systematic reviews of observational studies in nutritional epidemiology. Epidemiol Health. 2016; 38: e2016014. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHartling L, Milne A, Hamm MP, et al.: Testing the Newcastle Ottawa Scale showed low reliability between individual reviewers. J Clin Epidemiol. 2013; 66(9): 982–93. PubMed Abstract | Publisher Full Text\n\nHeneghan C, Aronson JK, Spencer E, et al.: Dataset 1 in: Oral hormone pregnancy tests and the risks of congenital malformations: a systematic review and meta-analysis. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16758.d222937"
}
|
[
{
"id": "40053",
"date": "05 Nov 2018",
"name": "David Healy",
"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\nI am fully supportive of this article on the effects of hormone pregnancy tests as it stands. I have no substantive criticism of the content or methods.\nI am of course interested in why the regulator (MHRA) did not find comparable results but this is not a matter that should be addressed in this article.\nThere is one extra point that this article may speak to which is that from some time it was thought that teratogens caused signature defects - such as the phocomelia of thalidomide. This may now be a minority position (I'm not sure of this point). The findings here do not support that point of view. I can understand if the authors may think that commenting on this point is a matter for others or for another article; I mention it for consideration.\n\nI have one very minor point about the layout which is that in the column where the numbers of women recruited to various studies is mentioned, the right justification of paragraphs leads to an odd spacing between 28 thousand and 671 - this doesn't apply when the page is resized.\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": "4363",
"date": "17 Jan 2019",
"name": "Carl Carl",
"role": "Author Response",
"response": "Many thanks for these positive comments.Signature defects are only likely to occur if the timing of exposure is the same in all cases. However, when the timing of exposure varies, fetuses will be affected in different ways, depending on the tissues that are developing at the time of exposure, giving rise to a variety of malformations. The large range of times of exposure during embryogenesis determines which developmental processes are most affected, resulting in a wide variety of potential defects as seen in our review and in alleged Primodos survivors. Furthermore, different fetuses may have different epigenetic susceptibilities to different teratogenic outcomes. The Primodos components norethisterone acetate and ethinyl estradiol induce developmental abnormalities in zebrafish embryos [Brown S, Fraga LR, Cameron G, Erskine L, Vargesson N. Sci Rep. 2018 Feb 13;8(1):2917. doi: 10.1038/s41598-018-21318-9]. Brown's data in Zebrafish show acetate and ethinylestradiol teratogenicity depends on dose and the embryonic stage of development, embryos at an early stage being more sensitive than those at a later stage.The comments are interested in why the regulator (MHRA) did not find comparable results but this is not a matter that should be addressed in this article. We agree with this issue - no change required.We have removed the spacing between numbers to eliminate the odd spacing effect."
}
]
},
{
"id": "40291",
"date": "29 Nov 2018",
"name": "Jesse Olszynko-Gryn",
"expertise": [
"Reviewer Expertise History of 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\nThis is a timely and much-needed paper that deserves to be widely read and cited. It provides the first systematic review and meta-analysis of old epidemiological data pointing towards a long-acknowledged association between HPTs and birth defects. Most of the paper is devoted to apparently rigorous statistical analysis. We leave constructive criticism of the statistics to other, more appropriately qualified reviewers. Instead, we confine our comments to the historical context and factual details presented in the paper. These, on the whole, are entirely satisfactory. But some minor errors — that do not significantly detract from the overall argument — should be amended:\n\n'Oral hormone pregnancy tests (HPTs), such as Primodos, containing ethinylestradiol and high doses of norethisterone, were given to over a million women from 1958 to 1978’ (p. 1).\n\nIt is worth clarifying that HPTs were available as injections from 1950 and in tablet form (e.g., Schering’s Orasecron, Roussel’s Amenorone Forte), in the UK, from at least 1956. See, for example, Britton (19561); and https://archive.org/details/b19974760M4180/page/n45?q=amenorone+1956. For an extended discussion, see Olszynko-Gryn (2014), available for download here. Furthermore, not all HPTs contained norethisterone; different companies used other types of synthetic progesterone, and the same goes for ethinylestradiol.\n\n‘Oral hormone pregnancy tests (HPTs), such as Primodos (known as Duogynon in Germany), were used from 1958 to 1978, before urine pregnancy tests were available’ (p. 3).\nContrary to popular belief, urine pregnancy tests were in fact widely though unevenly available between 1958 and 1978 and HPTs were never the dominant method of pregnancy testing. For a detailed timeline of pregnancy testing in the UK, please see Olszynko-Gryn et al. (20182), esp. pp. 35-36. It would also be helpful to clarify that HPTs were removed from UK market in 1978, but earlier and later elsewhere. See Olszynko-Gryn et al. (20182) for details (pp. 41-42).\n\n‘The test principle was that menstruation would be induced in those who were not pregnant’ (p.3).\n\nAt the time HPTs were variously described as ‘clinical’, ‘hormonal’, or ‘withdrawal bleeding’ pregnancy tests and it would be more precise to refer to their effect as inducing menstrual-like withdrawal bleeding, which is not identical to menstruation.\n\nWorth mentioning that Gal 1967, though a highly significant intervention, was not the first published warning against HPTs; these began to appear as early as 1956, in response to marketing literature aimed at GPs. See Britton (19561) and Olszynko-Gryn et al. (20182), p. 36.\n\n‘However, we believe that this systematic review shows an association of oral HPTs with congenital malformations’ (p. 17).\n\nMore optionally, the authors might consider reflecting on the extent to which the association they identify implies a causal association. An association between the use of HPTs and birth defects has long been recognised and was rarely in dispute. Many experts explained the association in terms of a suspected though as unknown direct mechanistic effect of HPTs on the developing human embryo. Others, however, preferred to explain the association in terms of underlying factors, e.g., a patient history of miscarriage or birth defects. This view, which still has traction in some quarters, is discussed to some extent in Olszynko-Gryn et al. (20182) (pp. 39-41). The authors might usefully offer a fresh perspective based on their findings, in the Conclusion and/or in the interesting discussion of unknown mechanisms on p. 16.\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? I cannot comment. A qualified statistician is required.\n\nAre the conclusions drawn adequately supported by the results presented in the review? Yes",
"responses": [
{
"c_id": "4362",
"date": "17 Jan 2019",
"name": "Carl Carl",
"role": "Author Response",
"response": "Many thanks for these positive comments. We have amended the introduction with the following text: 'Oral hormone pregnancy tests (HPTs), such as Primodos (known as Duogynon in Germany), were available as injections from 1950 and in tablet form in the UK from 1956 onwards, before the modern forms of urine pregnancy tests became available [1]’ We have cited: Olszynko-Gryn J, Bjørvik E, Weßel M, Jülich S, Jean C. A historical argument for regulatory failure in the case of Primodos and other hormone pregnancy tests. Reprod Biomed Soc Online. 2018 Oct 23;6:34-44. Doi: 10.1016/j.rbms.2018.09.003. eCollection 2018 Aug. We have amended the introduction text as per the reviewer's suggestion :\"Warnings about HPTs in pregnancy first emerged in 1956: accumulating concerns over an increased risk of malformations led to their withdrawal in a number of countries at different times. Norway cancelled the indication in pregnancy for HPTs in 1970; the UK did so in 1978, when the manufacturers of Primodos, Schering AG (taken over by Bayer AG in 2008), voluntarily stopped marketing the product; in Germany, Duogynon was taken off the market in 1981 [ref Olszynko-Gryn J].'We have amended the introduction text as per the reviewer's suggestion:‘The test principle was that they would induce bleeding similar to menstruation in those who were not pregnant.’'Warnings about HPTs in pregnancy first emerged in 1956.'And referenced the Britton H.G. Pregnancy test. Br. Med. J. 1956;2(18 Aug.):419. paper The benefits of our systematic review include that it quantifies the magnitude of the association and tests the robustness of this association across multiple studies by meta-analysis. We, therefore, perceive that they are rationale and the objectives are clear."
}
]
},
{
"id": "40887",
"date": "14 Dec 2018",
"name": "Olalekan Uthman",
"expertise": [
"Reviewer Expertise 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\nThe manuscript reads well. The manuscript was well written. It has the potential to add to the body of knowledge in the field. The method was described in detail to allow for replication of the study. I have no major concerns.\n\nMy only comment: the author should change the abstract conclusion to: “This systematic review and meta-analysis of observational studies shows that the use of oral HPTs in pregnancy is associated with increased risks of congenital malformations”.\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": "4361",
"date": "17 Jan 2019",
"name": "Carl Carl",
"role": "Author Response",
"response": "Many thanks for these positive comments. The methods and results are clear about the type of studies included, and we consider the conclusions do not require the addition of the study type. This is in line with previous systematic reviews our group has published."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1725
|
https://f1000research.com/articles/8-108/v1
|
25 Jan 19
|
{
"type": "Research Article",
"title": "Broad-spectrum capture of clinical pathogens using engineered Fc-mannose-binding lectin enhanced by antibiotic treatment",
"authors": [
"Benjamin T. Seiler",
"Mark Cartwright",
"Alexandre L. M. Dinis",
"Shannon Duffy",
"Patrick Lombardo",
"David Cartwright",
"Elana H. Super",
"Jacqueline Lanzaro",
"Kristen Dugas",
"Michael Super",
"Donald E. Ingber",
"Benjamin T. Seiler",
"Mark Cartwright",
"Alexandre L. M. Dinis",
"Shannon Duffy",
"Patrick Lombardo",
"David Cartwright",
"Elana H. Super",
"Jacqueline Lanzaro",
"Kristen Dugas",
"Michael Super"
],
"abstract": "Background: Fc-mannose-binding lectin (FcMBL), an engineered version of the blood opsonin MBL that contains the carbohydrate recognition domain (CRD) and flexible neck regions of MBL fused to the Fc portion of human IgG1, has been shown to bind various microbes and pathogen-associated molecular patterns (PAMPs). FcMBL has also been used to create an enzyme-linked lectin sorbent assay (ELLecSA) for use as a rapid (<1 h) diagnostic of bloodstream infections. Methods: Here we extended this work by using the ELLecSA to test FcMBL’s ability to bind to more than 190 different isolates from over 95 different pathogen species. Results: FcMBL bound to 85% of the isolates and 97 of the 112 (87%) different pathogen species tested, including bacteria, fungi, viral antigens and parasites. FcMBL also bound to PAMPs including, lipopolysaccharide endotoxin (LPS) and lipoteichoic acid (LTA) from Gram-negative and Gram-positive bacteria, as well as lipoarabinomannan (LAM) and phosphatidylinositol mannoside 6 (PIM6) from Mycobacterium tuberculosis. Conclusions: The efficiency of pathogen detection and variation between binding of different strains of the same species could be improved by treating the bacteria with antibiotics, or mechanical disruption using a bead mill, prior to FcMBL capture to reveal previously concealed binding sites within the bacterial cell wall. As FcMBL can bind to pathogens and PAMPs in urine as well as blood, its broad-binding capability could be leveraged to develop a variety of clinically relevant technologies, including infectious disease diagnostics, therapeutics, and vaccines.",
"keywords": [
"Mannose-binding lectin",
"Lipopolysaccharide",
"Lipoteichoic acid",
"Biomarker",
"Bacteria",
"Antibiotics",
"Diagnostic"
],
"content": "Introduction\n\nMannose-binding lectin (MBL) is a key host-defense protein associated with the lectin pathway of the innate immune system1, and deficiency of MBL can lead to increased susceptibility to a wide-spectrum of infectious diseases2–4. MBL functions as a calcium-dependent, pattern-recognition opsonin that binds a range of carbohydrate molecules associated with the surfaces or cell walls of many different types of pathogens5. Collectively these microbial surface carbohydrate molecules, including for example, lipopolysaccharide endotoxin (LPS) and lipoteichoic acid (LTA), are referred to as pathogen-associated molecular patterns (PAMPs)6,7. MBL has the intrinsic ability to distinguish foreign PAMPs from self, subsequently activating the complement system and providing protection via antibody-dependent and independent mechanisms8,9.\n\nDue to the evolutionarily conserved recognition carbohydrate moieties of PAMPs, MBL is a broad-spectrum opsonin that can bind over 90 different species of pathogens, including Gram-negative and Gram-positive bacteria, fungi, viruses, and parasites10–14. MBL binding to these various pathogens has been demonstrated by means of flow cytometry14,15, radio-immunoassay13,16, enzyme-linked immunosorbent assay (ELISA)13,17, immunofluorescence and scanning electron microscopy (SEM)18, and Saccharomyces cerevisiae-induced MBL activation and bystander lysis of chicken erythrocytes19. However, many discrepancies in MBL binding have been described, depending on the method used. For example, use of flow cytometry revealed little to no MBL binding to Pseudomonas aeruginosa, while others have reported good binding of MBL to Pseudomonas aeruginosa using a hemolytic assay15,19.\n\nWe set out to address these conflicting results by leveraging the recent development of an engineered version of MBL that contains the carbohydrate recognition domain (CRD) and flexible neck regions of MBL fused to the Fc portion of human IgG1, which is known as FcMBL20. The engineered FcMBL lacks the regions of the native molecule that interact with MBL-associated serine proteases (MASPs) that activate complement and promote blood coagulation, and thus, it can be used to capture PAMPs from complex biological fluids, such as blood and urine, without activating effector functions of complement, coagulation, and phagocytosis. We have previously used FcMBL in extracorporeal therapies, such as hemofiltration, and in diagnostics to capture and detect Staphylococcus aureus from osteoarticular and synovial fluids of infected patients20–22. In the present study, we used a previously described sandwich enzyme-linked lectin sorbent assay (ELLecSA) in which both live and fragmented pathogens (PAMPs) are captured using FcMBL conjugated to magnetic beads and then detected with horseradish peroxidase (HRP)-labeled MBL23. This ELLecSA has demonstrated FcMBL binding to 85% (47 of 55) of pathogen species previously tested, and enabled rapid diagnosis of bloodstream infections by capturing and detecting PAMPs in whole blood from human patients23. Here we extend this testing to include 69 isolates from 57 more pathogen species using the ELLecSA. In total, we measure direct binding of FcMBL to over 190 pathogen isolates from over 95 different pathogen species, including bacteria, fungi, viral antigens, parasites, and bacterial cell wall molecules. As a result of this more extensive analysis, we demonstrate that FcMBL detection increases to 85% of the isolates and 87% (97 out of 112) of the pathogen species tested. Furthermore, we show that antibiotic treatment or mechanical disruption of the bacterial pathogens exposes previously concealed FcMBL binding sites on cell walls, thereby increasing the efficiency of pathogen detection and reducing variation between binding of different strains of the same species. We also show that FcMBL can detect PAMPs in urine as well as blood, making this potential diagnostic technology highly synergistic with standard of care antibiotic therapy.\n\n\nResults\n\nWe first set out to determine the range of pathogens that FcMBL can capture by screening multiple species of bacteria, fungi, viral antigens, and parasites using the ELLecSA detection technology. In the FcMBL ELLecSA, pathogen materials in experimental samples are captured with FcMBL immobilized on superparamagnetic beads (1 µm diameter), magnetically separated, washed, detected with human MBL linked to horseradish peroxidase (HRP), magnetically separated again, washed, and then tetramethylbenzidine (TMB) substrate is added to measure the amount of pathogen material bound (Figure 1A). Results were quantified by interpolating against an internal standard curve generated using yeast mannan in buffer [50 mM Tris-HCl, 150 mM NaCl, 0.05% Tween-20, 5 mM CaCl2, pH 7.4 (TBST 5 mM CaCl2)] (Figure 1B). In addition, we show that while the curve sensitivity is reduced when mannan is spiked into whole human blood, the limit of detection remained similar (1 ng/ml) as it does in buffer (Figure 1B).\n\n(A) Diagrammatic representation of the FcMBL ELLecSA methodology. Live and fragmented bacteria are captured by FcMBL, which has been coated on superparamagnetic beads via an N-terminal aminooxy-biotin on the Fc to allow oriented attachment to streptavidin (FcMBL Bead). FcMBL beads with captured bacteria are then magnetically separated and detected using recombinant human MBL linked to horseradish peroxidase (Human MBL-HRP). Tetramethylbenzidine (TMB) substrate is added to quantify captured bacteria and results are read at OD 450nm. (B) Mannan standard curve showing FcMBL binding to mannan in buffer and spiked into whole human blood (<24 h from draw), measured at optical density 450 nm. The mannan standard curve in buffer serves as an internal assay control, and is used to determine FcMBL binding to pathogen material in units of PAMPs/ml (where 1 ng/ml of mannan bound by FcMBL is equivalent to 1 PAMP unit. PAMP units are then multiplied by the dilution factor of the sample volume to give PAMPs/ml).\n\nInitially our focus was on screening bacteria and as such we compiled a comprehensive list of clinically relevant bacterial pathogens (Table 1)23. When we screened 82 different species of bacteria to compare FcMBL binding to live versus fragmented cells, we found that FcMBL detected 59 out of 82 live microbes (72%) and that more could be detected (70 out of 82; 85%) after they were fragmented by treating with antibiotics, or mechanically disrupted using zirconia/silica beads in a mixer mill (Table 1)23. The antibiotics we used in this study were clinical grade cefepime, ceftriaxone, meropenem, amikacin, gentamicin, and vancomycin, to provide enough coverage to target this diverse range of bacteria. We dosed each bacterial class with a single appropriate antibiotic dose (≤1 mg/ml) to obtain acute fragmentation within 4 hours.\n\nMultiple species of bacteria, including multiple isolates (number of isolates), were screened by FcMBL ELLecSA to determine FcMBL binding. Total number detected of both live and fragmented bacterial isolates is shown. Fungi were screened and total number detected for live isolates shown. Purified or inactivated viral antigens, parasites, and bacterial antigens were tested directly in TBST 5 mM CaCl2 buffer, and number detected shown. Test samples were performed in duplicate. NT, not tested.\n\nTo determine if inducing bacterial fragmentation via antibiotic treatment or bead milling would reduce variation in FcMBL binding between different strains of the same species, we screened 134 isolates from 21 of the 88 Gram-positive and Gram-negative bacterial species, including strains tested by Cartwright et al. (Figure 2)23. As before, FcMBL bound a greater proportion of the pathogens when fragmented (113/134 = 84%) than when live and intact (77/134 = 57%). For some bacterial species such as Enterobacter cloacae, Escherichia coli, Klebsiella oxytoca, and Klebsiella pneumoniae we found that mechanical disruption or antibiotic-induced fragmentation greatly increased FcMBL binding, whereas other bacteria like Pseudomonas aeruginosa, Yersinia pseudotuberculosis, and MRSA bound equally well when live and intact (Figure 2). With the exception of Proteus mirabilis and Enterococcus faecalis, which FcMBL did not bind at all, the capture of fragmented bacteria was equal to or greater than that of live bacteria (Figure 2).\n\nThe graph is divided between Gram-negative bacterial isolates [Gram (-)] and Gram-positive bacterial isolates [Gram (+)]. Data are presented as the number of bacterial isolates bound live and fragmented within the total isolates tested for each species: A. baumannii (n = 3), E. aerogenes (n = 4), E. cloacae (n = 8), E. coli (n = 28), K. oxytoca (n = 7), K. pneumoniae (n = 9), P. mirabilis (n = 4), P. aeruginosa (n = 6), S. enteriditis (n = 2), S. paratyphi A (n = 2), S. typhimurium (n = 6), S. marcescens (n = 5), Y. pseudotuberculosis (n = 6), E. faecalis (n = 5), S. aureus (n = 17), S. aureus (MRSA) (n = 3), S. epidermidis (n = 4), S. hominis (n = 2), S. agalactiae (n = 2), S. pneumoniae (n = 9), S. viridans (n = 2).\n\nThese findings are consistent with past studies that showed the efficiency of MBL binding to live bacteria differs between isolates from the same bacterial genus and species, possibly due to differences in encapsulation15,16. To illustrate that the heterogeneity of MBL binding to live isolates of the same species can be reduced by using antibiotics to disrupt previously cryptic binding sites, we used two different isolates of Escherichia coli, and two different isolates of Streptococcus pneumoniae. E. coli 41949 and S. pneumoniae 3 exhibited equivalent FcMBL binding whether they were live or fragmented with antibiotics (1 mg/ml cefepime or ceftriaxone, respectively, for 4 hours), whereas fragmented forms of E. coli RS218 and S. pneumoniae 19A isolates bound much more effectively to FcMBL than living forms (Figure 3A–D). This difference was further supported visually using scanning electron microscopy (SEM) in which magnetic FcMBL beads could be seen to bind both live and fragmented versions of E. coli 41949 and S. pneumoniae 3, but with E. coli RS218 and S. pneumoniae 19A, the FcMBL beads only bound to fragmented material (Figure 3E–H). FcMBL binding increases upon fragmentation, which is correlated with LPS release measured using a limulus amebocyte lysate (LAL) assay: equal amounts of LPS were detected for E. coli 41949 whether live or fragmented, whereas LPS levels were higher in antibiotic treated E. coli RS218 (Figure 3I, J). These results suggest that antibiotic treatment results in exposure of previously cryptic PAMPs in the cell wall, including toxins such as LPS and LTA, which leads to greatly increased binding of FcMBL.\n\n(A–D) PAMPs/ml detection by FcMBL ELLecSA of both live (107 CFU/ml) and fragmented bacteria using antibiotics (cefepime 1 mg/ml or ceftriaxone 1 mg/ml). (A) E. coli 41949, (B) S. pneumoniae 3, (C) E. coli RS218, and (D) S. pneumoniae 19A. (E–H) Scanning electron microscopy images showing FcMBL bead (128 nm) capture of both live and fragmented bacteria using antibiotics (cefepime 1 mg/ml or ceftriaxone 1 mg/ml). (E) E. coli 41949, (F) S. pneumoniae 3, (G) E. coli RS218, and (H) S. pneumoniae 19A. (I,J) LPS endotoxin measurement (LAL assay) using 107 CFU/ml of both live and fragmented bacteria using antibiotics (cefepime 1 mg/ml). (I) E. coli 41949 and (J) E. coli RS218.\n\nIn these studies, we found that 12 bacterial species, including multiple species of Enterococcus and Proteus, failed to bind to FcMBL even when treated for 4 hours with combinations of antibiotics (500 µg/ml vancomycin and 500 µg/ml amikacin for Gram-positive isolates or 500 µg/ml cefepime and 500 µg/ml amikacin for Gram-negative isolates) (Table 1)23. Importantly however, FcMBL was able to detect 84% (172 out of 204) of the bacterial isolates (Figure 4), which includes 9 of the 10 pathogens responsible for most healthcare-associated infections in acute care hospitals in the U.S., with Enterococcus species being the one exception24.\n\n(Left) Percent of live and fragmented Gram-positive (n = 78) and Gram-negative (n = 117) bacterial isolates bound by FcMBL ELLecSA. (Right) Chart showing total number of isolates tested by FcMBL ELLecSA for bacteria, fungi, viral antigens, parasites, and bacterial antigens, total number FcMBL detected, and total percent detected overall. ***p-value < 0.0001; Pearson’s chi-squared test. NS, not significant.\n\nWe further explored FcMBL’s ability to bind bacterial cell wall components because when bead mill or antibiotics were used to disrupt the membranes of Gram-negative isolates (n = 117), there was a significant boost in FcMBL detection efficiency with fragmented cells (89%) versus live intact cells (58%) (Figure 4), which is likely due to exposure of LPS that is present in high concentrations in their cell wall25. Similarly, FcMBL also detected a greater percentage (77%) of fragmented Gram-positive isolates (n = 78), versus live intact cells (65%) (Figure 4). Thus, to better understand some of the major targets that FcMBL binds when bacteria are fragmented, we extended our analysis using purified samples of LPS and LTA25,26.\n\nUsing the ELLecSA, we screened LPS purified from Gram-negative bacteria (Serratia marcescens, Klebsiella pneumoniae, and Salmonella enterica serovar enteritidis), as well as LTA from Gram-positive bacteria (Enterococcus hirae, Staphylococcus aureus, and Streptococcus pyogenes) in TBST 5 mM CaCl2, as well as in more clinically relevant human whole blood samples. We found that FcMBL was able to detect LPS from all 3 Gram-negative species, with S. marcescens being the best (1 ng/ml limit of detection in buffer and 3.9 ng/ml in blood) (Figure 5A–C). FcMBL also bound to E. hirae LTA very well (15.6 ng/ml limit of detect in buffer and 62.5 ng/ml in blood) (Figure 5D), which is consistent with past findings27. Also consistent with past findings, we found that FcMBL binds S. aureus LTA through the carbohydrate recognition domain (Figure 5E and Supplementary Figure 1)28,29. Notably, however, FcMBL also bound S. pyogenes LTA (Figure 5F), which is in contrast to the past finding that MBL binds well to E. hirae LTA but poorly to LTA from S. pyogenes due to lack of glycosyl substituents27,30.\n\nLPS from (A) S. marcescens, (B) K. pneumoniae, and (C) S. enterica serovar enteritidis, and LTA from (D) E. hirae, (E) S. aureus, and (F) S. pyogenes were spiked into either TBST 5 mM CaCl2 buffer or whole human blood at indicated concentrations.\n\nWe next tested FcMBL’s ability to bind lipoarabinomannan (LAM) and its biosynthetic precursors, phosphatidylinositol mannoside 1 & 2 and 6 (PIM1,2 and PIM6) from Mycobacterium tuberculosis (TB) strain H37Rv31,32. LAM released from metabolically replicating or degrading TB bacteria has been detected in both blood and urine33,34. Thus, we assessed the ability of FcMBL to capture and detect LAM, as well as PIM1,2 and PIM6, spiked into both of these complex biological fluids as well as buffer. Our initial screen in buffer confirmed that FcMBL can detect LAM and PIM6 at levels down to 4 ng/ml, but it did not detect PIM1,2 (Figure 6A–C). FcMBL also bound to LAM in both blood and urine but its binding sensitivity was reduced as it could only detect 15.6 ng/ml and 250 ng/ml, respectively. FcMBL binding to PIM6 exhibited a similar sensitivity in buffer, but it could only detect 62.5 ng/ml and 15.6 ng/ml in blood and urine, respectively.\n\nGlycolipids spiked into (A) buffer (TBST 5mM CaCl2), (B) whole human blood, and (C) urine. FcMBL detected LAM and PIM6 at 4 ng/ml in buffer, but not PIM1,2. Sensitivity of LAM and PIM6 is reduced in whole human blood and urine.\n\nIn addition to screening multiple bacteria, we also tested FcMBL’s ability to bind to 12 different species of fungi, 10 viral antigens, 2 species of parasites, and 6 purified bacterial cell wall antigens (Table 1)23. In contrast to studies with bacteria, FcMBL was found to bind 100% of live fungal cells from all 12 species and 13 isolates tested (Figure 4). Of the two parasites tested in this preliminary analysis, only Trichomonas vaginalis was bound by FcMBL, whereas 80% of the viral antigens screened and 92% of the purified bacterial cell wall antigens were detected (Figure 4 and Figure 7). The handful of pathogen material FcMBL did not detect included the E1 protein from Chikungunya virus, the NS1 protein from Tick-borne encephalitis virus, Plasmodium falciparum, and PIM1,2 from TB. In total, the overall FcMBL binding profile respectively detected 85% (194 out of 229) of isolates, and 87% (97 out of 112) of the different pathogen species tested.\n\nFcMBL ELLecSA screening of Chikungunya E1 (3.0 µg/ml), Cytomegalovirus (CMV) (107 PFU), Dengue serotype 1 VLP (0.36 µg/ml), Human immunodeficiency virus (HIV) gp120 (0.1 µg/ml), Ebola GP1 (0.2 µg/ml), Influenza H1N1 HA (0.1 µg/ml), Influenza H1N1 NA (1 µg/ml), Respiratory syncytial virus (RSV) glycoprotein g (10.0 µg/ml), Tick-borne Encephalitis NS1 (5 µg/ml), and Zika lysate (0.24 µg/ml).\n\nRaw data for the present study are available on OSF29.\n\n\nDiscussion\n\nMBL has been reported to bind to over 90 different pathogen species as well as PAMPs released from these microbes based on studies in which binding was assessed by means of flow cytometry, ELISA, radio-immunoassay, immunofluorescence and SEM, or hemolytic assays12–19; however, different results have been obtained with different methods. Here we explored the broad-spectrum binding capabilities of an engineered form of MBL, known as FcMBL, using a previously described magnetic ELLecSA detection assay to quantify binding of MBL to over 190 isolates from over 95 different pathogen species, which include bacteria, fungi, viral antigens, parasites, and bacterial cell wall antigens23. FcMBL was previously shown to bind PAMPs released from 47 of 55 (85%) pathogen species tested, including 38 species of bacteria and 9 species of fungi23. The FcMBL ELLecSA also was able to detect infectious PAMPs in whole blood of sepsis patients, regardless of antibiotic therapy (blood culture positive or negative) with a detection sensitivity and specificity of 85% and 89%, respectively23. In the present study, we utilized the ELLecSA to extend this testing to include 69 isolates from 57 more pathogen species. In total, our results confirm that FcMBL binds to 85% of the isolates and 97 of the 112 species tested, which corresponds to an increased detection sensitivity of 87%.\n\nMBL binding to different clinical bacterial isolates of the same species has previously produced conflicting results15,16. These same studies also described that most Gram-negative isolates (encapsulated strains) bound little or no MBL. We have reported similar results as we previously found that FcMBL only bound 38% of live clinical E. coli isolates tested; however, upon fragmentation and release of PAMPs, FcMBL detection of these same isolates increased to 92%23. Broader examination of Gram-negative bacteria in the present study revealed a similar pattern: FcMBL only detected 68/117 (58%) of live isolates, but when these same microbes were treated with antibiotics or bead mill, the detection sensitivity significantly increased to 89% (104/117 isolates). Apparently, by treating bacteria with antibiotics or bead mill we were able to disrupt the encapsulated cell wall, exposing and presenting previously hidden PAMPs, thereby increasing binding and reducing variability between isolates within the same bacterial species. However, even with cell wall disruption, FcMBL did not bind 12 bacterial species, including multiple isolates of E. faecalis and P. mirabilis. These microbes likely lack the complex polysaccharide antigens which FcMBL and MBL bind. Alternatively, the binding sites might still be present, but if they are, they remain inaccessible due to the unique structure of their cell wall (e.g. carbohydrate conformation, sugar density or composition). Alternatively, the antibiotics we used might not be optimal for disrupting the cell walls of these bacteria.\n\nTo emphasize the ability of FcMBL to be used to detect the presence of a systemic pathogenic infection even when blood cultures are negative, we tested its ability to bind LPS and LTA that are major PAMP-associated toxins released by multiple species of bacteria. FcMBL was able to detect both LPS and LTA from all 6 bacterial species tested in both buffer and blood, although detection sensitivity was consistently higher in buffer. In addition, we explored whether FcMBL binds to the antigenic PAMPs, LAM, PIM1,2, and PIM6 from M. tuberculosis H37Rv because these are active virulence factors associated with TB pathogenesis, and hence, they are critically important targets for point-of-care diagnostic and vaccine applications35–38. We found that FcMBL can detect LAM and PIM6, but not PIM1,2, in buffer, urine, and blood; this difference in binding is likely due to the fact that PIM1,2 has 4 fewer branched mannose residues than PIM639.\n\nPreliminary viral antigen detection in the ELLecSA was encouraging, demonstrating FcMBL binding to 8 of 10 (80%) species. We screened these viral antigens at a range of 0.1 µg/ml to 10 µg/ml that is within or below the range at which viral proteins are known to induce an immune response in vaccines40–42.\n\nWhile the FcMBL ELLecSA cannot distinguish between different types of infections, we have previously shown that it can be used to rapidly (<1 h) detect the presence of blood infections in whole blood samples from patients suspected of sepsis regardless of whether or not they have positive blood cultures23. Use of the FcMBL ELLecSA in conjunction with other tools, such as C-reactive protein (CRP) and Procalcitonin (PCT), could help inform and assist the physician in deciding if there is an infection, whether hospitalization is critical, or whether antibiotics should be administered when a patient first enters a care center. In addition, we have found that we can combine the FcMBL capture of PAMPs with additional molecular diagnostic tools, such as PCR, to distinguish different pathogen types based on molecular composition of the samples22.\n\nIn summary, FcMBL’s ability to both bind to numerous types of infectious pathogens and capture many of the cell wall PAMPs released by these microbes when treated with antibiotics in complex biological fluids, further demonstrates the potential value of using FcMBL capture for rapid detection of bloodstream infections, even when blood cultures are negative. As a result of extending our FcMBL-based ELLecSA studies to a broader range of different pathogens, we determined a higher (87%) binding efficiency than that observed in preliminary studies. We also now better understand how FcMBL interacts with different bacterial strains of the same species when mechanically disrupted or fragmented by antibiotics. To our knowledge, this is the broadest range and largest number of pathogens and PAMPs that have been shown can be detected by a single blood opsonin or lectin. The ability of FcMBL to detect cell wall fragments also synergizes well with standard of care antibiotic therapy, and it’s broad-range pathogen capture and detection can be leveraged to develop a wide range of infectious disease diagnostics, therapeutics, and vaccines.\n\n\nMethods\n\nBacteria, fungi, viral antigens, parasites, and bacterial cell wall antigens were obtained from a multitude of sources which include: Abcam (Cambridge, USA), AERAS (Rockville, USA), American Type Culture Collection (Manassas, USA), Biodefense and Emerging Infections Resources (BEI Resources) (Manassas, USA), Boston Children’s Hospital (Boston, USA), Brigham and Women’s Hospital Crimson Biorepository (Boston, USA), Hospital Joseph-Ducuing (Toulouse, France), Sigma-Aldrich (St. Louis, USA), Sino Biological (Beijing, China), and The Native Antigen Company (Oxford, United Kingdom).\n\nIn addition, the following defined strains were used in this study: Streptococcus pneumoniae 3 (ATCC 6303), Streptococcus pneumoniae 19A (ATCC 700674), Escherichia coli 41949 (Multiple O antigens:H26) (Crimson Biorepository), and Escherichia coli RS218 (NMEC O18:H7) (Kindly provided by James R. Johnson from the University of Minnesota). LPS from Serratia marcescens (L6136), Klebsiella pneumoniae (L4268), Salmonella enterica serovar enteritidis (L6011), and LTA from Enterococcus hirae (L4015), Staphylococcus aureus (L2515), and Streptococcus pyogenes (L3140) were purchased through Sigma-Aldrich (St. Louis, USA). Mycobacterium tuberculosis H37Rv components, which include lipoarabinomannan (LAM, NR-14848) and phosphatidylinositol mannoside 1,2 and 6 (PIM1,2, NR-14846 and PIM6, NR-14847), were obtained from BEI Resources (Manassas, USA). Parasites, Trichomonas vaginalis (TV01-1000) and Plasmodium falciparum Circumsporozoite protein (ab73857), were purchased from The Native Antigen Company (Oxford, UK) and Abcam (Cambridge, USA), respectively. Viral antigens: Chikungunya E1 (CHIKV-E1), Dengue serotype 1 VLP (DENV1-VLP), Ebola GP1 (EBOVKW95-GP1-100), Tick-borne Encephalitis NS1 (TBEV-NS1-100), and Zika lysate (ZIKV-LYS-100) were purchased from The Native Antigen Company (Oxford, UK). Cytomegalovirus (CMV) was kindly provided by Brigham & Women’s Hospital (Boston, USA). Influenza H1N1 HA (11055-VNAB), Influenza H1N1 NA (11058-VNAHC), and Respiratory syncytial virus (RSV) glycoprotein g (11070-V08B2) were purchased from Sino Biological (Beijing, China). Human immunodeficiency virus (HIV) gp120 (ab174070) was purchased from Abcam (Cambridge, USA).\n\nBacteria were subcultured in RPMI (Thermo Fisher Scientific, USA) 10 mM glucose to a McFarland of 0.5 (equivalent to 108 CFU/ml) (Becton Dickinson, USA). Bacteria were grown to this logarithmic phase to ensure cell viability, and RPMI was used because it does not contain interfering MBL-binding nutrients, such as yeast extract. The culture was then split—live bacteria were kept on ice while the other half were fragmented. Fragmented bacterial PAMPs were generated using antibiotics or mechanical disruption. Antibiotic treatment included the appropriate use of one of the following: cefepime (NDC 25021-121-20), ceftriaxone (NDC 60505-6104-4), meropenem (NDC 63323-507-20), amikacin (NDC 0703-9040-03), gentamicin (NDC 63323-010-02), or vancomycin (NDC 0409-4332-49), at ≤1 mg/ml for ≥4 hours at 37°C 225 rpm. Mechanical disruption consisted of bead mill treatment at 30 Hz for 10 min using 0.1 mm zirconia/silica beads (BioSpec Products, USA) in a Mixer Mill MM 400 machine (Verder Scientific, Inc., USA). Testing by FcMBL ELLecSA was performed on titers of live bacteria at ≤107 CFU/ml, and on the same concentration of bacteria after fragmentation. LPS endotoxin from Gram-negative bacteria was quantified using a limulus amebocyte lysate (LAL) assay ([Endosafe®] Charles River Laboratories, USA).\n\nFungi species were primarily subcultured in RPMI 10 mM glucose; however other media, such as potato dextrose broth (Teknova, USA), were used to facilitate growth. In these cases, the fungal cells were pelleted at 3,000 × g for 5 minutes at 22°C (Eppendorf 5424, USA), washed 3x in 50 mM Tris-HCl, 150 mM NaCl, 0.05% Tween-20, 5 mM CaCl2, pH 7.4 (TBST 5 mM CaCl2) (Boston BioProducts, USA) to remove residual growth media, and then resuspended in TBST 5 mM CaCl2. Testing by FcMBL ELLecSA was performed on titers of live fungi at ≤107 CFU/ml. Purified or inactivated viral antigens, parasites, and bacterial antigens were resuspended or diluted in TBST 5 mM CaCl2 for testing directly by FcMBL ELLecSA. Trichomonas vaginalis was screened at 0.46 µg/mL and Plasmodium falciparum Circumsporozoite protein was screened at 50 µg/mL. Concentrations of viral antigens tested are indicated in the legend of Figure 7.\n\nAnti-protein A (catalog number ab19483) was purchased from Abcam (Cambridge, USA).\n\nFresh whole human blood (sodium heparin) was purchased from Research Blood Components, LLC. (Boston, USA), and normal single donor human urine (IR100007) was purchased from Innovative Research Inc. (Novi, USA).\n\nThe key metric used to quantify direct FcMBL binding to pathogen-associated molecular patterns (PAMPs) from bacteria, fungi, viral antigens, parasites, and bacterial cell wall antigens is a 96 well ELLecSA, which has been previously published23. The assay uses FcMBL coated superparamagnetic beads (1 µm MyOne Dynabead [Thermo Fisher Scientific, USA]) where FcMBL, biotinylated at the N termini of the Fc protein using an N-terminal amino-oxy reaction, is coupled to streptavidin beads in an oriented array (Figure 1A). Each sample was screened using 100 µl or 200 µl of test sample added to 900 µl or 800 µl of assay solution respectively, which contains 5 µg of the FcMBL beads at 5 mg/ml and 10 mM glucose in TBST 5 mM CaCl2 to total 1 ml (50 µl heparin is added if testing blood). PAMPs in the test sample are captured by FcMBL for 20 minutes at 22°C 950 rpm in a plate shaker (Eppendorf, USA). Using an automated magnetic-handling system (KingFisherTM Flex [not shown]) (Thermo Fisher Scientific, USA), captured PAMPs are washed two times using TBST 5 mM CaCl2, and detected with human MBL (Sino Biological, China) linked to horseradish peroxidase (MBL-HRP). Non-specific MBL-HRP is removed by 4 washes in TBST 5 mM CaCl2, and PAMPs are quantified with 1-step ultra tetramethylbenzidine (TMB) substrate (Thermo Fisher Scientific, USA). Finally, the reaction is quenched with 1 M sulfuric acid and results are read at the optical density 450 nm wavelength. Quantification of bound PAMPs is determined using a standard curve generated using yeast mannan (catalog number M3640, Sigma-Aldrich, USA) in TBST 5 mM CaCl2—a known target for MBL (1 ng/ml mannan = 1 PAMP unit)43. PAMP units are multiplied back by the dilution factor of the test sample volume to give PAMPs/ml. Previously, a receiver operating characteristic comparison was performed for a small pilot sepsis patient study in which sepsis blood was analyzed versus non-infected controls to determine an optimal ELLecSA threshold of 0.45 PAMP units23. Therefore, in this study we define and report FcMBL binding to a sample as having ≥5 PAMPs/ml. To confirm specificity of FcMBL binding, a negative control (FcMBL null) was used alongside FcMBL in the ELLecSA. FcMBL null was engineered by introducing two residue mutations, E347A and N349A, into aktFcMBL (GenBank accession: KJ710775.1) to remove functional binding of the CRD of MBL. FcMBL null was purified and used to coat beads in the same fashion as FcMBL described above for direct comparison. FcMBL null beads did not support any binding to yeast mannan, and supported less than half of binding to S. aureus compared with FcMBL, as the Fc portion of FcMBL binds S. aureus protein A22 (Supplementary Figure 1)29.\n\nData analyses on FcMBL binding to Gram-positive and Gram-negative live and fragmented bacterial isolates was performed using the statistical R language. Categorical variables are described as frequency (percentage). Comparisons between nominal dichotomous variables were performed with Pearson’s chi-square, when all contingency table cells were >5. Results were deemed as statistically significant when the null hypothesis could be rejected with >95% confidence. An unpaired two-tailed t-test was performed on FcMBL and FcMBL null binding to S. aureus using GraphPad Prism 7.0b (OS X). A p-value of < 0.05 was determined to be statistically significant. Dataset analysis is indicated in the figure legends.\n\nFor visualization of live and fragmented bacteria on FcMBL beads, bacteria were captured with 128 nm FcMBL beads (Ademtech, France), spun down onto 13=-mm coverslips and fixed with 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (Electron Microscopy Sciences, USA) for 1 hour. Cover slips were incubated in 1% osmium tetroxide in 0.1 M sodium cacodylate (Electron Microscopy Sciences, USA) for 1 hour. Ascending grades of ethanol dehydrated the sample before being chemically dried with hexamethydisilazane (Electron Microscopy Sciences, USA). Samples were then placed in a desiccator overnight. Dried samples were mounted on aluminum stubs, sputter-coated with a thin layer of gold particles, and imaged using a Zeiss Supra55VP microscope.\n\n\nData availability\n\nData for this study on the broad-spectrum capture of clinical pathogens using engineered Fc-mannose-binding lectin (FcMBL) enhanced by antibiotic treatment are available from OSF. DOI: https://doi.org/10.17605/OSF.IO/GW4X729.\n\nSupplementary figure 1. FcMBL and FcMBL null binding to S. aureus. 106 CFU/ml live S. aureus without and with treatment of 1 µg/ml anti-protein A, detected by ELLecSA. Anti-protein A antibody blocks the Fc-mediated binding of FcMBL null but has no significant effect on FcMBL binding. **p-value < 0.01; *p-value < 0.05; unpaired two-tailed t-test; NS: not significant. DOI: https://doi.org/10.17605/OSF.IO/GW4X729.\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\nThis work was supported by the Defense Advanced Research Projects Agency (DARPA) grants N66001-11-1-4180 and W911NF-16-C-0050, a Global Health Innovation Partnership (GHIP) grant from the Bill and Melinda Gates Foundation, and the Wyss Institute for Biologically Inspired Engineering.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe thank Vasanth Chandrasekhar for assistance in making the FcMBL beads, Shanda Lightbown for SEM images of S. pneumoniae 3, and Seth Kroll for image processing.\n\n\nReferences\n\nTakahashi K: Mannose-binding lectin and the balance between immune protection and complication. Expert Rev Anti Infect Ther. 2011; 9(12): 1179–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKilpatrick DC: Mannan-binding lectin: clinical significance and applications. Biochim Biophys Acta. 2002; 1572(2–3): 401–13. PubMed Abstract | Publisher Full Text\n\nNuytinck L, Shapiro F: Mannose-binding lectin: laying the stepping stones from clinical research to personalized medicine. Per Med. 2004; 1(1): 35–52. PubMed Abstract | Publisher Full Text\n\nEisen DP, Minchinton RM: Impact of mannose-binding lectin on susceptibility to infectious diseases. Clin Infect Dis. 2003; 37(11): 1496–505. PubMed Abstract | Publisher Full Text\n\nLarvie M, Shoup T, Chang W, et al.: Mannose-binding lectin binds to amyloid β protein and modulates inflammation. J Biomed Biotechnol. 2012; 2012: 929803. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMogensen TH: Pathogen recognition and inflammatory signaling in innate immune defenses. Clin Microbiol Rev. 2009; 22(2): 240–73. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSaito T, Owen DM, Jiang F, et al.: Innate immunity induced by composition-dependent RIG-I recognition of hepatitis C virus RNA. Nature. 2008; 454(7203): 523–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKang HJ, Lee SM, Lee HH, et al.: Mannose-binding lectin without the aid of its associated serine proteases alters lipopolysaccharide-mediated cytokine/chemokine secretion from human endothelial cells. Immunology. 2007; 122(3): 335–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWallis R: Interactions between mannose-binding lectin and MASPs during complement activation by the lectin pathway. Immunobiology. 2007; 212(4–5): 289–99. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDegn SE, Thiel S: Humoral pattern recognition and the complement system. Scand J Immunol. 2013; 78(2): 181–93. PubMed Abstract | Publisher Full Text\n\nFraser IP, Koziel H, Ezekowitz RA: The serum mannose-binding protein and the macrophage mannose receptor are pattern recognition molecules that link innate and adaptive immunity. Semin Immunol. 1998; 10(5): 363–72. PubMed Abstract | Publisher Full Text\n\nDommett RM, Klein N, Turner MW: Mannose-binding lectin in innate immunity: past, present and future. Tissue Antigens. 2006; 68(3): 193–209. PubMed Abstract | Publisher Full Text\n\nPolotsky VY, Belisle JT, Mikusova K, et al.: Interaction of human mannose-binding protein with Mycobacterium avium. J Infect Dis. 1997; 175(5): 1159–68. PubMed Abstract | Publisher Full Text\n\nTownsend R, Read RC, Turner MW, et al.: Differential recognition of obligate anaerobic bacteria by human mannose-binding lectin. Clin Exp Immunol. 2001; 124(2): 223–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNeth O, Jack DL, Dodds AW, et al.: Mannose-binding lectin binds to a range of clinically relevant microorganisms and promotes complement deposition. Infect Immun. 2000; 68(2): 688–93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Emmerik LC, Kuijper EJ, Fijen CA, et al.: Binding of mannan-binding protein to various bacterial pathogens of meningitis. Clin Exp Immunol. 1994; 97(3): 411–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSaifuddin M, Hart ML, Gewurz H, et al.: Interaction of mannose-binding lectin with primary isolates of human immunodeficiency virus type 1. J Gen Virol. 2000; 81(Pt 4): 949–55. PubMed Abstract | Publisher Full Text\n\nKlabunde J, Berger J, Jensenius JC, et al.: Schistosoma mansoni: adhesion of mannan-binding lectin to surface glycoproteins of cercariae and adult worms. Exp Parasitol. 2000; 95(4): 231–9. PubMed Abstract | Publisher Full Text\n\nKuipers S, Aerts PC, van Dijk H: Differential microorganism-induced mannose-binding lectin activation. FEMS Immunol Med Microbiol. 2003; 36(1–2): 33–9. PubMed Abstract | Publisher Full Text\n\nKang JH, Super M, Yung CW, et al.: An extracorporeal blood-cleansing device for sepsis therapy. Nat Med. 2014; 20(10): 1211–6. PubMed Abstract | Publisher Full Text\n\nDidar TF, Cartwright MJ, Rottman M, et al.: Improved treatment of systemic blood infections using antibiotics with extracorporeal opsonin hemoadsorption. Biomaterials. 2015; 67: 382–92. PubMed Abstract | Publisher Full Text\n\nBicart-See A, Rottman M, Cartwright M, et al.: Rapid Isolation of Staphylococcus aureus Pathogens from Infected Clinical Samples Using Magnetic Beads Coated with Fc-Mannose Binding Lectin. PLoS One. 2016; 11(6): e0156287. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCartwright M, Rottman M, Shapiro NI, et al.: A Broad-Spectrum Infection Diagnostic that Detects Pathogen-Associated Molecular Patterns (PAMPs) in Whole Blood. EBioMedicine. 2016; 9: 217–227. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMagill SS, Edwards JR, Bamberg W, et al.: Multistate point-prevalence survey of health care–associated infections. N Engl J Med. 2014; 370(13): 1198–208. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRosenfeld Y, Shai Y: Lipopolysaccharide (Endotoxin)-host defense antibacterial peptides interactions: role in bacterial resistance and prevention of sepsis. Biochim Biophys Acta. 2006; 1758(9): 1513–22. PubMed Abstract | Publisher Full Text\n\nSharma P, Dube D, Singh A, et al.: Structural basis of recognition of pathogen-associated molecular patterns and inhibition of proinflammatory cytokines by camel peptidoglycan recognition protein. J Biol Chem. 2011; 286(18): 16208–17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPolotsky VY, Fischer W, Ezekowitz RA, et al.: Interactions of human mannose-binding protein with lipoteichoic acids. Infect Immun. 1996; 64(1): 380–3. PubMed Abstract | Free Full Text\n\nIp WK, Takahashi K, Moore KJ, et al.: Mannose-binding lectin enhances Toll-like receptors 2 and 6 signaling from the phagosome. J Exp Med. 2008; 205(1): 169–181. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIngber D: “Broad-Spectrum Capture of Clinical Pathogens Using Engineered Fc-Mannose-Binding Lectin (FcMBL) Enhanced by Antibiotic Treatment.” OSF. 2019. http://www.doi.org/10.17605/OSF.IO/GW4X7\n\nLynch NJ, Roscher S, Hartung T, et al.: L-ficolin specifically binds to lipoteichoic acid, a cell wall constituent of Gram-positive bacteria, and activates the lectin pathway of complement. J Immunol. 2004; 172(2): 1198–202. PubMed Abstract | Publisher Full Text\n\nVergne I, Gilleron M, Nigou J: Manipulation of the endocytic pathway and phagocyte functions by Mycobacterium tuberculosis lipoarabinomannan. Front Cell Infect Microbiol. 2015; 4: 187. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRojas RE, Thomas JJ, Gehring AJ, et al.: Phosphatidylinositol mannoside from Mycobacterium tuberculosis binds alpha5beta1 integrin (VLA-5) on CD4+ T cells and induces adhesion to fibronectin. J Immunol. 2006; 177(5): 2959–68. PubMed Abstract | Publisher Full Text\n\nIskandar A, Nursiloningrum E, Arthamin MZ, et al.: The Diagnostic Value of Urine Lipoarabinomannan (LAM) Antigen in Childhood Tuberculosis. J Clin Diagn Res. 2017; 11(3): EC32–EC35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSakamuri RM, Price DN, Lee M, et al.: Association of lipoarabinomannan with high density lipoprotein in blood: implications for diagnostics. Tuberculosis (Edinb). 2013; 93(3): 301–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLawn SD: Point-of-care detection of lipoarabinomannan (LAM) in urine for diagnosis of HIV-associated tuberculosis: a state of the art review. BMC Infect Dis. 2012; 12: 103. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoonyarattanakalin S, Liu X, Michieletti M, et al.: Chemical synthesis of all phosphatidylinositol mannoside (PIM) glycans from Mycobacterium tuberculosis. J Am Chem Soc. 2008; 130(49): 16791–9. PubMed Abstract | Publisher Full Text\n\nKallert S, Zenk SF, Walther P, et al.: Liposomal delivery of lipoarabinomannan triggers Mycobacterium tuberculosis specific T-cells. Tuberculosis (Edinb). 2015; 95(4): 452–62. PubMed Abstract | Publisher Full Text\n\nShin HJ, Franco LH, Nair VR, et al.: A baculovirus-conjugated mimotope vaccine targeting Mycobacterium tuberculosis lipoarabinomannan. PLoS One. 2017; 12(10): e0185945. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRodriguez ME, Loyd CM, Ding X, et al.: Mycobacterial phosphatidylinositol mannoside 6 (PIM6) up-regulates TCR-triggered HIV-1 replication in CD4+ T cells. PLoS One. 2013; 8(11): e80938. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCox MM, Izikson R, Post P, et al.: Safety, efficacy, and immunogenicity of Flublok in the prevention of seasonal influenza in adults. Ther Adv Vaccines. 2015; 3(4): 97–108. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchiller JT, Castellsagué X, Garland SM: A review of clinical trials of human papillomavirus prophylactic vaccines. Vaccine. 2012; 30 Suppl 5: F123–F138. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCunningham AL, Lal H, Kovac M, et al.: Efficacy of the Herpes Zoster Subunit Vaccine in Adults 70 Years of Age or Older. N Engl J Med. 2016; 375(11): 1019–1032. PubMed Abstract | Publisher Full Text\n\nSuper M, Thiel S, Lu J, et al.: Association of low levels of mannan-binding protein with a common defect of opsonisation. Lancet. 1989; 2(8674): 1236–9. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "43622",
"date": "28 Jan 2019",
"name": "Damon Eisen",
"expertise": [
"Reviewer Expertise Infectious Diseases physician Clinical scientist with extensive history of MBL biology 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\nThe authors, who are the inventors of the FcMBL capture assay, have undertaken rigorous experimental work to produce a clinically oriented paper that extends the range and numbers of microbes tested in prior publications. The current publication describes the detection of bacteria, fungi, parasites and viral antigens by the enzyme-linked lectin sorbent assay (ELLecSA). Microbiologically relevant quantities of these microbes were tested in vitro and high rates of detection by ELLecSA are reported. As before, fragmentation of bacteria using antibiotics and physical fragmentation increased the numbers of bacteria detected.\nThe mechanisms of this increased sensitivity of detection were explored by the quantification of FcMBL-binding to the major virulence determinants of gram-negative and gram-positive bacteria. Greater sensitivity of binding to LPS was shown compared with LTA. This result provides some explanation for FcMBL's failure to detect E. faecalis.\nThe current publication does not include any further investigation of patients with the sepsis syndrome relying on the author's prior work to imply that a broader range of bacterial causes of sepsis will be detected. This is a reasonable assumption.\n\nFurther work using the same in vitro techniques could incorporate assessment of FcMBL's detection of rickettsia, leptospires and treponemes. These bacteria all cause diseases that are predominantly diagnosed by serology. Expedited diagnosis of blood stream or central nervous system infection would be clinically beneficial and direct appropriate antimicrobial treatment in patients with appropriate epidemiological characteristics. This could be particularly relevant to use of FcMBL technology as part of biodefence.\n\nAs the authors emphasise, the use of FcMBL can only indicate the presence of a microbial pathogen in the bloodstream due to the non-specific characteristics of binding to PAMPs. A positive ELLecSA may provide clinicians confidence that a patient presenting with the sepsis-syndrome has an infectious cause. Further testing in such patients will give additional information on the validity of the test and may provide additional encouragement for its incorporation in a microbiological testing algorithm.\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": "47919",
"date": "29 May 2019",
"name": "Anna Swierzko",
"expertise": [
"Reviewer Expertise Immunology",
"Microbiology."
],
"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\nDr B. T. Seiler and co-workers present the next application of magnetic beads coated with genetically engineered form of the human MBL CRD linked to IgG Fc domain (FcMBL beads). Previously, the same group of researchers reported usefulness of that invention (in a form of biospleen or extracorporeal hemoadsorption device) in the efficient clearance of microorganisms’ cells as well as their PAMPs from blood. The presented manuscript is the valuable extension of work by Cartwright et al. (2016)1, where FcMBL-beads were used for the rapid detection of broad spectrum pathogens in the whole blood (ELLecSA). The authors demonstrated interaction of FcMBL with bacteria, fungi and parasites. The authors confirmed also their finding that antibiotic treatment or mechanical disruption increases FcMBL-binding. No case of decreased binding after antibiotic treatment or bead milling was observed, however some isolates, even after disruption were still not recognised by FcMBL. Among them there were Enterococcus faecalis and Proteus isolates. These finding are in agreement with previous reports by Geiss-Liebisch et al. (2012)2 and Man-Kupisinska et al. (2018)3. Geiss-Liebisch demonstrated that wall teichoic acids and rhamnopolysaccharide of E. faecalis prevent peptidoglycan from MBL binding. Whereas Man-Kupisinska, despite observed common interaction of MBL with the conservative inner core of numerous enterobacterial lipopolysaccharides, did not find MBL interaction with 42 of 70 tested Proteus vulgaris, Proteus mirabilis and Proteus penneri isolated LPSs. The authors should consider including this information into the manuscript.\nThe authors also presented ability of FcMBL to capture isolated bacterial and viral antigens. They also described detection of M.tuberculosis mannosylated LAM in spiked blood and urine. Surprisingly, the sensitivity in urine was markedly reduced, not only in comparison to buffer but also in comparison to blood. Previously the authors described that there are inhibitors (probably immunoglobulins) in joint fluids preventing FcMBL interaction with bacteria. Treatment with proteases and hyaluronidase was proposed to improve pathogen capture. Since rapid and sensitive method of LAM detection in urine could be of big importance in TB diagnosis, it would be important to identify these inhibitors.\nThis is a very important and well done study. FcMBL-coated beads are demonstrated as a sensitive alternative to the time consuming and often false negative blood culture. It can also be a fantastic tool, helpful in the screening and simple identification of MBL ligands (for example – after SDS-PAGE-WB and immunostaining with PAMP-specific antibodies).\n\nMinor points\nMannosylated LAM Mycobaterium tuberculosis H37Rv strain should be described as “ManLAM” Page 4: probably should be “including multiple isolates” instead of “including multiple species” On the page 4, referring to Table 1 - the authors stated that they tested FcMBL binding to 12 different species of fungi. However , in Table 1 only 3 isolates belonging to 3 species are listed and in Figure 4 – results of FcMBL interaction with 13 fungi isolates is presented. It is confusing. On page 3 the Authors stated that pathogens were treated with antibiotic within 4 hours, however according to the method section – “bacteria were treated with antibiotics for ≥4 hours”. It should be clarified.\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/8-108
|
https://f1000research.com/articles/8-98/v1
|
24 Jan 19
|
{
"type": "Research Article",
"title": "Microplastics in the surface sediments from the eastern waters of Java Sea, Indonesia",
"authors": [
"Defri Yona",
"Syarifah Hikmah Julinda Sari",
"Feni Iranawati",
"Syamsul Bachri",
"Wulan Cahya Ayuningtyas",
"Syarifah Hikmah Julinda Sari",
"Feni Iranawati",
"Syamsul Bachri",
"Wulan Cahya Ayuningtyas"
],
"abstract": "Background: This study aimed to investigate the abundance of microplastics in the eastern water of Java Sea. The study areas are well known for the high population and high industrial activities that can contribute to the plastic pollution. Methods: Microplastics were sampled from the surface sediments at five different stations representing different local activities: fish landing area (St 1), mangrove forest (St 2), abandoned shrimp pond (St 3), river mouth (St 4) and open sea (St 5). Results: Three types of microplastics were found; the most common was plastic fragments (54.34±6.39%) followed by fibers (41.45±4.59%) and films (4.21±3.90%). The highest abundance of microplastics was observed in the mangrove area (896.96±160.28 particles/kg), dominated with fragments and fibers. Films were found in greatest quantities in the fish landing area, but compared to the other types of microplastics, the abundance was much lower (80.73±37.62 particles/kg). Domestic wastes and fisheries activities were the main causes of the high microplastics in the study areas. Conclusions: The results of this study showed that microplastic pollution is a serious problem that needs to be paid attention not only from the government but also from the local people. Plastics management waste is needed.",
"keywords": [
"Anthropogenic activities",
"Java Sea",
"Microplastics",
"Surface Sediments"
],
"content": "Introduction\n\nThe presence of microplastics in the aquatic environment has become a global concern. Microplastics are small pieces of plastic less than 5 mm in size (Barnes et al., 2009; Hidalgo-Ruz et al., 2012). There are two types of microplastics based on their source, the primary microplastics from manufactured plastics in microscopic sizes such as scrubbers and pellets (Isobe, 2016), or secondary microplastics which derived from the breakdown of bigger plastic products such as fragments, fibers or films (Zhu et al., 2018; Zobkov & Esiukova, 2017).\n\nOnce in the aquatic environment, microplastics might float in the water column or sink to the bottom, depending on the particle density (Barnes et al., 2009; Kowalski et al., 2016). There have been many studies conducted to analyze the presence of microplastics in the water (Chae et al., 2015; Isobe et al., 2015; Iwasaki et al., 2017), sediments (Alomar et al., 2016; Imhof et al., 2017; Wang et al., 2017; Zobkov & Esiukova, 2017) or both (Frère et al., 2017; Zhu et al., 2018). The sinking behavior of microplastics to the bottom sediments might be the result of biofouling, which can increase its density, the size and shape of microplastics and also fluid density (Kowalski et al., 2016).\n\nSediments have been considered to be major sinks of microplastics. With the capability of microplastics to concentrate other organic pollutants or heavy metals and also their durability and resistance to degradation, the accumulation of microplastics in sediments can bring harm to marine and human life. Thus, this study was conducted to evaluate microplastics contamination in the eastern water of Java Sea, which is busy with many human activities. The types of microplastics and the abundances were used to investigate the influence of anthropogenic factors on the spatial distribution of microplastics in the study areas.\n\n\nMethods\n\nSampling was conducted in the eastern water of Java Sea, Gresik, Indonesia. There were five sampling stations that represented different local activities. Station 1 was located in the fish landing area that is busy with fisheries activity, especially in the morning. Station 2 was located in the mangrove area that is frequently inundated during high tide. The mangrove area is vulnerable to the plastic wastes from other places that are carried away by the current and tides. Station 3 was located in the abandoned shrimp pond and was once known for shrimp culture. However, this activity had stopped and many ponds have been abandoned. Some local people discard their waste in these ponds, including plastic waste. Station 4 was located in the river mouth of Bengawan Solo River that connected to the open sea. Bengawan Solo River is the longest river in the Java Island, passing through many cities in the Central and East Java Provinces and along the way could bring domestic wastes to its end point in the Java Sea. Station 5 was located in the open sea, about 1 km from the river mouth (Figure 1).\n\nSampling was conducted in March 2018 and samples were collected in the surface sediment using an Ekman Grab soft sediment sampler. In total, three replicates were obtained and about 500 g of sediment samples from each replicate of the sampling station were stored in sealed plastic bags. Sediment samples were taken to the laboratory for further analysis. For more information about the study areas regarding the present of the plastic wastes, we also held informal discussions with the local fishermen, especially those who have concern for the mangrove ecosystem. Since this was considered low-risk personal communication, ethical approval and consent were not sought.\n\nMicroplastics analysis was conducted by modifying the section 3.5 of the NOAA method (Masura et al., 2015). In the laboratory, 150 g of sediment samples were oven-dried for 24 hours at 90°C. Density separation was performed by adding 20 ml 0.05 M Fe(II) and 20 ml 30% H2O2. The samples were then homogenized on a hotplate with stirrer for 30 minutes at 60°C. To remove organic materials, an additional 20 ml of 30% H2O2 and NaCl were added and left overnight. After one night, the floating microplastics were collected by filtration using an 0.3-mm mesh filter. Visual identification of microplastics was conducted under a microscope with three distinct rules to separate the types of microplastics. A fragment is a particle that cannot be torn apart with tweezers, with sharp and broken edges of irregular shape and size of degradational plastic; a fiber is a particle that equally thick throughout the entire length and is not tapered at the end; and film that is very thin, part of the sheets of plastic bags and similar (Dai et al., 2018; Hidalgo-Ruz et al., 2012; Nor & Obbard, 2014; Zobkov & Esiukova, 2017). The results of this study were compared with the results from other studies to understand more of the sources of plastic pollution.\n\nNormality test was performed to determine the data distribution and to decide whether to use parametric or nonparametric tests for the statistical analysis. The abundance of microplastics proven to be distributed normally, thus, one-way ANOVA was used to compare the abundance of microplastics among the sampling stations (p < 0.01) and the post hoc Tukey`s test was run to confirm the differences of microplastics between sampling stations. Kruskal-Wallis H test was conducted to test the difference in type of microplastics found in the study areas in which the distributions were found to be nonparametric. All statistical tests were carried out using SPSS 16.0 for Windows.\n\n\nResults and discussion\n\nMicroplastics were detected in the surface sediments of the sampling stations in the eastern waters of Java Sea. Anthropogenic activities mostly contributed to the present of microplastics in the study areas. The areas have been known to be highly populated. This is also home to many industrial activities which discharge wastewater to the eastern water of the Java Sea. Microplastics appeared in the range of 206.04−896.96 particle/kg (Figure 2). The highest abundance of microplastics was found at Station 2 in the mangrove area (896.96±160.28 particles/kg), followed by Station 1 (772±336.75 particles/kg) and Station 5 (639.51±121.58 particles/kg). Stations 3 and 4 were observed to contain rather similar amounts of microplastics, which were three times lower than the abundance in the mangrove area (206.04±84.49 particles/kg and 215.54±64.58 particles/kg, respectively). According to one-way ANOVA, there was a statistically significant different in the abundance of microplastics among sampling stations (p < 0.01). Tukey’s post-hoc test revealed that the abundances of microplastics at Stations 1 and 2 were significantly different to the abundances at Stations 3 and 4 (p = 0.05). There was no statistically significant difference between the abundance of microplastics at Station 5 and the other stations. Raw data on microplastic abundance, along with all other raw data, are available on OSF (Yona, 2018).\n\nMicroplastics in the mangrove ecosystems have been studied in the Singapore`s coastal areas (Nor & Obbard, 2014), where there were much lower amounts compared to those found in this study. Even though the abundances were very much different between these studies, but the reasons of the occurrence of the microplastics in the mangrove ecosystems were rather similar. High occurrence of microplastics at Station 2 in the mangrove area was observed, which could be the result of root system of mangroves, which can trap many different type of rubbish including plastics. Moreover, tides that frequently inundated mangrove beds could bring more plastic wastes from the surrounding waters.\n\nDue to their small size and ability to float in the water column, microplastics can be transported for long distance by ocean currents (Iwasaki et al., 2017). Eastern water of Java Sea is the end point of a very big river, Bengawan Solo, which passes many cities in Java Island and carried plastic wastes along its way to the ocean. Therefore, microplastics found in this study might not just from the local sources, but also from faraway places. According to information from local fishermen, during west monsoon season (November-February), mangroves in the study area are filled with plastic waste from the Bengawan Solo River. Similar results were also found in the study in the Saigon River canal system crossing a megacity, Ho Chi Minh City, in which the source of the plastic pollution was from the land-based due to local habits and waste management (Lahens et al., 2018).\n\nOn the other hand, limited interaction of Station 3, which located in the shrimp pond, with the surrounding waters resulted in a low abundance of microplastics. Even so, the abundance of microplastics in the pond was not that low (206.04 particles/kg). This may be because the pond has been abandoned for quite some times and some people from local village may have discarded their rubbish inside the pond.\n\nThere were three type of microplastics found in the study areas: plastic fragments, plastic fibers and plastic films (Figure 3). Fragments and fibers dominated most of the sampling stations, while films occurred in very low number compared to the other two types of the microplastics. In total, half of the microplastics found were plastic fragments (54.34±6.39%), followed by plastic fibers (41.45±4.59%) and plastic films (4.21±3.90%). Kruskal-Wallis H test showed that there was a statistically difference in the type of microplastics (fragment, fiber and film) among the study sites (p < 0.01).\n\nFragments contributed the most to the composition of microplastics in the study areas. This type of microplastic is the result of fragmentation of large plastic pieces into smaller particles (Alomar et al., 2016), and mostly the contribution comes from the domestic waste. Fisheries activity in the study areas might contribute to the present of fiber, as mostly local fishermen use plastic fishing nets to catch their fish. A high abundance of fibers was also observed in the North Yellow Sea as the result of constant use of plastic fishing and nets and ropes as the main fishing tools (Zhu et al., 2018).\n\nThe highest abundance of plastic fragments appeared at Station 2 in the mangrove area (537.25 ± 160.28 particle/kg). The highest abundance of plastic fibers was also found at Station 2 compared to the other stations. On the other hand, plastic films were detected the highest at Station 1 in the fish landing area.\n\nCompared to the results from the other studies (Table 1), the levels of microplastics found in this study were similar to values obtained in the Western Mediterranean Sea (Alomar et al., 2016) and much higher than those in the North Yellow Sea (Zhu et al., 2018). The dominant type of microplastics found among the studies was also different. The study in the North Yellow Sea found that plastic films predominated; in the Bohai Sea the most common type was plastic fibers (Dai et al., 2018), while this study obtained the highest percentage of plastic fragments. The results revealed that microplastics found in this study are mostly from the degradation of the plastic wastes from human activities as stated by Barnes et al. (2009) that fragments are the result of the breakdown of a wide range of everyday plastic products. The lack of awareness from the citizen on how dangerous plastic materials are to the environment is the main reason for plastic pollution (Derraik, 2002). Efforts from the government and the community are therefore needed to combat plastic use and production.\n\n\nConclusion\n\nMicroplastics were found in all of the samples from the study area, with the highest levels found in the mangrove area. Fragments were the most common type of microplastic observed, followed by fibers and then small amount of films. The results showed that plastic contamination in the eastern waters of the Java Sea were mostly from anthropogenic activities, especially domestic waste. This plastic waste was not just from the local sources but also from the long-distance sources carried away by the Bengawan Solo River that end in the eastern water of Java Sea.\n\n\nData availability\n\nRaw data on the microplastics at each location are given on OSF. DOI: https://doi.org/10.17605/OSF.IO/H3ZDQ (Yona, 2018).\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\nThis research was conducted with funding from Fisheries and Marine Science Faculty, Brawijaya University (Dana PNBP FPIK Universitas Brawijaya tahun 2018).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe also would like to thanks all members of Marine Resources Exploration and Management Research Group, Brawijaya University and Pak Abdul Mughni and also the Banyuurip Head Village (Ikhsanul Haris) for the helping during the sampling process.\n\n\nReferences\n\nAlomar C, Estarellas F, Deudero S: Microplastics in the Mediterranean Sea: Deposition in coastal shallow sediments, spatial variation and preferential grain size. Mar Environ Res. 2016; 115: 1–10. PubMed Abstract | Publisher Full Text\n\nBarnes DK, Galgani F, Thompson RC, et al.: Accumulation and fragmentation of plastic debris in global environments. Philos Trans R Soc B Biol Sci. 2009; 364(1526): 1985–1998. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChae DH, Kim IS, Kim SK, et al.: Abundance and Distribution Characteristics of Microplastics in Surface Seawaters of the Incheon/Kyeonggi Coastal Region. Arch Environ Contam Toxicol. 2015; 69(3): 269–278. PubMed Abstract | Publisher Full Text\n\nDai Z, Zhang H, Zhou Q, et al.: Occurrence of microplastics in the water column and sediment in an inland sea affected by intensive anthropogenic activities. Environ Pollut. 2018; 242(Pt B): 1557–1565. PubMed Abstract | Publisher Full Text\n\nDerraik JG: The pollution of the marine environment by plastic debris: a review. Mar Pollut Bull. 2002; 44(9): 842–852. PubMed Abstract | Publisher Full Text\n\nFrère L, Paul-Pont I, Rinnert E, et al.: Influence of environmental and anthropogenic factors on the composition, concentration and spatial distribution of microplastics: A case study of the Bay of Brest (Brittany, France). Environ Pollut. 2017; 225: 211–222. PubMed Abstract | Publisher Full Text\n\nHidalgo-Ruz V, Gutow L, Thompson RC, et al.: Microplastics in the marine environment: a review of the methods used for identification and quantification. Environ Sci Technol. 2012; 46(6): 3060–3075. PubMed Abstract | Publisher Full Text\n\nImhof HK, Sigl R, Brauer E, et al.: Spatial and temporal variation of macro-, meso- and microplastic abundance on a remote coral island of the Maldives, Indian Ocean. Mar Pollut Bull. 2017; 116(1–2): 340–347. PubMed Abstract | Publisher Full Text\n\nIsobe A: Percentage of microbeads in pelagic microplastics within Japanese coastal waters. Mar Pollut Bull. 2016; 110(1): 432–437. PubMed Abstract | Publisher Full Text\n\nIsobe A, Uchida K, Tokai T, et al.: East Asian seas: A hot spot of pelagic microplastics. Mar Pollut Bull. 2015; 101(2): 618–623. PubMed Abstract | Publisher Full Text\n\nIwasaki S, Isobe A, Kako S, et al.: Fate of microplastics and mesoplastics carried by surface currents and wind waves: A numerical model approach in the Sea of Japan. Mar Pollut Bull. 2017; 121(1–2): 85–96. PubMed Abstract | Publisher Full Text\n\nKowalski N, Reichardt AM, Waniek JJ: Sinking rates of microplastics and potential implications of their alteration by physical, biological, and chemical factors. Mar Pollut Bull. 2016; 109(1): 310–319. PubMed Abstract | Publisher Full Text\n\nLahens L, Strady E, Kieu-Le TC, et al.: Macroplastic and microplastic contamination assessment of a tropical river (Saigon River, Vietnam) transversed by a developing megacity. Environ Pollut. 2018; 236: 661–671. PubMed Abstract | Publisher Full Text\n\nMasura J, Baker J, Foster G, et al.: Laboratory Methods for the Analysis of Microplastics in the Marine Environment: recommendations for quantifying synthetic particles in waters and sediments. NOAA Technical Memorandum NOS-OR&R-48. 2015. Reference Source\n\nNor NH, Obbard JP: Microplastics in Singapore's coastal mangrove ecosystems. Mar Pollut Bull. 2014; 79(1–2): 278–283. PubMed Abstract | Publisher Full Text\n\nWang J, Peng J, Tan Z, et al.: Microplastics in the surface sediments from the Beijiang River littoral zone: Composition, abundance, surface textures and interaction with heavy metals. Chemosphere. 2017; 171: 248–258. PubMed Abstract | Publisher Full Text\n\nYona D: MICROPLASTICS IN THE SURFACE SEDIMENTS FROM THE EASTERN WATER OF JAVA SEA, INDONESIA. 2018. http://www.doi.org/10.17605/OSF.IO/H3ZDQ\n\nZhao J, Ran W, Teng J, et al.: Microplastic pollution in sediments from the Bohai Sea and the Yellow Sea, China. Sci Total Environ. 2018; 640–641: 637–645. PubMed Abstract | Publisher Full Text\n\nZhu L, Bai H, Chen B, et al.: Microplastic pollution in North Yellow Sea, China: Observations on occurrence, distribution and identification. Sci Total Environ. 2018; 636: 20–29. PubMed Abstract | Publisher Full Text\n\nZobkov M, Esiukova E: Microplastics in Baltic bottom sediments: Quantification procedures and first results. Mar Pollut Bull. 2017; 114(2): 724–732. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "43569",
"date": "06 Feb 2019",
"name": "Atsuhiko Isobe",
"expertise": [
"Reviewer Expertise transport and fate of oceanic microplastics"
],
"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 paper on microplastic abundance in the bottom sediment east of the Java Sea. I have only trivial points to be improved.\n\nP. 3, \"Methods\": Please add \"Bengawan Solo River\" into Figure 1.\n\nP. 4, \"Microplastic analysis\": What was the maximal size of microplastics in your analysis? I recognized that the minimal sizes was 0.3 mm according to the mesh filter. The maximal size was however not specified.\n\nP.4, \"Microplastic analysis\": How did you avoid the contamination of fibers floating in the air in both the field and laboratory? We sometimes set a \"blank\" to monitor air-borne microfibers during the experiment. If you did not conduct the monitoring, please add the sentence: the abundance of microfibers detected in this study might be overestimated by the contamination due to air-born microfibers.\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": "4416",
"date": "11 Feb 2019",
"name": "Syarifah Sari",
"role": "Author Response",
"response": "We would like to say thank you to Prof. Atsuhiko Isobe for the comments and suggestions. Our responses to his concerns on making the improvements to our paper are presented below: P. 3, \"Methods\": Please add \"Bengawan Solo River\" into Figure 1. We will add the Bengawan Solo River into Figure 1 in our next revised version of the manuscript. P. 4, \"Microplastic analysis\": What was the maximal size of microplastics in your analysis? I recognized that the minimal sizes was 0.3 mm according to the mesh filter. The maximal size was however not specified The maximal size was 5 mm. We conducted filtration using stacked arrangement of sieves in the size of 5 mm and 0.3 mm to collect microplastic samples. P.4, \"Microplastic analysis\": How did you avoid the contamination of fibers floating in the air in both the field and laboratory? We sometimes set a \"blank\" to monitor air-borne microfibers during the experiment. If you did not conduct the monitoring, please add the sentence: the abundance of microfibers detected in this study might be overestimated by the contamination due to air-born microfibers. We did not conduct the monitoring of air-born microfibers, therefore, we will add your suggestion to the revised manuscript."
}
]
},
{
"id": "43573",
"date": "19 Mar 2019",
"name": "Alex Sen Gupta",
"expertise": [
"Reviewer Expertise Physical Oceanography",
"Climate Science"
],
"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\nMicroplastics in the surface sediments from the eastern waters of Java Sea, Indonesia Defri Yona, Syarifah Hikmah Julinda Sari, Feni Iranawati, Syamsul Bachri, Wulan Cahya Ayuningtyas\n\nThis paper uses coastal sediment samples from sites in the east Java Sea to examine abundance and composition of microplastics. The article is generally well written although I note below a few places where the language could be improved. The work really constitutes a report rather than a scientific article as while it provides useful baseline information, it does not provide any mechanistic advance, or specific policy advice. I would suggest this article is sound subject to minor revisions.\n\nAbstract:\n“that needs to be paid attention” > “that needs greater attention”.\n\n“Plastics management waste is needed.” > “Plastics waste management is needed.”\n\nPage 3:\nParagraph 1:\n“the primary microplastics from manufactured plastics”. All plastics are manufactured, do you mean “used in the manufacture of plastic products\"? Suggested edit: \"There are two types of microplastic: primary microplastics including pellets and scrubbers that are used in the manufacture of products and secondary microplastics which are...\".\n\nParagraph 2:\n“and also fluid density” > “and also by fluid density”.\n\nParagraph 3:\n“Sediments have been considered to be major sinks of microplastics.” Reference needed.\n\n“With the capability of microplastics to concentrate other organic pollutants or heavy metals and also their durability and resistance to degradation, the accumulation of microplastics in sediments can bring harm to marine and human life.” Reference needed or rephrase e.g. “could potentially cause harm...”.\n\n“Thus, this study was conducted to evaluate microplastics contamination in the eastern water of Java Sea, which is busy with many human…” > “Thus, this study was conducted to evaluate microplastics contamination in the eastern water of the Java Sea, which is subject to intensive human…”.\n\nParagraph 4:\n“that are carried away by the current and tides” > “that are carried by the currents and tides”.\n\n“shrimp pond and was once” > “shrimp pond that was once”.\n\n“this activity had stopped” > “this activity has stopped”.\n\n“in the Central and East Java Provinces and along the way could bring domestic wastes to its end point in the Java Sea.” > “in Central and East Java Provinces and along its path could accumulate domestic wastes that ultimately is transported to the Java Sea.”\n\nIt would be useful to give a brief introduction to the local oceanography, as this would help to understand the distribution of microplastics: what does the circulation look like? How important are tides and tidal currents?\n\nParagraph 5:\n“In total, three replicates were obtained and about 500 g of sediment samples from each replicate of the sampling station were stored” > “In total, three replicates were obtained at each site comprising of about 500 g of sediment per replicate. Samples were stored…”.\n\n“the present of the plastic wastes” > “the presence of plastic waste…”.\n\nFigure 1:\nI’ve looked at a number of other maps of this region (to try and orient myself) and the coastline looks quite different to the one shown here (e.g. see google map satellite). The inset map isn’t very helpful, the red square doesn’t have the correct size. It would be also be more helpful to zoom out to show east Java and Madura island. What is the yellow area in the inset map?\n\nPage 4:\nParagraph 4:\n“Anthropogenic activities mostly contributed to…”. But aren’t all plastics related to anthropogenic activity?\n\n“activities mostly contributed to the present of” > “activities mostly contributed to the presence of”.\n\n“The areas have been known to” > “The areas are to”.\n\n“206.04−896.96 particle/kg”. You don’t need so many significant figures. The nearest gramme would be more than adequate given the huge range. Can you explain the units? Is this no. of particles per kg of water? I know this has been commonly used but I’m not clear as to the relevance of using particles as a measure. Are 10 small particles considered the same as 10 large particles?\n\n“mangrove area (896.96Å}160.28 particles/kg)”. Above the upper limit for the range was 896.96. Here you have 896+/-160 so the total range is higher than stated above. Needs to be clearer what you mean by ‘range’.\n\n“three times lower”. 200x3=600. Its more than 4 times.\nFigure 2: what are the error bars? The total range? Presumably they should go in both directions. Makes more sense to show the actual values of the 3 replicates.\n\nPage 5:\nParagraph 1:\n“in the Singapore`s coastal areas” > “in Singapore`s coastal areas”.\n\n“these studies, but the reasons” > “these studies, the reasons”.\n\n“Even though the abundances were very much different between these studies, but the reasons of the occurrence of the microplastics in the mangrove ecosystems were rather similar.”\n\nHow do you know this? Either you have evidence for this or it is just a guess, in which case you need to use more careful language, e.g. “In both cases we suspect that the high abundance relates to large tidally driven transport of water through the mangroves and the capture of particles by the complex root systems.” NB you haven’t provided any information about the tidal circulation.\n\nParagraph 2:\n“…can be transported for long distance by ocean currents (Iwasaki et al., 2017). Eastern water of Java Sea is the end point of a very big river, Bengawan Solo…” I don’t follow the link. In the first sentence you are talking about ocean currents. In the second you are referring to estuarine flow.\n\n“and carried plastic wastes” > “and carries plastic wastes”.\n\n“might not just from” > “might not just come from”.\n\n“but also from faraway places” > “but may have been transported considerable distances”.\n\n“during west monsoon” > “during the west monsoon”.\n\n“Similar results were also found in the study in the Saigon River canal” > “The importance of estuarine contamination was also found in a study of the Saigon River canal”.\n\n“pollution was from the land-based” > “pollution was land-based”.\n\nParagraph 3:\n“which located” > “which is located”.\n\n“resulted in a low abundance of microplastics” > “resulted in a low abundance of microplastics relative to the other sites.”\n\n“and some people from local village may have discarded their rubbish inside the pond.” > “and dumping of waste from the local village into the pond may have occurred.”\n\nParagraph 4:\n“while films occurred in very low number compared to the other two types of the microplastics.” > “while films occurred in very low numbers in comparison.”\n\nParagraph 5:\n“Fragments contributed the most to the composition of microplastics in the study areas”. You just said this, no need to repeat.\n\n“and mostly the contribution comes from the domestic waste.” > “with the largest contribution coming from domestic waste.”\n\n“to the present of fiber, as mostly local fishermen use” > “to the presence of fiber, as local fishermen primarily use”.\n\n“fishing and nets” > “fishing nets”.\n\nParagraph 6:\n“were detected the highest at Station 1” > “were most common at Station 1”.\n\nPage 6:\nTable 1: What’s the difference between an item and a particle?\n\n“…North Yellow Sea … in the Bohai Sea…”. Describe where these are and the local setting.\n\nFibres seem to be overwhelming in Bohai - did the authors explain what the source was? Would be nice to mention here. Also what are the source of films in the north Yellow Sea? Would be interesting to offer some explanation as to why the sources are different in different areas.\n\nParagraph 1:\n“The lack of awareness from the citizen on how dangerous plastic materials are…”. So what are the dangers?\n\nParagraph 2:\n“…mostly from anthropogenic activities…” What plastic isn’t related to anthropogenic activity?\n\n“carried away by” > “carried by”.\n\nThere seems to be a lot of speculation. Why does station 4 have so much less plastic than station 5, if the primary source is from the estuary?\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": []
}
] | 1
|
https://f1000research.com/articles/8-98
|
https://f1000research.com/articles/8-94/v1
|
23 Jan 19
|
{
"type": "Research Article",
"title": "Triglycerides, independent of Ferriman Gallwey Score, is a main determinant of free testosterone index in PCOS",
"authors": [
"Andon Hestiantoro",
"Putri Deva Karimah",
"Amalia Shadrina",
"Budi Wiweko",
"R. Muharam",
"Brilliant Putri Kusuma Astuti",
"Putri Deva Karimah",
"Amalia Shadrina",
"Budi Wiweko",
"R. Muharam",
"Brilliant Putri Kusuma Astuti"
],
"abstract": "Background: Polycystic Ovarian Syndrome (PCOS) is the most common endocrinopathy in women of reproductive age, affecting 5-20% of women worldwide. Hyperandrogenism, as the primary characteristic of PCOS, is not always present in every patient. The hyperandrogenic phenotype of PCOS patients is influenced by both hormonal and metabolic dysfunctions. Therefore, this study aims to determine the correlation between hormone profile, lipid profile, and clinical profile with free testosterone index in subjects with PCOS. Methods: This prospective cross-sectional study was conducted in the Dr. Cipto Mangunkusumo General Hospital between July 2014 and December 2016. The study involved 76 women with PCOS, who were classified into 2 subgroups: 39 subjects in the hyperandrogenism group and 37 subjects in the non-hyperandrogenism group. Each subject underwent physical examination, blood sample collection, and USG examination. Bivariate analysis was done using independent t-tests and Mann Whitney U-tests, while multivariate analysis was done using logistic regression. Results: Triglyceride and testosterone level showed weak (r = 0.232, p = 0.044) and moderate (r = 0.460, p ¡ 0.001) positive correlation with FTI, while SHBG level showed moderate negative correlation (r = -0.483, p ¡ 0.001). Triglyceride was also found to be determinant of hyperandrogenism condition in PCOS patient (OR 0.02, 95% CI 0.00–0.04, p = 0.013). However, there was no significant difference observed between FGS and hyperandrogenism (p = 0.43). Conclusions: Triglycerides, testosterone, and SHBG were associated with hyperandrogenism in PCOS patients, while FGS showed no such association.",
"keywords": [
"Polycystic Ovarian Syndrome",
"Hyperandrogenism",
"Free Testosterone Index",
"Hirsutism",
"Ferriman Gallwey Score",
"Triglyceride"
],
"content": "Introduction\n\nPolycystic Ovarian Syndrome (PCOS) is the most common endocrinopathy in women of reproductive age, affecting 5–20% of women, depending on the diagnosis criteria used1. PCOS is a complex disorder with a wide range of physical manifestation which is primarily characterized by hyperandrogenism or hyperandrogenemia, chronic anovulatory cycle, and polycycstic ovarian morphology. PCOS patients are considered to be at increased risk of developing several co-morbidities, such as diabetes, dyslipidemia, dysmetabolic syndrome, hypertension, obesity, obstructive sleep apnea, mental health disorders, and increased risk of cardiovascular diseases2–4.\n\nHyperandrogenism is one of the hallmark pathophysiological features of PCOS. Along with insulin resistance, hyperandrogenism causes metabolic derangements and some cutaneous symptoms, such as hirsutism, acne, and androgenic alopecia. The extent of hyperandrogenic clinical presentation varies among individuals and is affected by several factors, such as genetic polymorphism, inappropriate epigenetic reprogramming, metabolic factors, and other environmental factors3,5. An intricate interrelationship between environmental factors and aberrant micro-RNA expression has been proposed as the epigenetic mechanism underlying the development of PCOS5. It can be implicitly recognized that the phenotypic heterogeneity may illustrate the differences in their underlying genetic and metabolic pathophysiology, including the abnormalities of insulin regulation and lipid metabolism.\n\nBesides its prominent aspect in the pathophysiology of PCOS, hyperandrogenism also possesses a fundamental role in achieving a PCOS diagnosis. Hyperandrogenism, as one of the three diagnostic criteria of PCOS, can be identified with physical examination and laboratory evaluation. Hirsutism is the most frequently found clinical manifestation of hyperandrogenism, accounting for approximately 70–80% of all PCOS cases3,6,7. The modified Ferriman-Gallwey Score (mFGS) is used to measure the degree of terminal hair growth in several body sites and total score of ≥ 8 is considered hirsutism8. Biochemical assessment of hyperandrogenism involves the evaluation of several parameters, such as free testosterone index (FTI), free androgen index (FAI) or dehydroepiandosterone sulfate. Elevated levels of one of those markers indicate the presence of hyperandrogenemia9. However, elevated androgen levels are not always accompanied by obvious peripheral manifestations. Many studies found that only half of women with hirsutism have elevated levels of androgen hormones and only one-third of women with elevated androgen hormones have hirsutism7. Hirsutism is a multifactorial condition and androgen only plays a partial role in its occurrence. These findings suggest that hirsutism might not be the most suitable marker for identifying the elevation of androgen levels in PCOS patients.\n\nIt is important to evaluate further which factors correlate with testosterone levels reflected with free testosterone index in subject PCOS. We hypothesized that the hyperandrogenic phenotype of PCOS patients is influenced by both hormonal and metabolic dysfunctions. Therefore, this study aims to determine the association between hormone profile, lipid profile, and clinical profile with free testosterone index in subjects with PCOS.\n\n\nMethods\n\nThis was a cross-sectional study conducted at Dr. Cipto Mangunkusumo General Hospital, Jakarta from July 2014 until December 2016. Sample size was determined using Lemeshow sample size formula as presented below:\n\nN = (Z α2 x p(1-p)/d2\n\nN represents the minimum sample size. Z α represents standard normal deviation corresponding to 100% α (1.96) p represents the proportion of PCOS patients in Cipto Mangunkusumo General Hospital (45.7 %) d represents precision (12 %) According to that formulation, the minimum sample size of this study was 66.2 subjects. To account for potential post-enrollment drop out, additional 10% subjects were assigned to the study, resulting in a minimum sample size of 72 subjects. Subjects were recruited in-person consecutively from patients who came to the gynecology clinic with the chief complaint of irregular menstrual cycle. The recruitment of subjects was terminated once the minimum sample size has been achieved. Data and sample collection were performed immediately after patients agreed to participate in this study and were conducted during patient’s visit to gynecology clinic. All subjects were recruited based on the inclusion criteria, which included women between the age of 18 and 40 years old, who had been diagnosed with PCOS based on Rotterdam consensus criteria, have not consumed any PCOS or hormonal medication in the past 3 months, and were willing to participate in this study. Women who were pregnant or currently breastfeeding; those with history of uterine and other adnexal abnormalities, disorder of adrenal gland function, primary hypothalamic – hypophyseal disorder, ovarian tumor, disorder of prolactin secretion, unexplained abnormal uterine bleeding, and with previous history of thromboembolic or cerebrovascular disorders were excluded from this study.\n\nThe study protocol was approved by Ethics Committee of Faculty of Medicine Universitas Indonesia and Dr. Cipto Mangunkusumo Hospital, with reference number of 818/UN2.F1/ETIK/X/2016. Prior to the beginning of this study, subjects were informed about the protocol of the study and were asked to sign written consent form.\n\nThose who met the aforementioned criteria then underwent medical history taking and physical examination, including waist circumference, body weight, and body height measurements; hirsutism index through FGS10; and gynecologic examination.\n\nFGS is a visual instrument which is widely used to evaluate the excess growth of terminal hair in several body areas. There are 9 androgen sensitive areas which are assessed in FGS, and each area is assigned with value from 1 to 4 according to the thickness of the hair growth. These areas are lip, chin, chest, upper abdomen, lower abdomen, upper arm, thigh, upper back, and lower back. The cut off point at which hirsutism diagnosis is made varies depending on race and ethnicity. However, in general a total score equal to or more than 8 signifies hirsutism.\n\nUltrasonography was done to assess polycystic features. Blood samples were collected during the initial study for fasting plasma glucose, 2-hour postprandial plasma glucose, insulin, Homeostatic Model Assessment Insulin Resistance (Homa-IR), prolactin, LDL, HDL, Triglyceride, SHBG, TSH, LH, FSH, FTI, and testosterone examination. Subjects were instructed to avoid eating or drinking anything for 9–12 hours before the blood sample is collected. Approximately 10 mL of venous blood samples were drawn from each subject and were collected in several Vacutainers® blood collection tubes. After initial blood collection, subjects were instructed to have a full-course meal or a meal with at least 75 g of carbohydrates. Two hours following the meal, venous blood samples were drawn again to evaluate the 2 hour postprandial plasma glucose. All specimens were stored in a −70ª freezer before being transported to the lab for further analysis. Biochemical analysis was conducted at Prodia clinical laboratory. Fasting and 2-hour postprandial plasma glucose were determined using ARCHITECT Glucose Reagent Kit (Abbott Diagnostics, Illinois, USA). Plasma insulin was determined using ARCHITECT Insulin Reagent Kit (Abbott Diagnostics, Illinois, USA). HOMA-IR was calculated according to this formula = fasting insulin (in µIU/ml) X fasting plasma glucose (in mg/dL). Prolactin level was determined using ADVIA Centaur Prolactin Kit (Siemens Healthineers Global, New York, USA). LDL-C level was determined using Sekisui Cholesterol LDL Kit (Siemens Healthineers Global, New York, USA). HDL-C level was determined using Sekisui Cholesterol HDL Kit (Siemens Healthineers Global, New York, USA). Triglyceride level was determined using ADVIA Chemistry Triglyceride Reagent Kit (Siemens Healthineers Global, New York, USA). SHBG level was determined using ADVIA Centaur Immunoassay Kit (Siemens Healthineers Global, New York, USA). TSH level was determined using ADVIA Chemistry TSHs Reagent Kit (Siemens Healthineers Global, New York, USA). LH level was determined using ADVIA Centaur LH Kit (Siemens Healthineers Global, New York, USA). FSH level was determined using ADVIA Centaur FSH Kit (Siemens Healthineers Global, New York, USA). Testosterone level was determined using Testosterone II Kit (Roche Diagnostics, Risch-Rotkreuz, Switzerland). Free Testosterone Index (FTI) was calculated according to this formula = Total Testosterone (in nmol/L) /SHBG (in nmol/L) x 100.\n\nThe subjects were classified into two groups according to the results of their FTI tests: hyperandrogenism and non-hyperandrogenism groups. The diagnosis of hyperandrogenism was made according to subject’s FTI level. Subjects with FTI measurement equal to or greater than 5 were classified into hyperandrogenism group, while subjects with FTI measurement less than 5 were classified into non-hyperandrogenism group. Comparative analysis was performed between these two groups based on the variables mentioned above. The primary study outcome of this study was to determine the correlation between FGS and FTI, and also factors that contribute to hyperandrogenism phenotype in PCOS patients.\n\nData obtained from the subjects were recorded in case registration forms and were analyzed using Statistical Package for the Social Sciences (SPSS) version 20. Univariate analysis was done by converting valid data into tables containing mean and median values, as well as their distribution. Bivariate analysis was conducted using independent T-test and Mann-Whitney U-test to compare variables, such as age, body mass index (BMI), body weight, body height, waist circumference, FGS, fasting blood glucose, 2-hour postprandial plasma glucose, insulin, Homa-IR, prolactin, LDL, HDL, Triglyceride, SHBG, TSH, LH, FSH, and testosterone, between hyperandrogenism and non-hyperandrogenism groups. P < 0.05 was considered significant. Among those variables, multivariate analysis using logistic regression was performed on variables with p < 0.25. Finally, correlation analysis was performed using Pearson’s and/or Spearman’s test according to the normality of data distribution. Correlation analysis was performed to quantify the association between the biochemical parameters and free testosterone index in PCOS patients.\n\n\nResults\n\nThe 76 subjects participating in this study were classified into two groups: 37 in the PCOS without hyperandrogenism group and 39 in the PCOS with hyperandrogenism group. Most of the subjects were in their mid-20s, overweight (according to Asia Pacific WHO classification), and had central obesity. The subjects’ body weights ranged from 50.5 to 101 kg and the FGS ranged from 1 to 11. Most of the subjects showed good blood glucose profiles: 90.8% subjects had normal fasting blood glucose and 53% subjects had normal 2-hour post-prandial blood glucose. Approximately 94.7% and 82.9% subjects had abnormal LDL and HDL levels, respectively, but most (76.3%) had normal triglyceride levels. The median of insulin level was 13.6 µIU/ml, Homa-IR was 2.89, prolactin level was 9.4 ng/ml, SHBG level was 21.48 nmol/l, TSH level was 1.62 µIU/ml, LH level was 10.7 µIU/ml, FSH level was 6.3 µIU/ml, and testosterone level was 37.55 ng/dl. Details of subject characteristics are shown in Table 1.\n\nWithin bivariate analysis, significant differences between hyperandrogenism and non-hyperandrogenism group were observed in some characteristics, such as triglyceride level (p = 0.01), SHBG level (p = 0.01), and testosterone level (p = 0.04). The hyperandrogenism group had significantly higher level of triglyceride and testosterone, but lower SHBG level. On the other hand, no significant difference was observed between hirsute appearance, which was measured with FGS, and FAI (p = 0.43) (see Table 2).\n\nSpearman analysis was conducted to evaluate the correlation between these factors and FTI. From the analysis, it was found that triglyceride and testosterone level showed weak (r = 0.232) and moderate (r = 0.460) positive correlations with FTI, while SHBG level showed a moderate negative correlation (r = -0.483) (see Table 3).\n\nMultivariate analysis was conducted using logistic regression, including four variables such as triglyceride level, insulin level, Homa-IR level, and LDL level. Among these variables, triglyceride was found to be an important determinant of hyperandrogenism condition in PCOS patients (Table 4).\n\n\nDiscussion\n\nThis study was conducted to determine factors that influence the hyperandrogenism phenotype in PCOS patients, more specifically the concordance between its clinical features and biochemical parameters. Even though hyperandrogenism is one of the pivotal features of PCOS, not every patient with PCOS exhibits such hyperandrogenic phenotype. Pathophysiologically, hyperandrogenism is associated with intense ovarian steroidogenesis due to thecal cell hyperplasia. There are marked increases in GnRH and LH secretion, along with relative deficit in FSH secretion, which resulted in aberrant follicle growth and development, as well as reduced conversion of androstenedione and dehydroepiandrosterone to estrogen3,11. Insulin resistance also plays a significant role in the development of hyperandrogenism, by increasing the secretion pulse of LH and suppressing the production of SHBG in the liver, thus increasing the level of testosterone3.\n\nThis study found that there were three biochemical parameters that significantly differed between the two groups and correlated with hyperandrogenism phenotype, which were elevated testosterone and triglyceride levels, as well as decreased SHBG level. Testosterone and triglyceride levels had positive correlations with FTI, while SHBG showed negative correlation with FTI. There is an inverse relationship between SHBG and free testosterone level. SHBG is a glycoprotein that binds and transports sex steroids, such as testosterone and estradiol in the plasma. SHBG concentration is strongly influenced by various factors, such as sex steroid balance, drugs, thyroid hormone, insulin, dietary composition, and liver diseases. Lower level of SHBG means that less testosterone is bound, which results in a higher free testosterone concentration detected in blood plasma12,13.\n\nWe also found triglyceride levels to be a determinant factor of hyperandrogenism, with a weak positive correlation with FTI. Dyslipidemia is one of the most commonly found metabolic disturbances in PCOS, occurring in approximately 70% of PCOS patients. The pathogenic mechanisms underlying this condition are complex and are not yet fully understood. Previous studies found that obesity, hyperandrogenism, and hyperinsulinemic insulin resistance contributed to the development of hypertriglyceridemia in PCOS. Hyperandrogenism and hyperinsulinemia cause defective cathecolamine-induced lipolysis which eventually leads to the increased release of free fatty acids. Free fatty acids then stimulate hepatic overproduction of VLDL with more triglyceride compounds on each VLDL particles. Hyperandrogenism is also believed to have crucial roles in the upregulation of several genes which involved in the catabolism of lipoproteins, such as scavenger receptor B1 (SR-B1) and hepatic lipase (HL). Therefore, it is foreseeable that hyperandrogenic patients will have significantly higher level of triglyceridemia, compared to those with normal androgen concentration14,15.\n\nEven though not explicitly stated in this study, an independent relationship was observed between triglyceride level and insulin resistance. A moderate positive correlation was observed between HOMA-IR and triglyceride level in patients with PCOS (p < 0.001, r = 0.445). Triglyceride levels are considered a useful marker in identifying insulin resistance, particularly in patients with metabolic syndrome16. As stated above, insulin resistance, along with its compensatory hyperinsulinemia, contributed to triglyceride dysregulation in hyperandrogenic patients with PCOS. Hyperinsulinemia inhibits microsomal triglyceride protein expression which is crucial in the regulation of apolipoprotein B-100 and VLDL production. It also suppresses the removal of triglyceride-rich protein. Insulin resistant PCOS patients are more prone to dysregulation of lipid metabolism compared to those with normal insulin sensitivity (81% vs. 65%, respectively)14,15. A recently published study revealed a two-way relationship between androgen excess and insulin resistance. FAI as the indicator of hyperandrogenism can serve as an indicator of glucose tolerance, as an increase in FAI is usually followed by increases in blood glucose concentration, insulin level, and glucose resistance17.\n\nA novel concept, dysbiosis of gut microbiota (DOGMA), has been found to have considerable impact on the pathogenesis of PCOS, particularly through the development of insulin resistance and hypertriglyceridemia. A high-fat/high-sugar diet and obesity are the primary causes of DOGMA, driving increases in the growth of pathogenic microorganisms and suppress the growth good bacteria, which further leads to metabolic endotoxemia (the leakage of lipopolysaccharides produced by Gram-negative bacteria to systemic circulation) and chronic low-grade inflammatory conditions in the gut. Chronic low-grade inflammation interferes with islet β-cell proliferation and insulin receptor function, thus resulting in insulin resistance and compensatory hyperinsulinemia. In addition to that, DOGMA also plays a role in the development of dyslipidemia by modulating hepatic and systemic metabolism of lipid and glucose via the elicitation of short-chain fatty acids18–22.\n\nAside from DOGMA, vitamin D deficiency has also been implicated in insulin resistance and dyslipidemic condition commonly found in patients with PCOS. Many previous studies have indicated that PCOS patients with vitamin D deficiency tend to have higher levels of triglyceride and Homa-IR, compared to those with sufficient level of vitamin D concentration. Physiologically speaking, the vitamin D–vitamin D receptor (VDR) complex enacts an important role in regulating several genes, including those involved in glucose and lipid metabolism. Therefore, it is likely that interference in vitamin D concentration would also disrupt the metabolism of glucose and lipid23–25. One interesting finding to be noted in this study is the fact that no statistically significant difference was found between FGS and FTI. This finding implicates that the symptoms of hyperandrogenism in our PCOS subjects could not be assessed using FGS. The reason underlying this finding is the fact that clinical signs of hyperandrogenism are not particularly noticeable in PCOS patients in Asian countries, including Indonesia. Hirsutism appearance on each individual depends on their sensitivity to circulating androgens and its variation is influenced by ethnicity. Asian women tend to be less hirsute than Caucasian women, despite the elevated levels of androgen26. A number of controversies regarding the extent of testosterone level and FTI in predicting the severity of hirsutism have prevailed upon earlier studies6,27. Pathophysiologically, androgens play important role in the growth of terminal hair on several predilection areas which are normally hairy for men, but not for women7. However, prior clinical study discovered that only 68% hirsute PCOS patients were hyperandrogenic and only 63% hyperandrogenic PCOS patients were diagnosed with hirsutim7,28. This implicates that there are other factors that might contribute to hyperandrogenism other than hirsutism, vice versa.\n\n\nConclusions\n\nIn conclusion, a high FTI in PCOS patients is associated with high triglyceride levels, high testosterone levels, and low SHBG levels. Ferriman Gallwey score, as an indicator of hirsutism, shows no significant association with FTI. These associations mean that hyperandrogenic phenotype of PCOS patients is influenced by both hormonal and metabolic dysfunctions.\n\n\nData availability\n\nDataset 1. Anthropometric, hirsutism and blood sample data obtained from this study. DOI: https://doi.org/10.5256/f1000research.16815.d23173329.",
"appendix": "Grant information\n\nThis research received no specific grant from any funding agency in the public, commercial, and non-profit sectors.\n\n\nAcknowledgements\n\nThe authors would like to thank Yasmin Clinic and patients for their invaluable contributions to this research.\n\n\nReferences\n\nSirmans SM, Pate KA: Epidemiology, diagnosis, and management of polycystic ovary syndrome. Clin Epidemiol. 2013; 6: 1–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDe Leo V, Musacchio MC, Cappelli V, et al.: Genetic, hormonal and metabolic aspects of PCOS: an update. Reprod Biol Endocrinol. 2016; 14(1): 38. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDadachanji R, Shaikh N, Mukherjee S: Genetic Variants Associated with Hyperandrogenemia in PCOS Pathophysiology. Genet Res Int. 2018; 2018: 7624932. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatel S: Polycystic ovary syndrome (PCOS), an inflammatory, systemic, lifestyle endocrinopathy. J Steroid Biochem Mol Biol. 2018; 182: 27–36. PubMed Abstract | Publisher Full Text\n\nIlie IR, Georgescu CE: Polycystic Ovary Syndrome-Epigenetic Mechanisms and Aberrant MicroRNA. In Adv Clin Chem. Elsevier; 2015; 71: 25–45. PubMed Abstract | Publisher Full Text\n\nLeerasiri P, Wongwananuruk T, Indhavivadhana S, et al.: Correlation of clinical and biochemical hyperandrogenism in Thai women with polycystic ovary syndrome. J Obstet Gynaecol Res. 2016; 42(6): 678–683. PubMed Abstract | Publisher Full Text\n\nLanday M, Huang A, Azziz R: Degree of hyperinsulinemia, independent of androgen levels, is an important determinant of the severity of hirsutism in PCOS. Fertil Steril. 2009; 92(2): 643–647. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAswini R, Jayapalan S: Modified Ferriman-Gallwey Score in Hirsutism and its Association with Metabolic Syndrome. Int J Trichology. 2017; 9(1): 7–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChanukvadze D, Kristesashvili J, Kvashilava N: Correlation of biochemical markers and clinical signs of hyperandrogenism in women with polycystic ovary syndrome (PCOS) and women with non-classic congenital adrenal hyperplasia (NCAH). Iran J Reprod Med. 2012; 10(4): 307–14. PubMed Abstract | Free Full Text\n\nHohl A, Ronsoni MF, Oliveira Md: Hirsutism: diagnosis and treatment. Arq Bras Endocrinol Metabol. 2014; 58(2): 97–107. PubMed Abstract | Publisher Full Text\n\nJeffrey Chang R, Cook-Andersen H: Disordered follicle development. Mol Cell Endocrinol. 2013; 373(1–2): 51–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJayagopal V, Kilpatrick ES, Jennings PE, et al.: The biological variation of testosterone and sex hormone-binding globulin (SHBG) in polycystic ovarian syndrome: implications for SHBG as a surrogate marker of insulin resistance. J Clin Endocrinol Metab. 2003; 88(4): 1528–1533. PubMed Abstract | Publisher Full Text\n\nAl Kindi MK, Al Essry FS, Al Essry FS, et al.: Validity of serum testosterone, free androgen index, and calculated free testosterone in women with suspected hyperandrogenism. Oman Med J. 2012; 27(6): 471–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDiamanti-Kandarakis E, Papavassiliou AG, Kandarakis SA, et al.: Pathophysiology and types of dyslipidemia in PCOS. Trends Endocrinol Metab. 2007; 18(7): 280–285. PubMed Abstract | Publisher Full Text\n\nAlexander CJ, Tangchitnob EP, Lepor NE: Polycystic ovary syndrome: a major unrecognized cardiovascular risk factor in women. Rev Obstet Gynecol. 2009; 2(4): 232–9. PubMed Abstract | Free Full Text\n\nPark SY, Cho YJ, Lee SR, et al.: Triglyceride is a useful surrogate marker for insulin resistance in Korean women with polycystic ovary syndrome. Yonsei Med J. 2015; 56(3): 785–792. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang B, Wang J, Shen S, et al.: Association of Androgen Excess with Glucose Intolerance in Women with Polycystic Ovary Syndrome. Biomed Res Int. 2018; 2018: 6869705. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTremellen K, Pearce K: Dysbiosis of Gut Microbiota (DOGMA)--a novel theory for the development of Polycystic Ovarian Syndrome. Med Hypotheses. 2012; 79(1): 104–112. PubMed Abstract | Publisher Full Text\n\nGhazalpour A, Cespedes I, Bennett BJ, et al.: Expanding role of gut microbiota in lipid metabolism. Curr Opin Lipidol. 2016; 27(2): 141–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu R, Zhang C, Shi Y, et al.: Dysbiosis of Gut Microbiota Associated with Clinical Parameters in Polycystic Ovary Syndrome. Front Microbiol. 2017; 8: 324. PubMed Abstract | Publisher Full Text | Free Full Text\n\nInsenser M, Murri M, Del Campo R, et al.: Gut Microbiota and the Polycystic Ovary Syndrome: Influence of Sex, Sex Hormones, and Obesity. J Clin Endocrinol Metab. 2018; 103(7): 2552–2562. PubMed Abstract | Publisher Full Text\n\nGuo Y, Qi Y, Yang X, et al.: Association between Polycystic Ovary Syndrome and Gut Microbiota. PLoS One. 2016; 11(4): e0153196. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarg R, Malhotra J, Singh S, et al.: Relationship between vitamin d and insulin resistance in polycystic ovary syndrome women. Journal of South Asian Federation of Obstetrics and Gynaecology. 2017; 9(3): 211–215. Publisher Full Text\n\nKrul-Poel YH, Snackey C, Louwers Y, et al.: The role of vitamin D in metabolic disturbances in polycystic ovary syndrome: a systematic review. Eur J Endocrinol. 2013; 169(6): 853–865. PubMed Abstract | Publisher Full Text\n\nRodríguez-Rodríguez E, Ortega RM, González-Rodríguez LG, et al.: Vitamin D deficiency is an independent predictor of elevated triglycerides in Spanish school children. Eur J Nutr. 2011; 50(5): 373–378. PubMed Abstract | Publisher Full Text\n\nWang ET, Kao CN, Shinkai K, et al.: Phenotypic comparison of Caucasian and Asian women with polycystic ovary syndrome: a cross-sectional study. Fertil Steril. 2013; 100(1): 214–218. PubMed Abstract | Publisher Full Text\n\nAmiri M, Ramezani Tehrani F, Nahidi F, et al.: Association between biochemical hyperandrogenism parameters and Ferriman-Gallwey score in patients with polycystic ovary syndrome: A systematic review and meta-regression analysis. Clin Endocrinol (Oxf). 2017; 87(3): 217–230. PubMed Abstract | Publisher Full Text\n\nChang WY, Knochenhauer ES, Bartolucci AA, et al.: Phenotypic spectrum of polycystic ovary syndrome: clinical and biochemical characterization of the three major clinical subgroups. Fertil Steril. 2005; 83(6): 1717–1723. PubMed Abstract | Publisher Full Text\n\nHestiantoro A, Karimah PD, Shadrina A, et al.: Dataset 1 in: Triglycerides, independent of Ferriman Gallwey Score, is a main determinant of free testosterone index in PCOS. F1000Research. 2019. http://www.doi.org/10.5256/f1000research.16815.d231733"
}
|
[
{
"id": "43535",
"date": "05 Feb 2019",
"name": "Xue-Lian Li",
"expertise": [
"Reviewer Expertise Female reproductive endocrine-related diseases",
"especially PCOS."
],
"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\nPolycystic Ovary Syndrome (PCOS) is characterized by a series of endocrine and metabolism disturbances, such as insulin resistance, hyperandrogenism, sympathetic dysfunction and chronic low-grade inflammation state1 but the inter-relationships between these factors still remain unclear. Hyperandrogenism is the most important manifestation and diagnostic criteria of PCOS, and the Androgen Excess and PCOS Society (AES) has proposed that hyperandrogenism should be the essential condition to diagnose PCOS 2.There are three forms of serum testosterone (T), 60-65% was combined to sex hormones lobulin (SHBG) tightly, 35-40% was combined to albumin and free testosterone (FT) only consists 1-2% of total T. Which kind of serum androgen should be measured for diagnosis of PCOS remains controversial. Recently, it is believed assessments of free testosterone levels are more sensitive than the measurement of total T for establishing the existence of androgen excess3. Dr. Hestiantoro aims to determine the correlation between hormone profile, lipid profile and clinical profile with free testosterone index in subjects with PCOS, which is worthy of study.\nAccording to the Rotterdam diagnostic criteria, there are at least four phenotypes of PCOS: Subtype I - PCO & hyperandrogenism & oligo-ovulation, SubtypeⅡ - PCO & oligo-ovulation, SubtypeⅢ - hyperandrogenism & oligo-ovulation, Subtype Ⅳ - PCO + hyperandrogenism. Different phenotypes may display different endocrine disorders.\nDr. Hestiantoro has shown that high FTI in PCOS patients is associated with high triglyceride levels, high testosterone levels, and low SHBG levels, while Ferriman Gallwey score, as an indicator of hirsutism, shows no significant association with FTI. But another researcher who assessed the lipid profile in lean and non-lean PCOS patients, hyperandrogenemia was defined as free androgen index (FAI) ≥5, whose results show higher levels of total cholesterol, high-density lipoprotein cholesterol in lean patients with FAI <5 than in lean patients with FAI ≥ 5. There were no differences in lipid profile between non-lean patients with FAI ≥ 5 and non-lean patients with FAI <54. Another study has also confirmed these results with Ferriman-Gallwey scores(FGS) and triglycerides are significantly higher in PCOS patients5. In another study, PCOS patients with adrenal hyperandrogenism do not exhibit deterioration in insulin resistance and lipid profile despite the higher degree of total androgens6.\n\nSo what is the real correlation between hormone profile, lipid profile, and clinical profile with free testosterone index in subjects with PCOS? In my opinion, different PCOS phenotypes may display different endocrine and metabolic disorders, and FTI (free testosterone index) is a very valuable potential measurement to diagnose PCOS. It is of great significance to identify endocrine and metabolic characteristics of different phenotypes, but I am afraid the sample size of Dr.Hestiantoro's research is still too small to answer this question, and I suggest the authors may clarify PCOS patients to more detailed phenotypes and may have more interesting findings with a bigger sample size.\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? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "4479",
"date": "13 Mar 2019",
"name": "Andon Hestiantoro",
"role": "Author Response",
"response": "Dear Dr. Xue Lian Li,We are very sorry for the late reply. Thank you for your constructive comments and suggestions about our study.I admit that one of the limitations of this study is the small sample size, which makes the generalisation of these results to entire PCOS populations difficult. Therefore, future studies with greater sample size and longer observation durations should be established.As you mentioned previously in your reviews, I did not differentiate and classify the phenotype of PCOS in this study. Distinguishing PCOS phenotypes is an important aspect in the diagnosis and treatment of this syndromes, for different phenotypes might present in different manifestations and required different approach. This might be the reason why some of the findings in our study showed different or even contradictory results to another studies.Once again, thank you for your kind reviews. I look forward to be working and collaborating with you in some future researches. Cordially, Andon Hestiantoro"
}
]
},
{
"id": "44622",
"date": "07 Mar 2019",
"name": "Zheying Min",
"expertise": [
"Reviewer Expertise Mitochondrial Metabolism",
"Endocrine",
"Stem 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\nThis study describes the correlation between hormone profile, lipid profile and clinical profile with free testosterone index in subjects with PCOS. Triglycerides, testosterone, and SHBG were associated with hyperandrogenism in PCOS patients, while FGS showed no such association. Overall the paper is well structured and written, with results and a strong discussion. However I would have the following comments to the paper:\nThe average BMI of all PCOS patients were more than 25 (27.6). Could obesity influence the statistical analysis of the hormone? There are some evident errors in Tables. For example, In Table 1, is the average value of the body weight 0.6? The manuscript must be edited again for typo errors. In the discussion, “There are marked increases in GnRH and LH secretion, along with relative deficit in FSH secretion……” I think it should change deficit to deficiency. There are too many references in the discussion part, which should be moved to the introduction part. Because it only includes 9 references in the introduction.\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": "4480",
"date": "13 Mar 2019",
"name": "Andon Hestiantoro",
"role": "Author Response",
"response": "Dear Dr. Zheying Min,Thank you for your constructive comments and suggestions about our study. Please find below a point-by-point response to your reviews. The average BMI of all PCOS patients were more than 25 (27.6). Could obesity influence the statistical analysis of the hormone?There is a possibility that obesity affect the lipid and hormonal profiles of our subjects. Pathophysiologically speaking, obesity is closely related with dyslipidemia as well as insulin resistance, therefore statistical analysis with body mass index adjustment should be performed to independently study the association between PCOS and hormone concentrations. There are some evident errors in Tables. For example, In Table 1, is the average value of the body weight 0.6?I am sorry for my carelessness and thank you for alerting me about this mistakes. We will revise it soon. The manuscript must be edited again for typo errors. In the discussion, “There are marked increases in GnRH and LH secretion, along with relative deficit in FSH secretion……” I think it should change deficit to deficiency.I am sorry again for my carelessness and we also will revise it soon.There are too many references in the discussion part, which should be moved to the introduction part. Because it only includes 9 references in the introduction.We will re-arrange it and see if any of these references could be moved to introduction part.Once again, thank you for your kind reviews. I look forward to be working and collaborating with you in some future researches. Cordially, Andon Hestiantoro"
}
]
}
] | 1
|
https://f1000research.com/articles/8-94
|
https://f1000research.com/articles/8-93/v1
|
23 Jan 19
|
{
"type": "Research Article",
"title": "A high dose of total recombinant FSH suppresses granulosa cell apoptosis and maintains oocyte quality in endometriosis: A cross-sectional study",
"authors": [
"Budi Wiweko",
"Yassin Yanuar Mohammad",
"Naylah Muna",
"Kresna Mutia",
"Julianto Witjaksono",
"Nuri Purwito Adi",
"Mila Maidarti",
"Achmad Kemal Harzif",
"Gita Pratama",
"Kanadi Sumapraja",
"R. Muharam",
"Andon Hestiantoro",
"Yassin Yanuar Mohammad",
"Naylah Muna",
"Julianto Witjaksono",
"Nuri Purwito Adi",
"Mila Maidarti",
"Achmad Kemal Harzif",
"Gita Pratama",
"Kanadi Sumapraja",
"R. Muharam",
"Andon Hestiantoro"
],
"abstract": "Background: Endometriosis is one of the most common conditions causing infertility and an indication to undergo in vitro fertilization (IVF). High apoptosis rate and oxidative stress in patients with endometriosis are believed to negatively affect the IVF success rate. However, there have been conflicting results on the effect of endometriosis on IVF success, and there have been limited studies that directly assess endometriosis and its effect on oocyte quality. This study was performed to explore the correlation between mRNA BAX/BCL-2 expression and oocyte quality in endometriosis compared to non-endometriosis subjects. Methods: This was a cross-sectional study. 15 endometriosis and 15 non-endometriosis subjects were recruited through convenience sampling at Cipto Mangunkusumo Hospital, Jakarta. All subjects underwent follicle stimulation with recombinant follicle-stimulating hormone (FSH). Granulosa cells were collected and tested for BAX and BCL-2 expression and the results were compared to the oocyte quality and fertilization rate of the patients. Results: The total dose of recombinant FSH received by the endometriosis group was significantly higher compared with that of the non-endometriosis group (p = 0.005). There was a difference in BAX level (p = 0.029) and BCL-2 level (p<0.001) between groups. However, the BAX/BCL-2 ratio did not differ significantly (p = 0.787) between groups. No significant correlation was found between the BAX/BCL-2 ratio and any of the oocyte quality parameters measured. Conclusion: We found that there is a significantly higher dose in total dose recombinant FSH received by the endometriosis group compared with the non-endometriosis group. We also found that there was no significant difference in BAX/BCL-2 ratio between the endometriosis and non-endometriosis groups.",
"keywords": [
"Apoptosis",
"BAX/BCL-2 ratio",
"Endometriosis",
"Oocyte Quality",
"r-FSH"
],
"content": "Introduction\n\nEndometriosis is a condition in which endometrial tissues and glands are found outside the uterine cavity. This condition might cause pelvic pain and infertility. Ectopic endometrial tissue causes chronic inflammation, widespread fibrotic change, and adhesion1,2. Endometriosis occurs in 25–40% of women with infertility, and 30–50% of women with endometriosis also have infertility3,4. Endometriosis has been found to cause folliculogenesis and oocyte maturation disturbances, an increase in oxidative stress, and an imbalance in inflammatory cytokines that result in infertility5–8. In vitro fertilization (IVF) is one of the treatment options for patients with endometriosis. In 2016, in Indonesia, there were 8152 IVF cycles were performed, 7.03% of them were in women with endometriosis9.\n\nStudies on endometriosis affecting IVF success rate have shown conflicting results10. Some meta-analyses report that endometriosis affects the rate of pregnancy, miscarriage, and live birth during IVF. However, there has not been any adequate study that evaluates oocyte quality as an outcome of endometriosis. Oocyte quality is defined as the oocyte’s ability to undergo maturation and fertilization11. Some studies that evaluated oocytes quality from endometriosis patients found that oocyte quality were poorer regardless of sperm quality and uterine cavity condition12,13. Increased apoptosis rates have been found in endometriosis ovaries, and this has been linked to decreased oocyte quality14. Studies on apoptosis show a difference between endometriosis ovaries and non-endometriosis ovaries due to increased oxidative stress15. This condition triggers apoptosis through intrinsic and extrinsic pathways as well as meiotic spindle dysfunction and direct destruction through lipid peroxidase16,17. Apoptosis is regulated by specific genes that code protein and initiate the whole process. B-cell lymphoma/leukemia 2 (BCL-2) plays a crucial role in regulating apoptosis. This protein family consists of two categories: apoptosis inhibitors, BCL-2 and BCL-2-like 1 (BCL-2-LI or BCL-XL); as well as an apoptosis trigger, BCL-2-associated X protein (BAX)18. This study aimed to measure the gene expression of these apoptosis regulators, BCL-2 and BAX, and evaluate their correlation with oocyte quality in endometriosis and non-endometriosis IVF patients. The proposed conclusion was that there was significant difference in BAX/BCL-2 mRNA expression in patients with endometriosis and non-endometriosis.\n\n\nMethods\n\nSubjects were selected through convenience sampling. In total, 30 women were recruited for the study, who were undergoing IVF treatment at Yasmin Kencana Clinic, Cipto Mangunkusumo Hospital, Jakarta. Sample recruitment was done in June 2016 to August 2017.\n\nAll subjects found to positively have endometriosis, determined from a clinical examination and imaging procedures done by an obstetrician/gynecologist, were included in this study as a case group. However, subjects found to have other ovarian disturbances, history of ovarian removal, smoking, alcohol consumption, and sperm factor were excluded. For control group, we included all subjects with ovarian disturbance other than endometriosis and excluded subjects with male infertility factor.\n\nGranulosa cell sampling was conducted in the IVF Laboratory, while RNA isolation and quantitative real-time polymerase chain reaction (PCR) were done in the Integrated Laboratory Faculty of Medicine, Universitas Indonesia, between August and October 2017.\n\nThis study was conducted according to ethical standards in the Declaration of Helsinki and approved by the Faculty of Medicine Universitas Indonesia-Dr. Cipto Mangunkusumo General Hospital Research Ethical Committee with approval number LB.02.01/X.2/871/2017. All subjects in this study had the research clearly explained to them and have signed informed consent.\n\nRNA and DNA extraction from granulosa cells. Follicular fluid and granulosa cells were taken during ovum pickup, as part of routine IVF treatment. Granulosa cells were obtained during follicular aspiration and separated from the oocyte. These cells were then kept in a tube containing 500 uL RNA Later solution and stored at a temperature of -80°C.\n\nRNA was isolated from granulosa cells using the QIAamp RNA Blood Mini Kit (QIAGEN) according to the modified QIAamp RNA Blood Mini Handbook19. Buffer RLT and beta-mercaptoethanol were set with a 600 uL:6 uL ratio for each sample. Briefly, the granulosa cell samples in RNA Later solution were thawed and centrifuged at 20°C for 5 minutes (8000 × g). The supernatant was discarded, and RLT buffer mixed with beta-mercaptoethanol was added and homogenized. The sample was moved to a QIAshredder spin column and centrifuged for 8 minutes at 20°C (8000 × g). 600 uL Ethanol 70% was added to flow-through liquid, homogenized, and moved into QIAmp spin column. The QIAamp spin column was then centrifuged for 15 seconds (16.000 × g). The flow-through was then discarded and the column was washed using 600 uL RW1 buffer, centrifuged for 15 seconds (16.000 × g), and the flow-through was discarded. 500 uL RPE buffer was then added into the column for final washing, centrifuged for 15 seconds (16.000 × g), and the flow-through was discarded. This step was repeated twice. 25 uL RNase-free water was then added into the column as an elution solution to bind the RNA and the column was incubated for 10 minutes. The column then underwent centrifuge for 1 minute (16.000 × g) and the RNA collected at the collection tube under QIAamp spin column. The RNA concentration was measured with Nano Drop and stored at -80°C.\n\ncDNA was synthesized from RNA using the Qiagen Quantitect Reverse Transcription Kit according to the modified Quantitect Reverse Transcription Mini Handbook protocol20. The RNA template was mixed with RNase-free water with certain ratio according to RNA concentration obtained to gain equal cDNA concentration for all samples. This RNA and RNase-free water mix will obtain total volume of 14 uL. gDNA Wipeout buffer was then added into the mix and then incubated at 42°C for 10 minutes. Into each sample was then added Reverse-transcription master mix 1 uL, Quantiscript RT buffer 4 uL, and RT Primer mix 1 uL. the sample was then incubate at 42°C for 15 minutes and 95°C for 3 minutes. cDNA samples were then stored at -20°C until ready to be used.\n\nQuantitative real-time PCR was conducted using Qiagen Quantitect SYBR Green PCR Kit according to the modified Quantitect SYBR Green PCR Mini Handbook protocol21. qPCR was done by using absolute quantification, which measures sample by referring to standard curve. Standard curve was constructed using gBlocks oligonucleotide which made from amplicon product from each BAX and BCL-2 primer. The quantitative real-time PCR was performed using primers in Table 1.\n\nPCR master mix was firstly prepared and contain of 12.5 uL SYBR Green PCR mix, 0.25 uL forward primer, 0.25 uL reverse primer, and 10 uL nuclease-free water for each sample. After homogenized, the cDNA samples were then added into tube containing PCR mix and placed in real-time PCR instrument (Qiagen Rotor Gene-Q real-time PCR). Thermal profile of the instrument was set as shown in Table 2, while annealing temperature was set according do melting temperature in Table 1 for each gene, 58°C for BAX and 60°C for BCL-2.\n\nConstruction of standard curve was performed using the same PCR master mix composition, except for the cDNA samples. The cDNA was replaced by gBlocks oligonucleotide which has known concentration and diluted 5 times. The concentration of each gBlocks used were 1 until 1 × 10ˆ-8 ng/uL. Each gBlocks concentration was run duplicate to obtain standard curve with coefficient of determination value close to 1.\n\nThe standard curve was then used to measure each sample.\n\nOocyte quality assessment. Oocyte quality was measured by assessing morphology, maturation index, and fertilization rate. This assessment was performed under an Inverted Microscope (Olympus). Oocyte morphology was classified with Xia criteria, and mean score was calculated from the total number of oocytes obtained per subject22. The maturation index was defined as percentage of oocytes in metaphase II, whereas the fertilization rate was defined as percentage of metaphase II oocytes which successfully fertilized.\n\nDuring the stimulation phase, subjects were exposed to recombinant FSH, and the total dose of FSH might influence the result of the study; therefore, the total dose of FSH provided to the participants was noted.\n\nBAX/BCL-2 ratio was the independent variable, whereas the oocyte quality was the dependent variable. Statistical analysis was conducted with IBM SPSS (Statistical Package for Social Sciences) version 22, to test distribution and comparison between the endometriosis group and the control group for each variable. A correlation test was done to test any association between BAX/BCL-2 ratio and oocyte quality indicators in both groups. Variables with normally distributed data were tested with an independent t test, whereas other variables were tested with the Mann-Whitney test.\n\n\nResults\n\nA total of 30 subjects took part in this study: endometriosis group (n = 15) and control group (n = 15). Characteristics of these subjects are in Table 3. The patients’ age in both groups showed a normal distribution, with a mean age of 33.27±4.448 for endometriosis group and 32.67±3.559 for non-endometriosis group. Total dose of recombinant FSH received was calculated and there was a significant difference between the groups; mean dose of FSH received by the endometriosis group was 3760.0±1054.15 while the non-endometriosis group was 2763.3±700.82 (p=0.005). The total number of oocytes obtained during ovum pickup was normally distributed in both groups (Table 3). The mean number of oocytes was lower in the endometriosis group, although the difference was not significant.\n\nGene expression of BAX and BCL-2 were significantly different between the groups, with the concentrations of both genes being lower in the endometriosis group (Table 4). However, there was no statistically significant difference between the BAX/BCL-2 ratio between groups (p=0.787).\n\nOocyte quality was measured with three indicators: morphology, maturation index, and fertilization rate. We found no oocyte quality difference in the endometriosis group compared with the non-endometriosis group (Table 4).\n\n\nDiscussion\n\nBAX and BCL-2 are two proteins that play a crucial role in cell apoptosis through the intrinsic pathway. BAX is an apoptosis initiator, whereas BCL-2 is anti-apoptotic. A previous study by Tommi et al. found that the ratio between BAX and BCL-2 is one of the important indicators of apoptosis activity in cells23. The increases in BAX/BCL-2 ratio correspond to increased apoptosis activity and vice versa.\n\nIn this study, we found significant differences in BAX and BCL-2 mRNA concentrations in both groups. The expression of both genes was lower in the endometriosis group compared with the non-endometriosis group (p = 0.029 in BAX and p<0.001 in BCL-2). However, there was no difference in the BAX/BCL-2 mRNA ratio between groups (p = 0.787). This result showed that there is no difference in apoptosis activity in endometriosis granulosa cells compared with non-endometriosis cells. This finding is inconsistent with a previous study by Wiweko et al, which stated that BAX mRNA concentration in endometriosis granulosa cells is higher than in the control group24. However, the lower level of BAX mRNA does not imply less apoptotic activity. One interaction model of BAX and BCL-2 protein showed that while BAX induces apoptosis, BCL-2 inhibits the process25. In this case, the low level of BCL-2 in endometriosis might result in decreased anti-apoptotic activity. There was also a study that showed fewer follicles in rats with BCL-2 deficiency, and that higher BCL-2 expression would suppress apoptosis and atresia26.\n\nFilali et al. reported the essential role of BCL-2 mRNA in granulosa cells in determining oocyte quality. BCL-2 mRNA expression was found to be higher in the granulosa cells of mature oocytes, whereas there was no difference in BAX mRNA in the same cells. Moreover, BCL-2 mRNA correlates with the fertilization rate due to the lower apoptosis activity27. BAX/BCL-2 ratio has a more important role as an apoptosis marker, as mentioned earlier. However, in this study, we found no significant difference in apoptosis activity. This finding is inconsistent with a previous study by Wiweko et al. in 2017, which showed a higher BAX/BCL-2 ratio in the endometriosis group compared with the control group28.\n\nIn the present study, the oocyte number in the endometriosis group was lower than in the non-endometriosis group, but was not significant (p = 0.336). This finding is in line with studies by Rossi et al. and Yang et al. that showed a smaller number of oocytes in endometriosis subjects29,30. Regarding oocyte quality, we found no differences in the mean oocyte score, maturation index, and fertilization rate between the groups. The mean oocyte score showed no significant difference (p = 0.611). The oocyte score was assessed by Xia criteria, which analyzes oocyte morphology based on perivitelline space, polar body II, and cytoplasm. This result is inconsistent with Shebl et al, who reported that fewer normal oocytes were observed in patients with endometriosis compared with control subjects31. Similarly, the maturation index and fertilization rate in both groups did not show a marked difference (p=0.225 and p=0.693, respectively). This is inconsistent with other related studies that showed maturation index and fertilization rate in endometriosis were poor; Rossi et al. and Yang et al. reported that less MII oocytes were found in endometriosis patients29,30. On the other hand, Luca et al. stated that there was no difference in the number of MII oocytes in an endometriosis group compared with a control group32.\n\nInconsistent with our finding, Barnhart et al. showed a lower fertilization rate in endometriosis patients33. The fertilization rate in patients with severe endometriosis was higher compared with subjects with mild/moderate endometriosis and control subjects. However, Yang reported that a decrease in fertilization rate only occurred in subjects with endometriosis grade III/IV, not those with grade I/II, compared with control subjects30.\n\nIn this study, there was no correlation between BAX/BCL-2 mRNA ratio and mean oocyte score, maturation index, and fertilization rate in both groups. This result suggests that similar apoptosis activity will result in similar oocyte quality. Although we found no difference in BAX/BCL-2 mRNA ratio and oocyte quality, there was a significant discrepancy in recombinant FSH dose given to each group (p=0.005). The mean FSH dose in the endomteriosis group was 3760 IU, whereas the control group 2763.3 received IU. Shen et al. reported that FSH might play a role in protecting granulosa cells from oxidative stress, although the mechanism is currently unknown. FSH also protects mitochondria in granulosa cells undergoing oxidative stress34. Our study shows that subjects with endometriosis may benefit from higher dose of recombinant FSH to produce oocytes with similar quality to those in the control group. Therefore, further studies on FSH and its correlation with apoptosis activity and oocyte quality during ovum pickup are needed.\n\n\nConclusions\n\nAlthough we found lower levels of BAX mRNA and BCL-2 mRNA in patients with endometriosis, the ratio of these mRNAs was not statistically significant between groups. There was also no correlation between BAX/BCL-2 mRNA ratio and oocyte number, FSH dose, and oocyte quality in both groups. However, we noted a significantly higher dose of recombinant FSH given to the subjects with endometriosis compared with those in the non-endometriosis group. We suggest for future studies a similar protocol in collecting granulosa cells from size-matched follicles to minimize the discrepancy in BAX and BCL-2 mRNA concentrations. Moreover, examination and records on basal antral follicle and anti-müllerian hormone concentration should become a part of the standard IVF protocol.\n\n\nData availability\n\nFigshare: DATA_A High Dose of Total Recombinant FSH Suppress Granulosa Cells Apoptosis, https://doi.org/10.6084/m9.figshare.7483574.v135.\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\nFinancial support research grant from University of Indonesia—Dr Cipto Mangunkusumo General Hospital, Jakarta and Indonesia.\n\n\nAcknowledgements\n\nThe authors would like to thank Indonesian Reproductive Medicine Research and Training Center (INAREPROMED) and Yasmin Clinic Rumah Sakit Dr. Cipto Mangunkusumo Hospital teams for assistance and support in this study.\n\n\nReferences\n\nGiudice LC, Kao LC: Endometriosis. Lancet. 2004; 364(9447): 1789–99. PubMed Abstract | Publisher Full Text\n\nSomigliana E, Viganò P, Tirelli AS, et al.: Use of the concomitant serum dosage of CA 125, CA 19-9 and interleukin-6 to detect the presence of endometriosis. Results from a series of reproductive age women undergoing laparoscopic surgery for benign gynaecological conditions. Hum Reprod. 2004; 19(8): 1871–6. PubMed Abstract | Publisher Full Text\n\nHummelshoj L, Prentice A, Groothuis P: Update on endometriosis. Womens Health (Lond). 2006; 2(1): 53–6. PubMed Abstract | Publisher Full Text\n\nOzkan S, Murk W, Arici A: Endometriosis and infertility: epidemiology and evidence-based treatments. Ann N Y Acad Sci. 2008; 1127: 92–100. PubMed Abstract | Publisher Full Text\n\nGarrido N, Navarro J, Remohí J, et al.: Follicular hormonal environment and embryo quality in women with endometriosis. Hum Reprod Update. 2000; 6(1): 67–74. PubMed Abstract | Publisher Full Text\n\nMatsuzaki S, Schubert B: Oxidative stress status in normal ovarian cortex surrounding ovarian endometriosis. Fertil Steril. 2010; 93(7): 2431–2. PubMed Abstract | Publisher Full Text\n\nRichter ON, Dorn C, Rösing B, et al.: Tumor necrosis factor alpha secretion by peritoneal macrophages in patients with endometriosis. Arch Gynecol Obstet. 2005; 271(2): 143–7. PubMed Abstract | Publisher Full Text\n\nToya M, Saito H, Ohta N, et al.: Moderate and severe endometriosis is associated with alterations in the cell cycle of granulosa cells in patients undergoing in vitro fertilization and embryo transfer. Fertil Steril. 2000; 73(2): 344–50. PubMed Abstract | Publisher Full Text\n\nPERFITRI: Indonesian Association for In Vitro Fertization’s National Report 2017. IAIVF; 2018.\n\nGonzález-Comadran M, Schwarze JE, Zegers-Hochschild F, et al.: The impact of endometriosis on the outcome of Assisted Reproductive Technology. Reprod Biol Endocrinol. 2017; 15(1): 8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSanchez AM, Vanni VS, Bartiromo L, et al.: Is the oocyte quality affected by endometriosis? A review of the literature. J Ovarian Res. 2017; 10(1): 43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHull MG, Williams JA, Ray B, et al.: The contribution of subtle oocyte or sperm dysfunction affecting fertilization in endometriosis-associated or unexplained infertility: a controlled comparison with tubal infertility and use of donor spermatozoa. Hum Reprod. 1998; 13(7): 1825–30. PubMed Abstract | Publisher Full Text\n\nNavarro J, Garrido N, Remohi J, et al.: How does endometriosis affect infertility? Obstet Gynecol Clin North Am. 2003; 30(1): 181–92. PubMed Abstract | Publisher Full Text\n\nDíaz I, Navarro J, Blasco L, et al.: Impact of stage III-IV endometriosis on recipients of sibling oocytes: matched case-control study. Fertil Steril. 2000; 74(1): 31–4. PubMed Abstract | Publisher Full Text\n\nGarcia-Velasco JA, Arici A: Is the endometrium or oocyte/embryo affected in endometriosis? Hum Reprod. 1999; 14(Suppl 2): 77–89. PubMed Abstract | Publisher Full Text\n\nAgarwal A, Gupta S, Sikka S: The role of free radicals and antioxidants in reproduction. Curr Opin Obstet Gynecol. 2006; 18(3): 325–32. PubMed Abstract | Publisher Full Text\n\nSanchez AM, Somigliana E, Vercellini P, et al.: Endometriosis as a detrimental condition for granulosa cell steroidogenesis and development: From molecular alterations to clinical impact. J Steroid Biochem Mol Biol. 2016; 155(Pt A): 35–46. PubMed Abstract | Publisher Full Text\n\nDepalo R, Cavallini A, Lorusso F, et al.: Apoptosis in normal ovaries of women with and without endometriosis. Reprod Biomed Online. 2009; 19(6): 808–15. PubMed Abstract | Publisher Full Text\n\nQIAGEN: QIAamp®. RNA Blood Mini Handbook. 2010. Reference Source\n\nQIAGEN: QuantiTect®. Reverse Transcription Handbook. 2009. Reference Source\n\nQIAGEN: QuantiTect® SYBR® Green PCR Handbook.2011. Reference Source\n\nXia P: Intracytoplasmic sperm injection: correlation of oocyte grade based on polar body, perivitelline space and cytoplasmic inclusions with fertilization rate and embryo quality. Hum Reprod. 1997; 12(8): 1750–5. PubMed Abstract\n\nTommi V, editor. Regulation of apoptosis in the female reproductive system. University of Oulu; 2002. Reference Source\n\nWiweko B, Bowolaksono A, Narasati S, et al.: The Expression of Granulosa Cells BAX mRNA Among Endometriosis and Normal Women. Adv Sci Lett. 2017; 23: 7012–4. Publisher Full Text\n\nHadi R: Mekanisme Apoptosis Pada Regresi Sel Luteal. Maj Kesehat PharmaMedika. 2011; 3: 246–54. Reference Source\n\nHussein MR: Apoptosis in the ovary: molecular mechanisms. Hum Reprod Update. 2005; 11(2): 162–77. PubMed Abstract | Publisher Full Text\n\nFilali M, Frydman N, Belot MP, et al.: Oocyte in-vitro maturation: BCL2 mRNA content in cumulus cells reflects oocyte competency. Reprod Biomed Online. 2009; 19(Suppl 4): 4309. PubMed Abstract | Publisher Full Text\n\nWiweko B, Muna N, Gunawarti D, et al.: High bax-bcl-2 Ratio Expression on Granulosa Cells from Endometriosis Patients. Adv Sci Lett. 2017; 23(7): 6720–2. Publisher Full Text\n\nRossi AC, Prefumo F: The effects of surgery for endometriosis on pregnancy outcomes following in vitro fertilization and embryo transfer: a systematic review and meta-analysis. Arch Gynecol Obstet. 2016; 294(3): 647–55. PubMed Abstract | Publisher Full Text\n\nYang C, Geng Y, Li Y, et al.: Impact of ovarian endometrioma on ovarian responsiveness and IVF: a systematic review and meta-analysis. Reprod Biomed Online. 2015; 31(1): 9–19. PubMed Abstract | Publisher Full Text\n\nShebl O, Sifferlinger I, Habelsberger A, et al.: Oocyte competence in in vitro fertilization and intracytoplasmic sperm injection patients suffering from endometriosis and its possible association with subsequent treatment outcome: a matched case-control study. Acta Obstet Gynecol Scand. 2017; 96(6): 736–44. PubMed Abstract | Publisher Full Text\n\nLuca A, Nemescu D, Butnaru M, et al.: Ovarian stimulation outcome in infertile women with endometriosis undergoing IVF. Ginekol Pol. 2016; 87(1): 37–41. PubMed Abstract | Publisher Full Text\n\nBarnhart K, Dunsmoor-Su R, Coutifaris C: Effect of endometriosis on in vitro fertilization. Fertil Steril. 2002; 77(6): 1148–55. PubMed Abstract | Publisher Full Text\n\nShen M, Jiang Y, Guan Z, et al.: FSH protects mouse granulosa cells from oxidative damage by repressing mitophagy. Sci Rep. 2016; 6: 38090. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuna N: DATA_A High Dose ofTotal Recombinant FSH Suppress Granulosa Cells Apoptosis.xlsx. figshare. Dataset. 2018. https://www.doi.org/10.6084/m9.figshare.7483574.v1"
}
|
[
{
"id": "48952",
"date": "27 Jun 2019",
"name": "Chii-ruey Tzeng",
"expertise": [
"Reviewer Expertise Endometriosis in ART"
],
"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 entitled “A high dose of total recombinant FSH suppresses granulosa cell apoptosis and maintains oocyte quality in endometriosis: A cross-sectional study” - this study was conducted in 15 patients with endometriosis and 15 control patients without endometriosis. The draft is oversimplified with very preliminary results, and most data were published elsewhere. This study has a lack of novelty.\nThe sample size was too small to interpret the results with a proper power. The sample size estimation to get a proper power should be performed before conduction of the study.\n\nSince the sample size is small, the baseline characteristics of patients with endometriosis should be described more clearly, such as BMI, duration of infertility, primary or secondary infertility, baseline FSH/LH/E2/P4, antral follicle count, AMH, or the revised American Fertility Society (rAFS) score, etc. Such information is basic and important. The stage of endometriosis could also be a confounding factor to affect the results.\n\nThe size of endometriosis of ovary, unilateral or bilateral should also be mentioned.\n\nThe title is interesting. However, their data didn’t support their title. The gene expression of BAX and BCL-2 is only an association in patients with endometriosis. There was no ex-vivo functional assay, such as comparing the apoptosis activity in granulosa cells from patients with endometriosis under different doses of recombinant FSH, to support the title of manuscript. Thereafter, the title should be considered to be rewritten.\n\nThe conclusion section in the abstract should be rewritten. It was loosely organized and did not incorporate the title of this manuscript.\n\nWhether this was a retrospective or prospective study according to the statement in the methods section could not be determined clearly. Please describe it clearly.\n\nWhat are the diagnosis criteria of endometriosis in this manuscript, especially for clinical examination mentioned in the methods section? Besides, the grade of endometriosis should be illustrated.\n\nThe authors collected follicular fluid during ovum pickup, but we did not see any data from the analysis of follicular fluid.\n\nPlease describe the protocol of controlled ovarian hyperstimulation more clearly, including the medications used for COH, the method(s) of insemination, etc.\n\nIt was not proper to interpret the data from two small groups as normal distribution.\n\nThe goal of this study is not clear. One can not be sure if the authors tried to compare the results from endometriosis and non-endometriosis, or if they wanted to compare the dose effect of recombinant FSH in patients with endometriosis. Please make it more obvious.\n\nThe basic characteristics of patients should be more detailed, including AMH and/or AFC, COH protocol, COH duration, medication for triggering, maximal E2 level, etc.\n\nIn Table 4, the mean value of maturation index in the non-endo group is 1, but the range of maturation index is 0.25-1.00. Is the mean value correct?\n\nThe results of this manuscript were different from previous studies, but the authors did not explain the possible reason for the different findings in the discussion section. They published similar studies in 2017. This manuscript is very similar to their previous studies. Thereafter, there were no new insights from this draft.\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? 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? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-93
|
https://f1000research.com/articles/8-92/v1
|
23 Jan 19
|
{
"type": "Research Article",
"title": "The economic impact of critical congenital heart disease to the health system and families in Colombia",
"authors": [
"Cindy Lorena Chamorro Velásquez",
"Nestor Fernando Sandoval Reyes",
"Alejandra Taborda Restrepo",
"Sandra Vanessa Romero Ducuara",
"Maria Teresa Domínguez",
"Gloria Amparo Troncoso Moreno",
"Alejandra Fonseca Cuevas",
"Hernan Camilo Aranguren Bello",
"Pablo Andres Bermúdez Hernadez",
"Pablo Sandoval Trujillo",
"Rodolfo Jose Dennis",
"Darío Londoño Trujillo",
"Cindy Lorena Chamorro Velásquez",
"Nestor Fernando Sandoval Reyes",
"Alejandra Taborda Restrepo",
"Sandra Vanessa Romero Ducuara",
"Maria Teresa Domínguez",
"Gloria Amparo Troncoso Moreno",
"Alejandra Fonseca Cuevas",
"Hernan Camilo Aranguren Bello",
"Pablo Andres Bermúdez Hernadez",
"Pablo Sandoval Trujillo",
"Rodolfo Jose Dennis"
],
"abstract": "Background: Critical congenital heart disease (CCHD) make up a group of heart diseases present in newborns since the prenatal period and requiring early intervention through surgery or percutaneous interventions in the first year of life. Little is known about the societal economic impact associated with their care in low to middle income countries. We estimated direct medical costs, out-of-pocket expenditures and indirect costs of CCHD patients in Colombia. Methods: The methodology to estimate costs involved four stages: identification, measurement, and assessment of resources consumed, and total cost calculation. Regarding medical costs, hospital and ambulatory costs were estimated for the patient’s first year of life using clinical records of 73 patients and with thematic experts. A survey was carried out on 20 children´s caregivers to determine the out-of-pocket expenses and indirect costs. For this estimation, a descriptive analysis was made on the survey taking into account the reported salary. All costs are expressed in US dollars (2017 exchange rates). Results: The average direct medical hospital costs for CCHDs were $25,835 and the ambulatory costs reached $480. Indirect costs were $1,303 and out-of-pocket expenses were $2,058, which for families with an income lower than one monthly minimum wage (1 SMMLV) in 2017 correspond to $250. The impact on their budget was 57%. Conclusions: CCHDs represent an important economic impact both for the Colombian General Social Health System and for families. This study made it possible to estimate the costs that are not easily visible and thus quantified.",
"keywords": [
"Economic impact",
"Cost of Illness",
"Cost Analysis",
"Hospital Costs",
"Out-of-pocket Expenditures",
"Indirect Expenditures"
],
"content": "Abbreviations:\n\nCHD: Congenital Heart Disease; CCHD: Critical Congenital Heart Disease; COI: Cost of illness; FCI-IC: Fundación Cardioinfantil – Institute of Cardiology; SMMLV: Monthly Minimum Legal Wage\n\n\nIntroduction\n\nThe term critical congenital heart disease (CCHD) refers to a group of structural malformations of the heart which are present from the prenatal period. They represent more than one third of all Congenital Heart Diseases (CHD)1,2. Their incidence ranges from 1 in 15,000 to 1 in 26,000 live births and their prevalence corresponds to 147.4 cases per 100,000 live births3. CCHDs are an important cause of child mortality in the first month of life, and consequently it is necessary to start early surgical or interventional treatment in the first year of life to grant the patient’s survival2,4. 29.5% of newborns with CCHD are diagnosed late5, a fact associated with greater risk of morbidity and mortality6. The CCHD have a significant economic impact on the health systems and the patients’ families; thereby, no studies on indirect costs are available, and few of them address direct medical costs at world level7.\n\nThe purpose of carrying out a costs study on a disease (COI “Cost of Illness”), is to perform a descriptive analysis to identify and measure the costs of a disease in monetary terms to estimate its economic burden on society8,9. COI studies may take different perspective that determines the whole costs process; they estimate costs from the standpoint of society, the health system, the patient or the families. The perspective of society is the most comprehensive one and most widely recommended, since it includes all costs, independently from who incurs them. In consequence, besides direct medical costs, out-of-pocket expenses covered by the family, and indirect costs, are included. This society perspective allows to carry out a more complete analysis and include all costs of opportunity attributed to a disease. However, this perspective requires a larger collection and availability of information effort, a fact that hinders its use in COI for low prevalence diseases10,11.\n\nOverall, COI studies contemplate three cost categories: direct, indirect and intangible costs, the latter are rarely quantified, so it is accepted that only direct and indirect costs are considered. Direct medical costs are costs related with medical assistance, treatment, rehabilitation, etc.; non direct medical costs involve expenses in non-medical resources, i.e., transportation and nutrition outside the home. Lastly, indirect costs refers to the loss of productivity associated with the disease11.\n\nThe micro-costing, and gross-costing approaches are used to estimate direct medical costs. The micro-costing is a “bottom-up” process whereby every resource is identified; measuring, assessment and calculation of the total cost are carried out; contrarily, the gross-costing estimates costs from a “top-down” approach, drawing an average of the total cost. The human capital approach and the friction cost approach estimate indirect costs. The loss of productivity is associated with the friction period, i.e., the time the patient takes to be replaced by another employee and for production to return to its former level12.\n\nFrom the economic standpoint, the needs of the population are endless but the resources to meet them are limited. Additionally, factors like technological development, cause health expenses to increase. The estimation of costs is important in making decisions, as it helps to define and prioritize policies and intervention related with health assistance11. Also, it provides information to make future costs projections in medical assistance and allocate resources according to budget limitations, thus granting the efficiency of policies8,11.\n\nIn view of the importance of CCHD, the complexity presented in terms of costs for society and the scarce literature available on the subject, it was necessary for this study to estimate costs and hence approach the economic burden from a social perspective in Colombia. For this study we benefit from the information provided by Fundación Cardioinfantil – Institute of Cardiology (FCI-IC) and its social program13,14, to access a population that is difficult to study. Although many of the expenses incurred by the patients and their families are subsidized by FCI-IC’s social program, the use of resources is real, which allowed to approximate to the real cost incurred by these CCHD patients in Colombia.\n\n\nMethods\n\nThe method used to estimate the costs was micro-costing, as the information collected allowed to reach a level of detail in the identification and measuring of resources, and to carry out the assessment based on the national reference prices, and not only on those of a particular institution. Besides, this method identifies gaps in expenses and include only events related with CCHD. The calculations were made in Microsoft Excel 2016. We did not perform any statistical tests.\n\nThe seven CCHDs included in the estimation of costs were: Pulmonary Atresia, Tetralogy of Fallot, Tricuspid Atresia, Truncus Arteriosus, Hypoplastic Left Heart Syndrome, Total Anomalous Pulmonary Venous Drainage and Transposition of Great Arteries. The decision to include these CCHDs was made following the approaches and recommendations taken on the subject at world level, thus the prioritization of these cardiopathies in the United States was taken as reference15.\n\nThe cost estimation process followed four stages10: 1. Identification of resources; 2. Measuring the use of resources; 3. Assessment of resources; 4. Calculating total costs. This methodology was followed, and the “bottom- up” process, i.e., the estimation of direct medical hospital costs followed an identification of resources, measuring (frequency and quantity) and assessment of resources (as per Tariffs Manual ISS 2001+35%, market prices in Colombia)16, and lastly, the total costs calculation, based on clinical records for 73 patients with the seven CCHDs from the data base of the Institute of Congenital Heart Disease at FCI-IC. Direct annual medical costs were estimated for every CCHD, and for indirect costs and out-of-pocket expenses, 20 surveys were applied to caregivers of children with three Infrequent Congenital Heart Diseases since there is a higher level of complexity in the collection of information on CCHDs. The caregivers were a convenience sample identified from a cohort of the PINOCCHIO Program (Innovation Program in Uncommon Human Congenital Heart Diseases for Colombia). This cohort was built through three main sources at Fundacion Cardionfantil – Instituto de Cardiologia – FCI–IC: medical databases of patients from the Congenital Heart Disease Institute; Pediatric Cardiovascular surgery service; and patients diagnosed during pediatric brigades carried out by FCI-IC social program “Give a life” in 12 cities in Colombia from the period 2010 to October, 2018. The eligibility criteria for the 20 caregivers of children were patients from the PINOCCHIO Cohort with detectable critical congenital heart diseases with the following diagnosis: Ebstein’s anomaly, Interrupted Aortic Arch and Pulmonary Valve Stenosis and which were less than two years of age at the time of the survey. The caregivers filled a case report form (CRF) whereby information on out-of-pocket expenses and indirect costs was obtained. Before caregivers started to answer the survey, a research assistant explained the objective, asked for signed informed consent, and were present to respond to any question. A copy of the CRF translated to English is attached as extended data.\n\nThe perspective taken in cost estimation was that of society. All costs are expressed in US dollars, firstly estimated in Colombian pesos for 2017, and then the average exchange rate for 2017 was applied ($2,951.15 COP for one US dollar)17.\n\nThe collected information was filtered, and only the hospital events in which patients were admitted by causes directly related with CCHDs were selected. Then, a validation was made with experts on the amount of resources included in this data base to verify if they were similar to usual clinical practice. Then, an analysis of hospital events per patient was performed, including very detailed information on the resources used, grouped under large line items: surgical and non-surgical procedures, laboratories, diagnostic aids, blood bank, medication and surgical medical materials. Finally, a base case was drawn for a patient, including the average of hospital events (surgical and non-surgical) taking place in a year for every CCHD, weighing the frequency of these events and the average number of days in hospital for each.\n\nThe estimation of ambulatory costs was also carried out in four stages; the identification of resources for ambulatory events in the seven CCHD, was validated by clinical experts. To measure resources, we asked the same experts how many medical consultations, echocardiograms, EKGs and chest x-rays are performed on average on these patients in their first year of life. The assessment of resources was carried out based on Tariff Manual ISS 200116; finally, the following formula was applied Cost = frequency or quantity*prices.\n\nThe source for procedure prices was Tariff Manual ISS 200116, a reference for tariffs used for contracting health services in Colombia. The adjustment of these tariffs for the Colombian market is done by taking the base tariff present in 2001 plus an increase of 35% (20% for the minimum cost and 50% for the maximum cost to perform a sensitivity analysis).\n\nIn order to estimate out-of-pocket expenses and indirect costs covered by the caregiver, a survey was applied to 20 caregivers of patients younger than two years with a CCHD, as detailed above. The survey included specific questions on monthly income, occupation, days on leave from work and expenses associated with the disease to establish the impact of out-of-pocket expenses and the loss of productivity in the family budgets.\n\nThe methodology used to estimate the out-of-pocket expenses corresponds to a weighted average according to the records of every family. The line items included were: transportation costs (transfer to the city), payments caused by private medical consultation, payment of sliding scale fees18, co-payment expenses (contributions from patients), paperwork expenses, expenses caused by transfer to a temporary residence, on meals outside the home, purchase of medication and other as payment for care of other children, telephone calls, among others. A monthly average for out-of-pocket expenses was estimated considering the age of the patient, to then calculate the annual out-of-pocket expense undertaken by every family.\n\nRegarding indirect costs, the number of days of leave was calculated for the caregivers who were economically active at the time of the survey and the equivalent cost of productive days lost was established by means of the human capital method, based on the net salary reported in the surveys. The average daily income was calculated based on the distribution of income of all caregivers, leading to a summation of the frequency in the level of expenses times the average daily income.\n\nOut-of-pocket expenses and indirect costs were differentiated considering the need for a surgical procedure. Lastly, to understand the magnitude of these costs, the impact of these expenses on the annual budget of the families was calculated, differentiated by levels of income and measured in terms of the minimal wage in Colombia, which corresponded to $250 monthly in 201719.\n\nThe ethical approval for the project was awarded by the Clinical Research Ethics Committee at FCI-IC (CEIC) for the supply and access to the database of patients with CCHD and for the collection of information on care providers. The informed consent was provided to be filled in by the CRF of patients from the PINOCCHIO Program Cohort, whereby information on out-of-pocket expenses and indirect costs of every care provided was obtained from those answering the survey.\n\n\nResults\n\nThe average annual hospital cost for the seven CCHDs was $25,835. Diagnosis of Truncus Arteriosus caused the highest cost, at $42,419 and Anomalous Pulmonary Venous Drainage had the lowest cost, at $16,904. Table 1 shows the results for the seven CCHDs, including inpatient costs with and without a surgical event, and the total cost (base cost), and the lowest and highest cost.\n\nTable 2 shows the results of annual ambulatory costs. The highest ambulatory cost was found in the diagnosis for Hypoplastic Left Heart Syndrome, at $719, and the lowest cost was represented by the Tetralogy of Fallot, at $357. The average ambulatory costs for the seven CCHDs was $487.\n\nThe variables for the socio-demographic characteristics of the target population were collected by means of a survey applied to caregivers: 90% of patients with CCHD belonged to a socio-economic level equal or lower than 3 (in a scale ranging from 1 to 6, 1 being the lowest level20); 55% were female, and the average age for patients was 11 months. Regarding the characteristics of caregivers, 40% had a secondary school level of education; 53% were employed, 42% reported having another occupation (housewife, unemployed); 75% stopped doing their most frequent activity to take up the patient’s care and 54% had an income lower than one monthly minimum wage (SMMLV).\n\nIndirect costs were estimated at $1,303 per caregiver; the average number of days caregivers were on leave was 140 days. Figure 1 shows the impact of indirect costs on the caregiver’s budget depending on their level of income. The highest impact is found among caregivers with incomes lower than one monthly minimum wage.\n\nBesides, the differentiated indirect cost for patients undergoing surgery or not was estimated, where the difference between undergoing surgery and not is observed, reaching $498. When the patient undergoing surgery the indirect cost was $1,466 and in the other hand the cost was $969.\n\nThe average out-of-pocket expense incurred by a family with a CCHD patient is $171, approximately $2,058 per year. The line item most frequently reported by the largest number of families was transportation costs in the city and meals outside the home (Table 3).\n\nAn estimate of out-of-pocket expenses was also made, differentiated for patients undergoing surgery where the difference was $1,300. When the patient undergoing surgery, the out-of-pocket expenses was $2,383.\n\nTo understand the economic impact imposed on families, it is important to estimate the impact on the family budget by income levels. For families earning less than one SMMLV (up to $249) the out-of-pocket expense represented 57% of their budget, assuming an income of $249, the widest interval range.\n\n\nDiscussion\n\nThe results of this study are consistent with a high economic impact of CCHDs on the health system and on families in Colombia. The consumption of resources that is made evident represents high average annual direct medical costs, both for inpatient and ambulatory services. Even out-of-pocket expenses undertaken by the families have a high impact on their budget, mainly in families with lower incomes.\n\nThe scarcity of literature on this subject makes this study relevant, not only in Colombia but also in other low and middle income countries. In addition, it is noteworthy to highlight the incorporation of estimates on direct costs and out-of-pocket expenses, an economic burden for families which is rarely taken into account due to the difficulty in collecting this information. The estimation of costs in this study has been a comprehensive process, incorporating different sources of information to get an estimate of the data described, data bases, validation by experts, surveys to caregivers, complying with the cost estimation methodology involving the four stages previously described.\n\nStudies on costs related with CCHDs mostly refer to the process of screening followed to detect these heart diseases, but not to the economic burden caused by these diseases as a whole. One study made by Peterson et al.15 approached this topic, although it focused on comparing inpatient events among CCHD patients diagnosed timely and late, and not on estimating the total financial burden assigned to the disease. However, inpatient costs for the first year of life were estimated for 2011 prices, which reached $100,200 in patients diagnosed late versus a cost of $69,500 for patients timely diagnosed.\n\nRegarding indirect costs and out-of-pocket expenses for CCHD, Connor et al.21 carried out a qualitative study that explored the perceptions of parents on the costs associated with having a child with congenital heart disease regarding direct medical and non-medical costs undertaken by the family, and how this economic burden resulted in more stress and emotional impact. Mughal et al.22 developed an observational study at a hospital in Lahore (Pakistan), and identified that 63.1% of the families contributed with the cost associated with the CHD patient’s treatment, and 12.3% of them contributed at 100%. A study by Raj et al.23, conducted in India, estimated costs associated with CHD. Although not focusing specifically on CCHD, it is the study that approaches this topic more closely. The main results indicated that the connection between total inpatient expenses and the family income was 0.93 times the annual family income; the average time lost due to leave of absence by the father was 35 days and the loss of working days was 15 days on average. A quantitative estimation on the indirect costs and out-of-pocket expenses associated with CCHDs was not found in the literature.\n\nThere were some limitations in this study, mainly related with the low number of cases available due to the low prevalence of most CCHDs, a fact that hindered the collection of more data which may have led to imprecision. This situation led to having to carry out the surveys within a period of nine months. In addition, since records did not provide enough information to identify ambulatory costs, the data had to be obtained by consensus with experts.\n\n\nConclusions\n\nCCHDs represent a high economic impact on the health system and families, especially on those with lower incomes. Cost analysis is a relevant topic for the health system and it requires a systematic and complex process. In this case, different primary and secondary information sources were available: data bases, surveys to caregivers and validation with experts on the topic. In addition, the bottom-up approach as a way to estimate a more real cost can be highlighted.\n\nOne of the most relevant added values included in the analysis for this study was the social perspective, which includes the measuring of indirect costs and out-of-pocket expenses. This dimension led to a larger data collection effort, more so when dealing with low-prevalence diseases. The results deriving from this study show the impact of these costs on the budgets of families when one of their members is diagnosed with a CCHD.\n\n\nData availability\n\nIn the following repository, you can find the dataset used for ambulatory costs, hospital costs and the survey data:\n\nOpen Science Framework: Underlying data of economic impact of CCHD in Colombia. DOI http://dx.doi.org/10.17605/OSF.IO/9J8KT24\n\nThe survey asked to 20 caregivers has been deposited in the following repository:\n\nOpen Science Framework: Survey asked to 20 caregivers of CCHD in Colombia. DOI http://dx.doi.org/10.17605/OSF.IO/67S4G25\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\nThis study is financed by the Colombian Fund for the Financing of Science, Technology and Innovation, Francisco José de Caldas – COLCIENCIAS, to the Program for Innovation in Infrequent Congenital Human Cardiopathies (PINOCCHIO from its Spanish acronym) – Contract 662-2015, awarded to Fundacion Cardioinfantil – Instituto de Cardiologia.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nWe would like to thank the support staff of the Program for Innovation in Infrequent Human Congenital Heart Diseases for Colombia (PINOCCHIO), the Help Desk of the FCI-IC for their procurement of the databases, especially to Maricela Atara Martinez for her constant collaboration; to Jonathan Campos and Oscar Rodriguez, previous research assistants of the project during the initial period of information gathering; to Juan Leonardo Novoa for his support during the review period and validation of the clinical histories of the patients included in the databases; and Dayan Roa for her collaboration with identification and contacting patients´ caregivers.\n\n\nReferences\n\nOlney RS, Ailes EC, Sontag MK: Detection of critical congenital heart defects: Review of contributions from prenatal and newborn screening. Semin Perinatol. 2015; 39(3): 230–237. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHarold JG: Cardiology patient page. Screening for critical congenital heart disease in newborns. Circulation. 2014; 130(9): e79–81. PubMed Abstract | Publisher Full Text\n\nSchultz AH, Localio AR, Clark BJ, et al.: Epidemiologic features of the presentation of critical congenital heart disease: implications for screening. Pediatrics. 2008; 121(4): 751–757. PubMed Abstract | Publisher Full Text\n\nBruno CJ, Havranek T: Screening for Critical Congenital Heart Disease in Newborns. Adv Pediatr. 2015; 62(1): 211–226. PubMed Abstract | Publisher Full Text\n\nPeterson C, Ailes E, Riehle-Colarusso T, et al.: Late detection of critical congenital heart disease among US infants: estimation of the potential impact of proposed universal screening using pulse oximetry. JAMA Pediatr. 2014; 168(4): 361–370. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMahle WT, Newburger JW, Matherne GP, et al.: Role of pulse oximetry in examining newborns for congenital heart disease: a scientific statement from the American Heart Association and American Academy of Pediatrics. Circulation. 2009; 120(5): 447–458. PubMed Abstract | Publisher Full Text\n\nGrosse SD, Peterson C, Abouk R, et al.: Cost and Cost-Effectiveness Assessments of Newborn Screening for Critical Congenital Heart Disease Using Pulse Oximetry: A Review. Int J Neonatal Screen. 2017; 3(4): 34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGreenberg D, Mohamed Ibrahim MIB, Boncz I: What Are the Challenges in Conducting Cost-of-Illness Studies? Value Health Reg Issues. 2014; 4: 115–116. PubMed Abstract | Publisher Full Text\n\nByford S, Torgerson DJ, Raftery J: Economic note: cost of illness studies. BMJ. 2000; 320(7245): 1335. PubMed Abstract | Publisher Full Text | Free Full Text\n\nInstitute for Quality and Efficiency in Health Care: Working paper cost Estimation. 2009. Reference Source\n\nJo C: Cost-of-illness studies: concepts, scopes, and methods. Clin Mol Hepatol. 2014; 20(4): 327–337. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTarricone R: Cost-of-illness analysis. What room in health economics? Health Policy. 2006; 77(1): 51–63. PubMed Abstract | Publisher Full Text\n\nGarcía A, Moreno K, Ronderos M, et al.: Differences by Altitude in the Frequency of Congenital Heart Defects in Colombia. Pediatr Cardiol. 2016; 37(8): 1507–1515. PubMed Abstract | Publisher Full Text\n\nGarcía A, Caicedo M, Moreno K, et al.: Diferencias regionales en cardiopatías congénitas. Revista Colombiana de Cardiología. 2017; 24(2): 161–168. Publisher Full Text\n\nPeterson C, Grosse SD, Oster ME, et al.: Cost-effectiveness of routine screening for critical congenital heart disease in US newborns. Pediatrics. 2013; 132(3): e595–603. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCONSEJO DIRECTIVO DEL INSTITUTO DE SEGUROS SOCIALES: ACUERDO No. 256 DE 2.001. EL CONSEJO DIRECTIVO DEL INSTITUTO DE SEGUROS SOCIALES; 2001; [cited 2019 Jan 4]. Reference Source\n\nTasa de cambio del peso colombiano (TRM): Banco de la República (banco central de Colombia). 2012; [cited 2018 Apr 30]. Reference Source\n\nPreguntas Frecuentes. ¿Qué diferencia existe entre cuota moderadora,.... [cited 2018 Oct 15]. Reference Source\n\nSalarios | Banco de la República (banco central de Colombia). [cited 2019 Jan 4]. Reference Source\n\nDepartamento Administrativo Nacional de Estadística. Preguntas frecuentes estratificación.\n\nConnor JA, Kline NE, Mott S, et al.: The meaning of cost for families of children with congenital heart disease. J Pediatr Health Care. 2010; 24(5): 318–325. PubMed Abstract | Publisher Full Text\n\nMughal AR, Sadiq M, Hyder SN, et al.: Socioeconomic status and impact of treatment on families of children with congenital heart disease. J Coll Physicians Surg Pak. 2011; 21(7): 398–402. PubMed Abstract\n\nRaj M, Paul M, Sudhakar A, et al.: Micro-economic impact of congenital heart surgery: results of a prospective study from a limited-resource setting. PLoS One. 2015; 10(6): e0131348. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChamorro C, Sandoval N, Taborda A, et al.: Underlying data of economic impact of CCHD in Colombia. OSF. 2019; [cited 2019 Jan 12]. http://www.doi.org/10.17605/OSF.IO/9J8KT\n\nChamorro C, Sandoval N, Taborda A, et al.: Survey asked to 20 caregivers of CCHD in Colombia. OSF. 2019; [cited 2019 Jan 13]. http://www.doi.org/10.17605/OSF.IO/67S4G"
}
|
[
{
"id": "43544",
"date": "11 Feb 2019",
"name": "Scott D. Grosse",
"expertise": [
"Reviewer Expertise Health economics",
"public health",
"newborn screening",
"birth defects"
],
"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 summarizes clinic-based medical cost estimates during the first year of life for 73 Colombian infants with one of seven specified CCHDs and also provides survey-based estimates of out-of-pocket medical and non-medical costs and productivity costs for 20 families.\n\nThe following text should be revised for clarity: “The human capital approach and the friction cost approach estimate indirect costs. The loss of productivity is associated with the friction period, i.e., the time the patient takes to be replaced by another employee and for production to return to its former level12.” It should be made clear that those are two alternative approaches that can be used to estimate indirect or productivity costs and that the two methods typically yield very different estimates. The second sentence refers specifically to the friction cost approach. Most (>90%) COI studies use the human capital approach, which typically yields substantially larger estimates of productivity costs since there is no time limit. Pike and Grosse (2018)1 summarize the differences between those two approaches.\n\nThe study should not be described as following a micro-costing approach. A micro-costing or ingredients costing study assesses labor hours, consumables and supplies, depreciated equipment, floor space, utilities, et al. and the costs of each to calculate the cost of providing a service. Peterson et al. (2014)2 used micro-costing to assess the costs of CCHD screening in a sample of hospitals in New Jersey, USA. The present paper used information on accounting costs, which is the approach typically used to estimate costs of health care services.\n\nThe authors are to be commended for seeking to undertake a COI study from the societal perspective, which includes the lost productivity of parents who take leave from employment to help their children obtain medical care. However, only a subset of societal costs have been estimated. First, the study did not include lost productivity associated with caring for a child with CCHD outside of medical encounters. Second, only considering time taken away from paid employment understates the loss of household services displaced by the need to provide care for a sick child. That approach is valid, but it is a partial estimate of the cost of informal care giving. Third, another limitation is the omission of the largest productivity cost using the human capital approach, namely the lost productivity associated with preventable deaths. Fourth, another limitation is that the analysis did not include estimates of costs for governmental programs such as public education. It is well-established that children with CCHDs are much more likely to require special education services. For example, see Oster et al. (2017)3.\n\nThe analysis of out-of-pocket costs likely double-counted medical costs by including both the charge and the amount paid by families, which is part of the overall charge. To avoid double-counting for the societal perspective COI analysis, medical expenses should be subtracted from the out-of-pocket costs reported. It is valid to include those costs in an analysis from the family perspective, in which case the clinic costs would not be included, thereby eliminating any potential double-counting.\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? Yes",
"responses": []
},
{
"id": "44300",
"date": "26 Feb 2019",
"name": "Leonardo Arregocés",
"expertise": [
"Reviewer Expertise Health economics",
"particularly in economic evaluation methods."
],
"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 lack of Cost of Illness studies for Colombia makes this study particularly important. The societal perspective makes it even more valuable.\nThe paper has a good structure that covers briefly but clearly the basics on how costs are estimated. Although I would suggest having another read of the whole text and probably get it checked by an editor would help correct some usage errors that change the meaning of the text. For example, differentiated is a more mathematical term. Disaggregated could be a better word.\nThe first introductory paragraph mentions the prevalence CCHD with late diagnosis corresponds to a USA statistic which might be important to mention. It is likely that in a developing country could be different, probably higher.\nThe methods section could benefit from more clarity, particularly on the assumptions and decisions made. I am sure the study team had good reasons in making their decisions but is highly recommended that those reasons are clearly stated. For example, it is said that statistical tests were not performed without an explanation of why this was decided. Later in the text, seven CCHD conditions were selected based on a cost-effectiveness of screening for CCHD in the USA but is not stated why is this also considered a good selection for Colombia. The time horizon used seems quite short, particularly because the criteria to select this period are not presented. Like other types of economic analyses in health, a clear description of the time horizon and the reasons for selecting a specific value should be included.\nHospital costs were estimated from observed data which is a valuable source. It is mentioned that observed use of services was validated by experts. There is no description of criteria used to categorise a person as an expert nor the methods used for the validation. Observed data may be considered data of higher quality relative to experts’ opinion. It is not clear how discrepancies between both were solved. Moreover, the way the base case was constructed is poorly described. Moreover, a description of the contents and quantities of services included in the base cases is not presented.\nThe sources of prices are a Tariff Manual used by an already extinct institution. It is not clear why the 2001 tariff is inflated for a specific value. This might be important because not all contracts among insurers and providers are negotiated using these prices. Furthermore, it is not clear that prices in this tariff manual are really reflecting the costs of production of health services.\nOut of pocket expenditure is a big added value of the present study. The review of the survey used might have some problems, but I am not sure if this is because key questions got lost in translation. The interview is available in English. Probably the questions are well formulated in Spanish, although in English looks clunky. Despite this issue, I think is likely that monthly income might not be correctly estimated. Did the households income was estimated by adding the reported incomes? Routine household surveys could be used as an example of how income is estimated. It is not clear how the weighing of out-of-pocket expenses was made.\nRegarding the results I would suggest presenting a weighted average for each condition instead of a simple average. I believe is important to describe what the non-surgical events include. Figure 1 describes the relation between reported income and indirect costs. Although the graph selected might not be the best option it presents interesting data. A chart or table with like this one relating out-of-pocket expenditure to the household income would be valuable.\nIt was mentioned in the introduction that some of the expenses were subsidised by the FCI foundation. That assertion makes necessary to present how much of the out-of-pocket expenditure was subsidised and to whom. The economic burden of these conditions seems to be extraordinarily high, probably sinking more into poverty families with children affected with CCHD. Is possible that subsidies received by the poorest families kept them in a better economic situation.\nThe discussion should reflect on some important aspects such as the short time horizon selected, the recall bias on the survey, the impact on the estimates of the differences in data sources or the lack of other statistical analyses. Also, what I consider the most valuable contribution of this study is the estimation of out-of-pocket expenditure, which the authors did not reflected upon in the discussion. This COI study finds an extremely high economic burden in the first year, which way above the 10% usually used as a threshold for driving families into poverty. A reflection on the financial protection that the Colombian Health System is providing would be interesting.\nFinally, a reflection on how the producing estimates from a single source might affect the precision of the estimates. If the results presented here are to be used by decision-makers, what does an estimate from a single source that is a private provider in the capital city of the country might affect the estimates. Is it possible that costs in other cities might differ?\nUnfortunately the methodological description is too shot and not entirely described. It is not possible to reproduce the results because the lack of descriptions in the methods.\n\nMy congratulations to the research team for a fantastic effort and looking forward to their discussion and considerations of the suggestions presented 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? 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/8-92
|
https://f1000research.com/articles/7-1784/v1
|
12 Nov 18
|
{
"type": "Research Article",
"title": "Key Concepts for assessing claims about treatment effects and making well-informed treatment choices",
"authors": [
"Andrew D. Oxman",
"Iain Chalmers",
"Astrid Austvoll-Dahlgren",
"Informed Health Choices group",
"Iain Chalmers",
"Astrid Austvoll-Dahlgren"
],
"abstract": "Background: The Informed Health Choices (IHC) Key Concepts are standards for judgement, or principles for evaluating the trustworthiness of treatment claims and treatment comparisons (evidence) used to support claims, and for making treatment choices. The list of concepts provides a framework, or starting point, for teachers, journalists and other intermediaries for identifying and developing resources (such as longer explanations, examples, games and interactive applications) to help people to understand and apply the concepts. The first version of the list was published in 2015 and has been updated yearly since then. We report here the changes that have been made from when the list was first published up to the current (2018) version. Methods: We developed the IHC Key Concepts by searching the literature and checklists written for the public, journalists, and health professionals; and by considering concepts related to assessing the certainty of evidence about the effects of treatments. We have revised the Key Concepts yearly, based on feedback and suggestions; and learning from using the IHC Key Concepts, other relevant frameworks, and adaptation of the IHC Key Concepts to other types of interventions besides treatments. Results: We have made many changes since the Key Concepts were first published in 2015. There are now 44 Key Concepts compared to the original 32; the concepts have been reorganised from six to three groups; we have added higher-level concepts in each of those groups; we have added short titles; and we have made changes to many of the concepts. Conclusions: We will continue to revise the IHC Key Concepts in response to feedback. Although we and others have found them helpful since they were first published, we anticipate that there are still ways in which they can be further improved. We welcome suggestions for how to do this.",
"keywords": [
"concepts",
"critical thinking",
"critical appraisal",
"causal inference",
"treatment claims",
"informed decision making",
"epistemology"
],
"content": "Background\n\nYou cannot make informed decisions without information. For decisions about actions to improve or maintain the health of individuals or communities (‘treatments’) to be well-informed and not misinformed, you need reliable information about the effects of treatments. Unfortunately, we are bombarded with claims about the benefits and harms of treatments, many of which are not reliable. Therefore people need to learn how to distinguish reliable from unreliable claims.\n\nUnreliable claims about the benefits and harms of treatments are made in the mass media and social media, as well as in personal communications with family, friends, quacks, and health professionals1–10. They are made by governments, celebrities, journalists, advertisers, researchers, gurus, aunts, and uncles. They include claims about medicines, surgery and other types of “modern medicine”; lifestyle changes, such as changes in what you eat or how you exercise; herbal remedies and other types of “traditional” or “alternative medicine”; public health and environmental interventions; and changes in how healthcare is financed, delivered, and governed.\n\nMany, if not most people are unable to assess the reliability of these claims. For example, in a survey of a random sample of Norwegian adults, we found that less than 20% of respondents recognized that lung cancer can be associated with drinking alcohol but not necessarily caused by it11. This mirrors misleading claims that are commonly made in the media. For instance, stories about coffee frequently use language suggesting that cause and effect has been established, such as “coffee can kill you”, when reporting on associations that have been found between drinking coffee and various health outcomes12. Personal experiences (anecdotes) are often used as a basis for treatment claims, and people are more likely to trust anecdotes than research. For example, surveys in the UK have shown that only about one third (37%) of the public trust evidence from medical research, while about two thirds (65%) trust the experiences of friends and family13. In addition, anecdotes often exaggerate the alleged benefits of treatments (for cancer, for example) and ignore or downplay harms14. At the same time, people in need or desperation hope that treatments will work and ignore potential harms.\n\nConsequences of people’s inability to assess the reliability of treatment claims include overuse of ineffective and sometimes harmful treatments and underuse of effective treatments, both of which result in unnecessary suffering and waste15,16. For example, billions of dollars are wasted on alternative medicine and nutritional supplements for which there is no reliable evidence of benefits17,18. At the same time, millions of children die unnecessarily, in part because their parents do not seek and use effective treatments that are available to them19,20, and they don’t trust reliable claims about effective preventive treatments such as effective vaccines21.\n\nTo address this problem, the Informed Health Choices (IHC) group is developing and evaluating resources to help people learn how to assess the trustworthiness of treatment claims and make well-informed decisions about treatments22,23. The first step in this work was to identify the key concepts that people need to understand and apply to do this24,25. We refer to these as the IHC Key Concepts. We review and update this list of concepts yearly. In this article we report the changes that we have made to the IHC Key Concepts since they were first published24 and present the most recent (2018) version.\n\n\nMethods\n\nThe IHC Key Concepts are standards for judgment, or principles for evaluating the trustworthiness of treatment claims and treatment comparisons (research) used to support claims, and for making treatment choices. The list is intended to be relevant to people everywhere and to any type of treatment. Many of the concepts can be learned and used successfully by primary school children22,26,27. Although we have developed and framed the Key Concepts to address treatment claims, people in other fields have also found them relevant. Work to adapt these concepts to apply to interventions in other fields is ongoing, including agricultural, economic, educational, environmental, international development, management, nutrition, policing, social welfare, and veterinary interventions.\n\nThe IHC Key Concepts are a starting point for developing learning resources to help people make judgements about the trustworthiness of claims about the effects of treatments (and other interventions), and to make well-informed decisions about treatments. They are also the basis for a database of multiple-choice questions that can be used to assess people’s abilities to assess treatment claims and make treatment choices28. We have written the concepts and explanations in plain language. However, some of them may be unfamiliar and difficult to understand. The Key Concepts list is not intended to be a learning resource. It is a framework that can be used by teachers and others to identify and develop learning resources.\n\nTo develop the IHC Key Concepts, we first extracted all of the concepts addressed in Testing Treatments29, a book that was written to promote more critical public assessment of claims about the effects of treatments. We then searched the literature for other relevant material, including books and checklists for the public, journalists, and health professionals24. We also considered concepts related to making judgements about the certainty of evidence of the effects of treatments30.\n\nOur aim has been to include all concepts that are important for people to consider. At the same time, we have tried to minimise redundancy. We have organised the concepts in a way that we believe is logical, and we have sought feedback on this logic. The concepts are not organised based on how complex or difficult they are to understand and apply, or in the order in which they should be taught.\n\nWe have collected structured written feedback on the Key Concepts using a form with four questions (Box 1). We initially obtained feedback from 29 members of an international advisory group24. We have subsequently obtained responses to these questions at three workshops:\n\nGlobal Evidence Summit, Cape Town, South Africa, 14 September 2017\n\nEvidence Live, Oxford, UK, 20 June 2018\n\n25th Cochrane Colloquium, Edinburgh, UK, 17 September 2018\n\n1. Are concepts included that should not be?\n\n2. Are there important concepts that are missing?\n\n3. Are the concepts organised in a logical way?\n\n4. Do you have any other comments regarding the concepts?\n\nIn addition, we have sought feedback and suggestions from colleagues when we have presented the Key Concepts, and on our website. The Key Concepts are updated yearly, and once or twice each year the three authors review and discuss each new suggestion and feedback from workshops, and we reach a consensus on which, if any, changes to make to the Key Concepts. For each suggestion, we record our response and the rationale for it. We invite comments on planned revisions from the IHC group and others prior to finalising each update.\n\nThree other sources of input have contributed to changes that we have made to the IHC Key Concepts. First, experience from developing learning resources and teaching has led to changes. For example, development of primary school resources31 led to reorganising the concepts into three groups from the original six groups24.\n\nSecond, we are reviewing related frameworks for critical thinking32, including frameworks for teaching and learning critical thinking33–37; scientific reasoning, literacy, and thinking38–41; epistemic cognition42; causal inference43, problem solving44, and meta-cognition45; health literacy46–48; and evidence-informed decision making and evidence-based practice49–51. In addition to ideas for new concepts, this review has contributed to the development of lists of competences (required skills, knowledge, or capacity to do something) and dispositions (frequent and voluntary habits of thinking and doing) for thinking critically about treatments. We added these to the IHC Key Concept list in 2018.\n\nThird, adaptation of the IHC Key Concepts to claims and decisions about other types of interventions (such as educational, economic, and environmental interventions) has contributed to changes that we have made, including the decision to reorganise the Key Concept list in 2018.\n\n\nResults\n\nThe 2018 version of the IHC Key Concepts is the most recent version. It can be found as Supplementary File 1 and online52. Before reporting the changes that we made in this version and the reasons for those changes, we summarise the changes that we made to the IHC Key Concepts in 2016 and 2017.\n\nThe first version of the IHC Key Concepts, published in 201524, included 32 concepts in the following six groups:\n\nRecognising the need for fair comparisons of treatments\n\nJudging whether a comparison of treatments is a fair comparison\n\nUnderstanding the role of chance\n\nConsidering all the relevant fair comparisons\n\nUnderstanding the results of fair comparisons of treatments\n\nJudging whether fair comparisons of treatments are relevant\n\nIn 201653, we added two new concepts and reorganised the concepts into three groups. The two new concepts were:\n\nUnpublished results of fair comparisons may result in biased estimates of treatment effects.\n\nA lack of evidence is not the same as evidence of “no difference”.\n\nThe decision to reorganise the concepts into three groups grew out of our efforts to simplify the concepts and teach them to primary school children. The suggestion to use three groups - claims, comparisons, and choices - came from Matt Oxman, who had primary responsibility for writing the text for The Health Choices Book for primary school children54. The book, which has been shown to be an effective learning resource in a randomised trial with over 10,000 children in Uganda, is a story in comic book format which introduces and explains 12 Key Concepts.\n\nIn 201755, we added short titles for all the concepts and two new concepts:\n\nPeer-reviewed and published treatment comparisons may not be fair comparisons.\n\nComparisons designed to evaluate whether a treatment can work under ideal circumstances may not reflect what you can expect under usual circumstances.\n\nThe suggestion to add the short titles came from Douglas Badenoch, the project manager for the Testing Treatments websites54. The short titles were needed for the Critical thinking and Appraisal Resources Library (CARL) on the Testing Treatments - English website. CARL is a database of learning resources for teachers and others who are responsible for encouraging critical thinking about treatment claims56. It contains over 500 open-access learning resources in a variety of formats, including text, audio, video, webpages, cartoons, and lesson materials. Each resource is relevant to at least one IHC Key Concept and CARL can be searched or browsed using the Key Concepts.\n\nIn the 2018 version (Supplementary File 1), we merged two Key Concepts and added nine new concepts. We reorganised the concepts within each of the three main groups and added three subgroups to each of the three main groups of concepts. We also replaced all of the short titles and introduced emojis.\n\nWe removed the concept that “hope or fear can lead to unrealistic expectations about the effects of treatments” and incorporated this in the explanation of the concept “treatments may be harmful”. The explanation begins with “People often exaggerate the benefits of treatments and ignore or downplay potential harms.” We added: “Similarly, people in need or desperation hope that treatments will work and ignore potential harms.”\n\nThe nine new concepts were:\n\nWe can rarely, if ever, be 100% certain about the effects of treatments.\n\nPeople often recover from illness without treatment.\n\nMore data is not necessarily better data, whatever the source.\n\nIt is rarely possible to know in advance who will benefit, who will not, and who will be harmed by using a treatment.\n\nIndirect comparisons of treatments can be misleading.\n\nOutcomes should be assessed reliably in treatment comparisons.\n\nTreatment comparisons may be sensitive to assumptions that are made.\n\nVerbal descriptions of treatment effects can be misleading.\n\nThe problem and the treatment options being considered may not be the right ones.\n\nWe introduced three higher level concepts within each of the three groups of Key Concepts and reframed the titles of the three groups as shown in Box 2.\n\n1. Beware of treatment claims like these\n\n1.1 Beware of claims that are too good to be true.\n\n1.2 Beware of claims based on faulty logic.\n\n1.3 Beware of claims based on trust alone.\n\n2. Check the evidence from treatment comparisons\n\n2.1 Don’t be misled by unfair comparisons.\n\n2.2 Don’t be misled by unreliable summaries of treatment comparisons.\n\n2.3 Don’t be misled by how treatment effects are described.\n\n3. Make well-informed treatment choices\n\n3.1 What is the problem and what are the options?\n\n3.2 Is the evidence relevant?\n\n3.3 Do the advantages outweigh the disadvantages?\n\nWe did this in response to feedback that the organisation of concepts within the three main groups was not logical, and that having long lists of concepts was overwhelming. The subgroups of concepts, using these higher-level concepts, provides a more transparent logic for how the concepts are organised in each main group. Having just three higher level concepts for each group may also make it easier to get the gist of the concepts and make the list less overwhelming and easier to remember.\n\nThere were three reasons for changing the short titles used for each of the Key concepts. First, we had received feedback that the short titles were not consistent with some of the concepts and that some were not short; and it was difficult to come up with a short, catchy title that accurately reflected each concept. Second, we wanted short titles that were consistent with the new organisation of the concepts. Third, short titles that we were developing for posters and a website targeted at school children seemed to be a solution to this problem. We added emojis to make the poster and website that we are developing more appealing. When presenting these to colleagues and others, the emojis appeared to appeal across age groups and to reflect the content accurately, which also may help to convey the gist of the concepts. The full list of short titles for the Key Concepts and the emojis are shown in Box 3.\n\nIn addition to adding 13 new Key Concepts and removing one since the first version was published in 2015, and reorganising the concepts, we have modified several of them. Most of these changes have been in response to suggestions to add new concepts when we concluded that it made more sense to incorporate the suggestion in an existing concept. These changes are summarised in Table 1.\n\nIn addition to feedback from three workshops over the past two years, we have received 61 suggestions for revisions over the past three years. For many of these we concluded that no change was needed. Several suggestions were similar. We summarise these suggestions and our reasons for not making any changes in Table 2.\n\n\nDiscussion\n\nUp to now we have received much positive feedback, along with many suggestions for improvements, on the IHC Key Concepts, including positive feedback on the changes that we made in the 2018 version. Nonetheless, as can be seen from the results reported here, we have made many changes since the Key Concepts were first published in 2015. There are now 44 Key Concepts compared to the original 32; the concepts have been reorganised from six to three groups; we have added higher-level concepts within each of those groups; we have added short titles; and we have made changes to many of the concepts. We will continue to revise the IHC Key Concepts in response to feedback. Although we and others have found the concepts helpful since they were first published24, we anticipate that there will still be ways in which they can be further improved. We welcome suggestions on ways of doing this.\n\nThe most common misunderstanding in the feedback we have received is that the Key Concepts list is a learning resource intended for people with no relevant research background. As noted in the Methods section, the list of Key Concepts serves as the basis for developing learning resources. It is not designed as a learning resource. It is a framework, or starting point, for identifying and developing learning resources.\n\nAnother common misunderstanding is that the Key Concepts are organised in the order in which they should be taught or learned. We have organised the Key Concepts logically by grouping them first in three groups and then within those three groups using higher-level concepts (Box 2). This logic does not reflect the difficulty of the concepts or the order in which they should be learned.\n\nWhen teaching the concepts, it may make sense to start with ones in the first group, followed by ones in the second group, followed by ones in the third group. However, it does not necessarily make sense to teach them in that order or in the order that they are organised within each group. For example, at least 24 of the Key Concepts can be understood and applied by primary school children31, whereas other concepts are likely too difficult for primary school children to understand and use. Thus, it would obviously make sense to hop over those concepts when teaching primary school children.\n\nAlso, it is important not to try to teach or learn too much at one time. We initially tried teaching 24 Key Concepts to primary school children in one go, and found that was too much to teach in a single school term31. Our efforts to teach IHC Key Concepts to both primary school children and their parents support our initial hypothesis that the time to start learning these concepts is in primary school - if not even younger59. Ideally, these concepts should be taught and learned using a spiral curriculum60–62, that maps out what students should learn, where they should begin, and how they should progress to master these skills. Each cycle in a spiral curriculum reinforces what was learned previously while introducing new concepts. This can help teachers and students identify when milestones have been reached, build a foundation for later stages of learning, and guide the development of assessment tools and learning resources. We have not yet developed a spiral curriculum based on the IHC Key Concepts.\n\nDecisions about the suggestions we have received have been based on logic and discussion. Four criteria have emerged from these discussions, which we will use explicitly in further developing the IHC Key Concepts. New Key Concepts have to:\n\nbe within the scope of the IHC Key Concepts - standards for judgment, or principles for evaluating the trustworthiness of treatment claims and treatment comparisons (research) used to support claims, and to inform treatment choices\n\naddress ways in which treatment claims and comparisons are frequently misleading or ways in which poorly informed decisions are taken\n\nbe useful for people without a research background to use research, not just for researchers or for doing research\n\noverlap as little as possible with other Key Concepts\n\nIn addition to continuing to seek and review feedback and suggestions, we will further develop the Key Concepts by continuing to learn from using the IHC Key Concepts, other relevant frameworks, and adaptation of the IHC Key Concepts to other types of interventions. We also plan to summarise the evidence supporting each of the Key Concepts.\n\n\nConclusions\n\nThe IHC Key Concepts have proven useful in designing learning resources, evaluating them, and organising them25. The most recent version of the Key Concepts improves on previous versions by incorporating additional Key Concepts, organising the Key Concepts more logically and, we believe, making it easier to get the gist of the Key Concepts. Future improvements will be made based on feedback and suggestions, and ongoing evaluation.\n\n\nData availability\n\nDataset 1: Suggested revisions to the IHC Key Concepts and responses 2016-2018 https://dx.doi.org/10.5256/f1000research.16771.d22353263",
"appendix": "Grant information\n\nIC receives support through National Institute for Health Research funding for the James Lind Initiative. The IHC Key Concepts were developed as part of the Informed Health Choices Project, which was funded by the Research Council of Norway [220603/H10].\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nAllen Nsangi, Claire Glenton, Simon Lewin, Angela Morelli, Sarah Rosenbaum, Daniel Semakula, and Nelson Sewankambo were co-authors of the first version of the IHC Key Concepts. They and other members of the Informed Health Choices group also contributed to the subsequent development of the IHC Key Concepts. We are grateful to all of the people who have provided feedback and suggestions for improving the IHC Key Concepts.\n\n\nSupplementary material\n\nSupplementary File 1: Informed Health Choices (IHC) Key Concepts 2018\n\nClick here to access the data\n\n\nReferences\n\nWang MTM, Grey A, Bolland MJ: Conflicts of interest and expertise of independent commenters in news stories about medical research. CMAJ. 2017; 189(15): E553–E559. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWalsh-Childers K, Braddock J, Rabaza C, et al.: One step forward, one step back: changes in news coverage of medical interventions. Health Commun. 2016; 16: 1–14.\n\nSumner P, Vivian-Griffiths S, Bolvin J, et al.: Exaggerations and caveats in press releases and health-related science news. PLoS One. 2016; 11(12): e0168217. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchwitzer G: A guide to reading health care news stories. JAMA Intern Med. 2014; 174(7): 1183–6. PubMed Abstract | Publisher Full Text\n\nMoorhead SA, Hazlet DE, Harrison L, et al.: A new dimension of health care: systematic review of the uses, benefits, and limitations of social media for health communication. J Med Internet Res. 2013; 15(4): e85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchwartz LM, Woloshin S, Andrews A, et al.: Influence of medical journal press releases on the quality of associated newspaper coverage: retrospective cohort study. BMJ. 2012; 344: d8164. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGlenton C, Paulsen E, Oxman AD: Portals to Wonderland: health portals lead to confusing information about the effects of health care. BMC Med Inform Decis Mak. 2005; 5: 7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoynihan R, Bero L, Ross-Degnan D, et al.: Coverage by the news media of the benefits and risks of medications. N Engl J Med. 2000; 342(22): 1645–50. PubMed Abstract | Publisher Full Text\n\nCoulter A, Entwistle V, Gilbert D: Sharing decisions with patients: is the information good enough? BMJ. 1999; 318(7179): 318–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSansgiry S, Sharp WT, Sansgiry SS: Accuracy of information on printed over-the-counter drug advertisements. Health Mark Q. 1999; 17(2): 7–18. PubMed Abstract | Publisher Full Text\n\nOxman AD, Austvoll-Dahlgren A, Garratt A, et al.: Understanding of key concepts relevant to assessing claims about treatment effects: a survey of Norwegian adults. IHC Working Paper.2017; Accessed October 14, 2018. Reference Source\n\nSchwitzer G: A guide to reading health care news stories. JAMA Intern Med. 2014; 174(7): 1183–6. PubMed Abstract | Publisher Full Text\n\nAcademy of Medical Sciences: Enhancing the use of scientific evidence to judge the potential benefits and harms of medicines. London: Academy of Medical Sciences, 2017; Accessed October 14, 218. Reference Source\n\nFishman J, Ten Have T, Casarett D: Cancer and the media: how does the news report on treatment and outcomes? Arch Intern Med. 2010; 170(6): 515–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBrownlee S, Chalkidou K, Doust J, et al.: Evidence for overuse of medical services around the world. Lancet. 2017; 390(10090): 156–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGlasziou P, Straus S, Brownlee S, et al.: Evidence for underuse of effective medical services around the world. Lancet. 2017; 390(10090): 169–77. PubMed Abstract | Publisher Full Text\n\nFrass M, Strassl RP, Friehs H, et al.: Use and acceptance of complementary and alternative medicine among the general population and medical personnel: a systematic review. Ochsner J. 2012; 12(1): 45–56. PubMed Abstract | Free Full Text\n\nStarr RR: Too little, too late: ineffective regulation of dietary supplements in the United States. Am J Public Health. 2015; 105(3): 478–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJones G, Steketee RW, Black RE, et al.: How many child deaths can we prevent this year? Lancet. 2003; 362(9377): 65–71. PubMed Abstract | Publisher Full Text\n\nPierce H, Gibby AL, Forste R: Caregiver Decision-Making: Household Response to Child Illness in sub-Saharan Africa. Popul Res Policy Rev. 2016; 35(5): 581–97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuñoz DC, Llamas LM, Bosch-Capblanch X: Exposing concerns about vaccination in low- and middle-income countries: a systematic review. Int J Pub Health. 2015; 60(7): 767–80. PubMed Abstract | Publisher Full Text\n\nNsangi A, Semakula D, Oxman AD, et al.: Effects of the Informed Health Choices primary school intervention on the ability of children in Uganda to assess the reliability of claims about treatment effects: a cluster-randomised controlled trial. Lancet. 2017; 390(10092): 374–88. PubMed Abstract | Publisher Full Text\n\nSemakula D, Nsangi A, Oxman AD, et al.: Effects of the Informed Health Choices podcast on the ability of parents of primary school children in Uganda to assess claims about treatment effects: a randomised controlled trial. Lancet. 2017; 390(10092): 389–98. PubMed Abstract | Publisher Full Text\n\nAustvoll-Dahlgren A, Oxman AD, Chalmers I, et al.: Key concepts that people need to understand to assess claims about treatment effects. J Evid Based Med. 2015; 8(3): 112–25. PubMed Abstract | Publisher Full Text\n\nChalmers I, Oxman AD, Austvoll-Dahlgren A, et al.: Key Concepts for Informed Health Choices: a framework for helping people learn how to assess treatment claims and make informed choices. BMJ Evid Based Med. 2018; 23(1): 29–33. PubMed Abstract | Publisher Full Text\n\nNsangi A, Semakula D, Oxman AD, et al.: One year follow-up of the effects of the Informed Health Choices primary school intervention on the ability of children in Uganda to assess the reliability of claims about treatment effects: a cluster-randomised trial. Submitted.\n\nNsangi A, Semakula D, Glenton C, et al.: Resources to teach primary school children in low-income countries to assess claims about treatment effects: process evaluation. Submitted.\n\nAustvoll-Dahlgren A, Semakula D, Nsangi A, et al.: Measuring ability to assess claims about treatment effects: the development of the 'Claim Evaluation Tools'. BMJ Open. 2017; 7(5): e013184. PubMed Abstract | Free Full Text\n\nEvans I, Thornton H, Chalmers I, et al.: Testing Treatments, 2nd Edition London: Pinter and Martin, 2011. Reference Source\n\nGuyatt GH, Oxman AD, Vist GE, et al.: GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008; 336(7650): 924–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNsangi A, Semakula D, Oxman AD, et al.: Development of the Informed Health Choices resources to teach primary school children to assess claims about treatment effects in four countries. Submitted.\n\nOxman AD, Martinez GL: Comparison of the Informed Health Choices Key Concepts to other frameworks that are relevant to learning how to think critically about treatment claims, comparisons, and choices: protocol for a mapping review. IHC Working Paper, 2018; Accessed October 24, 2018. Reference Source\n\nPresseisen BZ: Critical thinking and thinking skills: State of the art definitions and practice in public schools. Philadelphia: Research for Better Schools, Inc., 1986. Reference Source\n\nFollman J: Critical thinking definitions. Inquiry. 1991; 8: 4–5.\n\nMoseley D, Baumfield V, Elliott J, et al.: Frameworks for Thinking, A Handbook for Teaching and Learning. Cambridge: Cambridge University Press 2005. Publisher Full Text\n\nMoore T: Critical thinking: seven definitions in search of a concept. Stud High Educ. 2013; 38(4): 506–22. Publisher Full Text\n\nGyenes A: Definitions of critical thinking in context. Ann Educ Stud. 2015; 20: 17–25.\n\nKind P, Osborne J: Styles of scientific reasoning: a cultural rationale for science education? Sci Educ. 2017; 101(1): 8–31. Publisher Full Text\n\nZimmerman C: The development of scientific reasoning skills. Dev Rev. 2000; 20(1): 99–149. Publisher Full Text\n\nLaugksch RC: Scientific literacy: a conceptual overview. Sci Ed. 2000; 84(1): 71–94. Publisher Full Text\n\nMiller JD: Scientific literacy: a conceptual and empirical review. Daedalus. 1983; 112(2): 29–48. Reference Source\n\nBendixen LD: Teaching for epistemic change in elementary classrooms. In: Bråten I, Sandoval WA, Greene JA, eds. Handbook of Epistemic Cognition. New York: Routledge, 2016. Reference Source\n\nGlass TA, Goodman SN, Hernán MA, et al.: Causal inference in public health. Annu Rev Publ Health. 2013; 34: 61–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmith MUed: Toward a Unified Theory of Problem Solving: Views from the Content Domains. Hillsdale, NJ: Lawrence Erlbaum, Associates Inc., 1991. Reference Source\n\nGascoine L, Higgins S, Wall K: The assessment of metacognition in children aged 4-16 years: a systematic review. Rev Educ. 2017; 5(1): 3–57. Publisher Full Text\n\nBröder J, Okan O, Bauer U, et al.: Health literacy in childhood and youth: a systematic review of definitions and models. BMC Public Health. 2017; 17(1): 361. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSørensen K, Van den Broucke S, Fullam J, et al.: Health literacy and public health: a systematic review and integration of definitions and models. BMC Publ Health. 2012; 12: 80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSykes S, Wills J, Rowlands G, et al.: Understanding critical health literacy: a concept analysis. BMC Publ Health. 2013; 13: 150. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMorgan RL, Kelley L, Guyatt GH, et al.: Decision-making frameworks and considerations for informing coverage decisions for healthcare interventions: a critical interpretive synthesis. J Clin Epidemiol. 2018; 94: 143–150. PubMed Abstract | Publisher Full Text\n\nAlonso-Coello P, Schünemann HJ, Moberg J, et al.: GRADE Evidence to Decision (EtD) frameworks: a systematic and transparent approach to making well informed healthcare choices. 1: Introduction. BMJ. 2016; 353: i2016. PubMed Abstract | Publisher Full Text\n\nAlbarquoni L, Hoffmann T, Straus S, et al.: Core competencies in evidence-based practice for health professionals: consensus statement based on a systematic review and Delphi survey. JAMA Netw Open. 2018; 1(2): e180281. Publisher Full Text\n\nOxman AD, Chalmers I, Austvoll-Dahlgren A, et al.: Key Concepts for assessing claims about treatment effects and making well-informed treatment choices. Version: 2018. Accessed October 14, 2018. Reference Source\n\nAustvoll-Dahlgren A, Oxman AD, Chalmers I, et al.: Key Concepts for assessing claims about treatment effects and making well-informed treatment choices. Version: 2016. Accessed October 14, 2018. Reference Source\n\nThe Informed Healthcare Choices Group: Teachers’ Guide for The Health Choices Book: Learning to think carefully when making choices about treatments. A health science book. Oslo: Norwegian Institute of Public Health; 2016; Accessed October 14, 2018. Reference Source\n\nAustvoll-Dahlgren A, Chalmiers I, Oxman AD, et al.: Assessing claims about treatment effects: key concepts that people need to understand. Version: 2017. Accessed October 14, 2018. Reference Source\n\nCastle J, Chalmers I, Atkinson P, et al.: Establishing a library of resources to help people understand key concepts in assessing treatment claims-The \"Critical thinking and Appraisal Resource Library\" (CARL). PLoS One. 2017; 12(7): e0178666. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAkl EA, Oxman AD, Herrin J, et al.: Using alternative statistical formats for presenting risks and risk reductions. Cochrane Database Syst Rev. 2011; (3): CD006776. PubMed Abstract | Publisher Full Text\n\nAkl EA, Oxman AD, Herrin J, et al.: Framing of health information messages. Cochrane Database Syst Rev. 2011; (12): CD006777. PubMed Abstract | Publisher Full Text\n\nSemakula D, Nsangi A, Oxman AD, et al.: Effects of the Informed Health Choices podcast on the ability of parents of primary school children in Uganda to assess the trustworthiness of claims about treatment effects, one-year follow-up: a randomised trial. Submitted.\n\nBruner JS: The Process of Education. Cambridge, MA: Harvard University Press, 1960. Reference Source\n\nHarden RM: What is a spiral curriculum? Med Teach. 1999; 21(2): 141–3. PubMed Abstract | Publisher Full Text\n\nMurray JW: Skills development, habits of mind, and the spiral curriculum: a dialectical approach to undergraduate general education curriculum mapping. Cogent Educ. 2016; 3(1): 1156807. Publisher Full Text\n\nOxman AD, Chalmers I, Austvoll-Dahlgren A: Dataset 1 in: Key Concepts for assessing claims about treatment effects and making well-informed treatment choices. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16771.d223532"
}
|
[
{
"id": "40591",
"date": "26 Nov 2018",
"name": "Catherine Mathews",
"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\nThis is a manuscript describing the refining of the key concepts which will ideally be used as a basis for developing interventions to help people discern between what are true and false healthcare claims. This manuscript describes the methods and rationale for revising the classification system of key concepts, and for revising the key concepts themselves. It also provides the \"results\": the revised list of key concepts with explanations and presentation format.\n\nThis is very important work: the IHC Key Concepts have been the foundation of education interventions that are being actively tested and implemented in several countries in the world, across rich and poor settings. It is an international initiative to promote critical thinking and evidence-based health care.\nThe manuscript describes excellent work, and it is written clearly and articulately.\n\nI have the following minor recommendations for improving the manuscript:\nThe authors write that “adaption of the IHC Key Concepts to claims and decisions about other types of interventions (such as educational, economic and environmental interventions), has contributed to the changes we have made.\" However, the authors have not clarified how the changes were informed by considering other types of interventions. I suggest they provide a brief explanation.\n\nOn page 3, the authors have written: “We initially obtained feedback from 29 members of an international advisory group”. I suggest they name the group in the manuscript, rather than only in the relevant reference.\n\nThe conclusion of the Abstract could be stronger, by referring to the overall purpose and potential impact of the revised key concepts.\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": "4369",
"date": "23 Jan 2019",
"name": "Andrew David Oxman",
"role": "Author Response",
"response": "Thank you. 1. We have clarified that it is feedback from those who have adapted the Key Concepts that has contributed to changes. 2. The advisory group did not have a formal name. 3. We edited the conclusion of the Abstract as suggested."
}
]
},
{
"id": "40589",
"date": "10 Dec 2018",
"name": "Richard Lehman",
"expertise": [
"Reviewer Expertise Shared Understanding of 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\nThis description of the development of Key Concepts is clear and comprehensive. It is a real pleasure to read such strong and well-chosen English in a biomedical paper. The authors' sensitivity to language and meaning is also clear from the thoughtful way they have handled suggestions for changes in the wording of Key Concepts themselves.\nI have no suggestions for changes in the text. The concepts themselves are of permanent value and their authors have given the world a model of openness and rigour. It would be good to hear about their dissemination plan as clearly they have been successful in many ways but deserve to reach the widest audience possible.\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": "4368",
"date": "23 Jan 2019",
"name": "Andrew David Oxman",
"role": "Author Response",
"response": "Thank you."
}
]
},
{
"id": "40587",
"date": "13 Dec 2018",
"name": "David Henry",
"expertise": [
"Reviewer Expertise evidence-based practice",
"clinical epidemiology",
"population data sceince"
],
"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\nA very useful list and teaching aid but make sure you read the book\nThe Informed Health Choices (IHC) group performs a critical role in creating and regularly updating a framework of concepts that should underpin the decisions that the public make about their health. We can attest to the importance of this initiative having adapted and used some of these concepts in a recent public engagement session on judging the legitimacy of health claims in Australia: (https://bond.edu.au/researchers/research-bond/research-week-2018/program-and-highlights). We strongly support the intention of this program and the sentiments that lie behind it. The education of school children is a particularly innovative and important aspect. The recent categorization of a lengthening list of concepts under Claims, Comparisons and Choices is helpful in aiding understanding of principles and placing them in context.\nRegarding the 2018 update, we raise two general questions about how the key concepts have developed:\n1. Did the authors use rigorous methods to develop the concepts? If not, does it matter?\n2. Is the overall tone too negative? Can healthy skepticism turn into cynicism?\n\n1. The report is billed as a research article, but the methods used by the authors to develop the IHC concepts are not described in detail. The original set of concepts was based on those identified in the book ‘Testing Treatments’, which itself was a compilation of concepts illustrated by historical examples compiled by the experienced authors. Although the book wasn’t written around a rigorous theoretical framework it is brilliant. It combines key messages (e.g. ‘new is not always better; ‘more is not always better’; ‘earlier is not always better’) with accurate and compelling examples. So, the concepts in the IHC list certainly have face validity. But should the authors have done more? For instance, there is overlap between several of the IHC concepts. Should they have used factor analysis to shorten the list and make it more usable? We think not. They were not designing an instrument to make valid and reliable measurements. Rather, the IHC concepts are a learning resource and their validity has been demonstrated in a randomized trial in children.\n\n2. Our concern about the IHC concepts is the generally negative tone. The framing language is always skeptical and to a degree this is understandable and hard to avoid. We are inundated with claims daily and cannot start from a position of equipoise in judging and acting upon those that are relevant to us. The book ‘Testing Treatments’ documents some of the most important advances in medical science, including vaccination, treatment of heart disease, treatment of HIV and some cancers. As the journalist Nick Ross said in a foreword to the first edition of the book “it warmly admires much of what modern medicine has achieved. Its ambitions are always to improve medical practice, not disparage it.” The IHC concept list as a stand alone document lacks this balance. The underlying tone could be interpreted (wrongly) as ‘nothing works’ and ‘be suspicious of all medical claims’. We believe the authors should look for opportunities to use positive framing for some concepts. The preamble to the list should acknowledge the massive progress made by modern medicine and public health, including large reductions in all-cause and some cause-specific mortality rates over the last 50 years. Much of this progress has been made in small increments that individually might not have seemed compelling, but in summation have been dramatic. In our view the main targets of the IHC concepts should be claims that are intentionally misleading, often made by those with vested interests. There are plenty to deal with.\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?\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": "4367",
"date": "23 Jan 2019",
"name": "Andrew David Oxman",
"role": "Author Response",
"response": "Thank you for these comments. 1. Did the authors use rigorous methods to develop the concepts? If not, does it matter? This is an interesting question, for which we do not believe there is a simple answer. First, we would like to clarify that the IHC Key Concepts provide a framework for developing learning resources, such as those that we tested in randomised trials in Uganda. They are not, by themselves, a learning resource. We believe that we have demonstrated the usefulness of the IHC Key Concepts as a framework for deciding what to teach, for designing learning resources, and for designing a measurement instrument. The results of the randomised trials and process evaluations of those resources suggest that many people have not learned some of these concepts; that they can be learned and used by children and their parents; that people value learning them; and that they help people to think critically about claims. The measurement instrument that we used in the trials was shown to be reliable and valid, and those studies provide some indirect support for both face validity and construct validity for the 13 included concepts (using Rasch analysis). The extent to which very similar concepts have been found to apply across a wide range of different types of interventions provides some additional validation. Second, it is unclear how “rigorous methods” should be conceptualised for developing a framework such as that drawing on the IHC Key Concepts. As noted in the Methods section, we are currently conducting a systematic review of related frameworks for critical thinking. We have not so far come across any standard methods for developing frameworks such as these. Our review considers several questions related to the development of frameworks for critical thinking. These include: Is there a clear description of the methods that were used? Is the basis for the framework clear? For example, was it based on another framework, a model or theory, a systematic review, an unsystematic review, a formal consensus process, or an informal consensus process? Are the criteria for including and excluding elements clear? We believe that we have provided sufficient information about what we have done up to now for others to be able to judge how rigorous or appropriate our methods have been. The basis for developing our framework up to now has been an unsystematic review and an iterative, informal consensus process. A systematic review and a formal consensus process might be considered to be more rigorous, and some might also consider starting with an explicit model or theory to be more rigorous. We are unaware of any evidence suggesting that any of those methods would result in a better framework - however one chooses to define ‘better’. None of the frameworks that we have reviewed so far have been formally evaluated. Some possible criteria for assessing how sensible a framework is are listed in Box 1. Those are questions that we put to our advisory group (as reported in our first report of the Key Concepts), and that we have continued to ask others, as reported in this article. For the most part, we have received positive feedback in response to those questions. However, we have received many suggestions for improvements, as reported in this article, and the framework has continued to improve. We could assess more rigorously than we have done up to now the extent to which the Key Concepts are sensible, and we will consider doing this in future. Something else that we plan to do more rigorously in future (as noted in the Discussion section) is to systematically summarise the evidence supporting each of the concepts. 2. Is the overall tone too negative? Can healthy skepticism turn into cynicism? We share this implied concern about the second group of concepts and we will likely reframe at least some of those in the next update. However, we believe the first group of concepts must be framed negatively. The last group of concepts are framed as questions. This same concern came up at a meeting in December at which people from different disciplines discussed the applicability of the IHC Key Concepts to interventions beyond those in health care. This included agricultural, economic, educational, environmental, development, health, informal learning, management, nutritional, planetary health, policing, social welfare, speech and language, and veterinary interventions. We arrived at a list of concepts that participants agreed are relevant to a wide range of interventions. We must reiterate, however, that the Key Concepts by themselves are not a learning resource. We believe that teachers and those developing teaching/learning resources must be aware of this concern, regardless of how the Key Concepts are framed, and that they must be careful to encourage their students to be healthy sceptics without being nihilists."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1784
|
https://f1000research.com/articles/8-84/v1
|
21 Jan 19
|
{
"type": "Research Note",
"title": "Cadmium SAD phasing at CuKα wavelength",
"authors": [
"Igor E. Eliseev",
"Anna N. Yudenko",
"Valeria M. Ukrainskaya",
"Oleg B. Chakchir",
"Anna N. Yudenko",
"Valeria M. Ukrainskaya",
"Oleg B. Chakchir"
],
"abstract": "Single-wavelength anomalous diffraction (SAD) is the most common method for de novo elucidation of macromolecular structures by X-ray crystallography. It requires an anomalous scatterer in a crystal to calculate phases. A recent study by Panneerselvam et al. emphasized the utility of cadmium ions for SAD phasing at the standard synchrotron wavelength of 1 Å. Here we show that cadmium is also useful for phasing of crystals collected in-house with CuKα radiation. Using a crystal of single-domain antibody as an experimental model, we demonstrate how cadmium SAD can be conveniently employed to solve a CuKα dataset. We then discuss the factors which make this method generally applicable.",
"keywords": [
"Protein crystallography",
"experimental phasing",
"single-wavelength anomalous diffraction",
"cadmium ions",
"Cd-SAD"
],
"content": "Introduction\n\nElucidation of atomic structures of macromolecules by X-ray crystallography requires knowledge of the phases of measured reflections. Nowadays this phase problem is most often solved by molecular replacement (MR), a computational technique which utilizes the known structure of a homologous molecule to estimate phases. However, in the case of de novo structure elucidation when an appropriate homologous structure is unavailable, phases should be determined experimentally. This is predominantly achieved by analyzing anomalous scattering produced either by atoms naturally occurring in the molecule, or intentionally introduced into crystal during growth or soaking. The two phasing methods exploiting the anomalous scattering, multiwavelength anomalous diffraction (MAD) and single-wavelength anomalous diffraction (SAD), were reviewed by Hendrickson1. Synchrotron radiation with tunable wavelength allows achieving the absorption edges of all elements with Z≥20 to maximize anomalous signal, thus making these methods remarkably versatile.\n\nOn the contrary, the choice of anomalous scatterer is minimal when data are to be collected in-house using a laboratory X-ray generator, most often equipped with a copper anode (λ=1.5418 Å, CuKα). Indeed, in some cases, even weak anomalous signal of sulfur (f′′=0.56e- at CuKα) can be used for phasing, as demonstrated in pioneering SAD work on crambin2. Similarly, zinc (f′′=0.68e- at CuKα) was proposed to be useful for in-house SAD experiments3. Perhaps the most impressive result came from the structural genomics project, where iodine ion soaks were systematically used for de novo SAD phasing of datasets collected with CuKα radiation4. Iodine has a strong anomalous scattering (f′′=6.9e- at CuKα), high solubility, and binds multiple hydrophobic sites or positively charged residues on protein surface. Iodine SAD appeared remarkably efficient for phasing the crystals of membrane proteins which possess patches of positively charged residues at the hydrophobic-hydrophilic interface, providing many binding sites for anions5.\n\nAnother attractive opportunity is to use cadmium ions, which have a great anomalous signal (f′′=4.7e- at CuKα) comparable to that of iodine, promote crystal growth6, and can substitute other divalent cations in metal-binding proteins. Despite all these advantages and its use in the very early SAD works7, Cd is rarely used in the phasing of protein crystals. Recently, a paper emphasizing the utility of cadmium ions for experimental phasing at the standard synchrotron wavelength of 1 Å was published8. In this short research note, we show how Cd-SAD can also be conveniently used for phasing datasets collected using CuKα radiation.\n\n\nMethods\n\nAs an experimental model for in-house cadmium SAD, we used a crystal of an anti-ErbB3 single-domain antibody BCD090-M2, which we recently studied9. The details of protein purification, characterization, and structural analysis are given in the paper9. Briefly, the protein was expressed in E. coli SHuffle cells as a SUMO fusion, purified by immobilized metal affinity chromatography, cleaved by TEV protease, and then polished by an additional step of high-resolution cation-exchange chromatography. The antibody was crystallized by hanging-drop vapor diffusion in two different forms: in a space group C2 without divalent cations (PDB accession number: 6EZW) and in P1 with two cadmium ions per unit cell (PDB accession number: 6F0D)9. Crystals of both types diffracted below 2 Å. The data were collected on a Kappa Apex II diffractometer (Bruker AXS) using CuKα radiation generated by a IμS microfocus X-ray tube. Both structures were solved by molecular replacement in Phenix software suite v. 1.1110. The dataset with cadmium (6F0D) with unmerged Friedel pairs was used for SAD analysis. For experimental phasing, we used a standard protocol employing SHELXC/D/E programs11 through HKL2MAP v. 0.4 graphical interface12. Data were processed with SHELXC v. 2016/1, anomalous substructure was solved by SHELXD v. 2013/2 and phasing and density modification were done by SHELXE v. 2018/2. The automatic model building and refinement were done in Phenix v. 1.1410, and manual refinement was done in Coot v. 0.8.9.113. Figures were prepared with PyMOL.\n\n\nResults and discussion\n\nThe phasing of protein crystals by SAD starts from finding the positions of an anomalous substructure, which is usually done by direct methods. First, the dataset was processed with SHELXC, and the statistical analysis of the anomalous signal is shown in Figure 1A and Table 1. The use of kappa goniometer for data collection allowed achieving high completeness (96.4%) and multiplicity (5.9) of anomalous pair measurements. The signal-to-noise ratio defined as ⟨d′′/σ(d′′)⟩ and the correlation coefficient CC1/2 indicate that useful anomalous signal is present almost in the whole resolution range. For further substructure solution, we implied a rather conservative high-resolution cut-off of 2.4 Å corresponding to CC1/2 (anom.) ~ 0.3.\n\nThe crystal of the single-domain antibody BCD090-M2 with cadmium ions was used as an experimental model for in-house Cd-SAD. (A) Strength of the anomalous signal represented by ⟨d′′/σ(d′′)⟩ and CC1/2 as a function of resolution. (B) Electron density modification in SHELXE as monitored by an increase in map contrast; solutions with original and inverted anomalous substructure give indistinguishable contrast due to centrosymmetry. (C) Cadmium ion binding site. (D) Schematic representation of the crystal unit cell.\n\nValues in parentheses are for the highest resolution shell.\n\nCFOM, combined figure of merit; CC, correlation coefficient.\n\nThe anomalous substructure was immediately solved by SHELXD as judged by high correlation coefficients (combined figure of merit = 55.6%), high occupancies of the two cadmium sites (1.00, 0.99), and the rapid drop in occupancy of the next site (0.17). The positions of Cd ions corresponded to the largest off-origin peak of the anomalous Patterson function at (0.58, 0.02, 0.03). The solution was used in SHELXE for phasing, electron density modification, and chain tracing. This yielded electron density maps with high contrast, and the solutions for original and inverted substructure were indistinguishable due to centrosymmetry (Figure 1B). As discussed previously14, centrosymmetric anomalous sites in SAD can impede interpretation of electron density maps, because the resulting map is a superposition of the true electron density with its negative mirror-image. However, in our case the major portion of the protein chain (87%) was traced after density modification. This incomplete model was further improved in phenix.autobuild, and then refined manually in Coot and phenix.refine giving final Rwork/Rfree of 17.8/21.0%.\n\nIn this particular case, structure determination by in-house Cd-SAD was almost as straightforward as an automated molecular replacement. The causes of this simplicity were the relatively small protein size, high completeness and multiplicity of the anomalous data, and the small number of high-occupancy cadmium sites. Furthermore, the recent theoretical study gives the following simple dependency for expected anomalous signal ⟨Sano⟩ ~ (Nrefl/nsites)1/2, where Nrefl is the number of independent reflections and nsites is the number of anomalous scatterers15. Our case with maximum Nrefl due to the lowest symmetry (P1) and only 2 anomalous sites appears virtually optimal for SAD. The high metal-binding affinity of cadmium sites was achieved through coordination with carbonyl oxygen of Glu114, and carboxylic groups of Asp100 and Asp116 (Figure 1C). By bridging these residues to the N-terminal Gly residue of the neighboring molecule, cadmium ions effectively defined crystal contacts (Figure 1D). Data associated with this study are available on OSF16.\n\n\nConclusion\n\nIn conclusion, we suggest that cadmium SAD can be generally applied for the phasing of protein crystals collected in-house using CuKα radiation. We see the following advantages of this approach: (1) cadmium has a great anomalous signal (f′′=4.7e- at CuKα); (2) cadmium ions frequently promote crystal growth and can substitute other divalent cations; (3) cadmium binding sites are complementary to that of iodine, another strong anomalous scatterer, and therefore Cd-SAD can be useful in cases where I-SAD does not work.\n\n\nData availability\n\nData for this study, including unmerged experimental intensities, structure factors and final atomic coordinates after refinement, are available on OSF. DOI: https://doi.org/10.17605/OSF.IO/KYH6D16.\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\nThis work is funded by the Ministry of Science and Higher Education of the Russian Federation (contract 14.577.21.0217, unique identifier RFMEFI57716X0217) and co-funded by CJSC Biocad.\n\n\nReferences\n\nHendrickson WA: Anomalous diffraction in crystallographic phase evaluation. Q Rev Biophys. 2014; 47(1): 49–93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHendrickson WA, Teeter MM: Structure of the hydrophobic protein crambin determined directly from the anomalous scattering of sulphur. Nature. 1981; 290(5802): 107–113. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim MK, Lee S, An YJ, et al.: In-house zinc SAD phasing at Cu Kα edge. Mol Cells. 2013; 36(1): 74–81. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAbendroth J, Gardberg AS, Robinson JI, et al.: SAD phasing using iodide ions in a high-throughput structural genomics environment. J Struct Funct Genomics. 2011; 12(2): 83–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMelnikov I, Polovinkin V, Kovalev K, et al.: Fast iodide-SAD phasing for high-throughput membrane protein structure determination. Sci Adv. 2017; 3(5): e1602952. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTrakhanov S, Kreimer DI, Parkin S, et al.: Cadmium-induced crystallization of proteins: II. Crystallization of the Salmonella typhimurium histidine-binding protein in complex with L-histidine, L-arginine, or L-lysine. Protein Sci. 1998; 7(3): 600–604. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobbins AH, McRee DE, Williamson M, et al.: Refined crystal structure of Cd, Zn metallothionein at 2.0 A resolution. J Mol Biol. 1991; 221(4): 1269–1293. PubMed Abstract | Publisher Full Text\n\nPanneerselvam S, Kumpula EP, Kursula I, et al.: Rapid cadmium SAD phasing at the standard wavelength (1 Å). Acta Crystallogr D Struct Biol. 2017; 73(Pt 7): 581–590. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEliseev IE, Yudenko AN, Vysochinskaya VV, et al.: Crystal structures of a llama VHH antibody BCD090-M2 targeting human ErbB3 receptor [version 2; referees: 2 approved]. F1000Res. 2018; 7: 57. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAdams PD, Afonine PV, Bunkóczi G, et al.: PHENIX: a comprehensive Python-based system for macromolecular structure solution. Acta Crystallogr D Biol Crystallogr. 2010; 66(Pt 2): 213–221. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSheldrick GM: Experimental phasing with SHELXC/D/E: combining chain tracing with density modification. Acta Crystallogr D Biol Crystallogr. 2010; 66(Pt 4): 479–485. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPape T, Schneider TR: HKL2MAP: a graphical user interface for macromolecular phasing with SHELX programs. J Appl Crystallogr. 2004; 37(5): 843–844. Publisher Full Text\n\nEmsley P, Lohkamp B, Scott WG, et al.: Features and development of Coot. Acta Crystallogr D Biol Crystallogr. 2010; 66(Pt 4): 486–501. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcCoy AJ, Read RJ: Experimental phasing: best practice and pitfalls. Acta Crystallogr D Biol Crystallogr. 2010; 66(Pt 4): 458–469. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTerwilliger TC, Bunkóczi G, Hung LW, et al.: Can I solve my structure by SAD phasing? Anomalous signal in SAD phasing. Acta Crystallogr D Struct Biol. 2016; 72(Pt 3): 346–358. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEliseev I: Cadmium SAD phasing at CuKα wavelength. OSF. 2019. http://www.doi.org/10.17605/OSF.IO/KYH6D"
}
|
[
{
"id": "43346",
"date": "18 Feb 2019",
"name": "Thomas E. Edwards",
"expertise": [
"Reviewer Expertise X-ray crystallography",
"small molecular and antibody therapeutic development",
"de novo phasing."
],
"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\nEliseev et al. demonstrate the ability to use cadmium divalent cations, X-ray data collected on an in house X-ray source (Cu Kalpha), and single wavelength anomalous dispersion (SAD) phasing to determine the structure of a protein, in this case the anti-ErbB3 single domain antibody (sdAb). They originally solved the structure by molecular replacement (MR) and reported it in a previous publication, but here they demonstrate the applicability of Cd-SAD to this data set to determine the structure. By our examination of the literature, at least two other structures have been solved by Cd-SAD but both of these structures were solved using data collected at a synchrotron source at a longer wavelength (PDB IDs 2X7K and 5AM6). This particular application provides an excellent alternative method to others presented in the literature and referenced in this paper. A couple of notes and comments:\nThe authors may help the reader understand the number of electron possible for phasing by referring the reader to the UW X-ray anomalous scattering web page (http://skuld.bmsc.washington.edu/scatter/) designed by Ethan Merritt. The authors do an excellent job of data collection on their in house instrument by using the kappa goniometer to increase completeness and enhance the anomalous signal. The authors should also site a second paper by Terwilliger et al. from the same issue as reference 15 but with pages 359-3741 which provides additional information on experimental design and execution. We performed an overnight soak at 5 mM CdCl2, collected a high resolution data set in house on a Cu Kalpha source, and obtained sufficient anomalous signal to solve the structure by Cd-SAD. This provided us with good confidence that this method should be generally applicable.\n\nOverall, the current paper by Eliseev provides a good alternative approach to experimental structure determination, complementing previously reported techniques.\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": "45676",
"date": "01 Apr 2019",
"name": "Igor Melnikov",
"expertise": [
"Reviewer Expertise Macromolecular crystallography",
"structural 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\nIn this experiment report the authors present their work on derivatising crystals of a single-domain antibody protein with cadmium ions and subsequent solving the structure via SAD. They demonstrate that home-lab X-ray crystallographic equipment is capable of solving crystal structures via SAD at CuKalpha X-ray energy. The authors then conclude by discussing the result and suggesting the method for use in home-lab source diffraction experiments. To sum up, I find the study to be clearly presented and well prepared.\nAdditional notes:\nTo my point of view, the data were collected competently to reach higher multiplicity (which is indispensable for SAD phasing) for P1 space group by exploiting kappa goniometer. The workflow for structure solution via Cd-SAD presented in this article seems to be relatively straightforward even considering the problem with centrosymmetric Cd sites (easily resolved by chain tracing). \"The phasing of protein crystals by SAD starts from finding the positions of an anomalous substructure, which is usually done by direct methods.\" - In fact, the process involves Patterson search coupled with direct methods-based calculations1 namely dual-space refinement. I might suggest adding a few words on how the authors introduced cadmium ions into the crystal (I see that it was co-crystallisation as reported in their previous article).\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": []
}
] | 1
|
https://f1000research.com/articles/8-84
|
https://f1000research.com/articles/8-82/v1
|
21 Jan 19
|
{
"type": "Review",
"title": "The Hertfordshire Cohort Study: an overview",
"authors": [
"Holly E. Syddall",
"Shirley J. Simmonds",
"Sarah A. Carter",
"Sian M. Robinson",
"Elaine M. Dennison",
"Cyrus Cooper",
"The Hertfordshire Cohort Study Research Group",
"Holly E. Syddall",
"Shirley J. Simmonds",
"Sarah A. Carter",
"Sian M. Robinson",
"Elaine M. Dennison"
],
"abstract": "The Hertfordshire Cohort Study is a nationally unique study of men and women born in the English county of Hertfordshire in the early part of the 20th century. Records that detail their health in infancy and childhood have been preserved, their sociodemographic, lifestyle, medical and biological attributes have been characterised in later life, and routinely collected data on their hospital use and mortality have been acquired. This paper provides an overview of the study since its inception in the 1980s, including its methods, findings, and plans for its future.",
"keywords": [
"Hertfordshire",
"Cohort Study",
"Epidemiology",
"Ageing",
"Musculoskeletal",
"Lifecourse"
],
"content": "Background to the Hertfordshire Cohort Study\n\nIn the 1980s, ecological studies showed that death rates from heart disease in the 212 local authority areas of England and Wales during 1968–78 were correlated with infant mortality rates in those areas in the early part of the 20th century1. This raised the possibility that environmental influences acting during the fetal and infant stages of development might increase risk of cardiovascular disease in later life.\n\nHowever, stronger evidence from a more robust epidemiological study design was required. The discovery of a large set of birth records for people born in Hertfordshire during the first half of the twentieth century provided an opportunity for the MRC Environmental Epidemiology Unit (MRC EEU) to establish a cohort study in which the early origins of disease in later life could be explored.\n\nIn 1911, a team of midwives and nurses was assembled in Hertfordshire with the aim of improving the health of children in the county. Midwives attended women during childbirth and health visitors routinely visited the child throughout infancy and childhood; weight at birth and one year of age, method of infant feeding, illnesses, and developmental milestones were all recorded. The information (Table 1) was transcribed into ledgers (Figure 1) at the County Health Department in Hertford. The ledgers cover almost all births in Hertfordshire from 1911 until the NHS was formed in 1948.\n\nSource: original Hertfordshire ledger, on loan from the Hertfordshire Local Studies Archive; currently stored securely at the MRC LEU.\n\nThe MRC EEU computerised the Hertfordshire ledgers and traced the cohort by linkage with the National Health Service Central Register (NHSCR). A mortality study of 15,000 men and women born between 1911–30 showed that lower weight at birth and one year of age was associated with increased risk of death from cardiovascular disease by 75 years of age2. Subsequently, numerous detailed physiological studies were conducted among community dwelling men and women born in Hertfordshire between 1920–30 and still living there in the early 1990s. Small size in early life was associated with greater risk of cardiometabolic disease3–5, osteoporosis6 and sarcopenia (the loss of muscle mass, strength and function with age)7 in later life. Collectively these studies contributed to the “developmental origins” hypothesis which suggests that the nourishment a baby receives in utero, and its development in infancy and early childhood, determine its risk of disease in later life8.\n\nThese early studies were important for establishing relationships between the early environment and physiological markers of disease, but they lacked detailed information on adult anthropometry and diet, and the men and women born 1920–30 were becoming frail and unable to take part in further studies. A second (younger) cohort, in whom the work could be continued, was therefore recruited between 1998–2004. During this time Professor Cyrus Cooper became Hertfordshire Cohort Study (HCS) principal investigator and the MRC EEU was reconfigured as the MRC Lifecourse Epidemiology Unit (LEU), a University Unit Partnership between the MRC and the University of Southampton.\n\n\nKey methods\n\nThe younger cohort comprises 3000 men and women born between 1931 and 1939 who were traced following the model of the early studies. These are the people to whom the term ‘The Hertfordshire Cohort’ specifically applies. The remainder of this paper applies only to them, unless otherwise stated.\n\nThe principal objective of the HCS is to evaluate interactions between the intra-uterine and early postnatal environment, adult diet and lifestyle, and genetics, in the aetiology of chronic disorders of ageing (osteoporosis, osteoarthritis, sarcopenia, obesity, cardiometabolic disease and type II diabetes mellitus). The study aims to place these interactions within a life-course model for disease pathogenesis, and to characterise the physiological mechanisms underlying the pathways to these chronic disorders.\n\nThe baseline recruitment of HCS participants has been described in detail previously9; key stages in the process are shown in Figure 2. In brief, the ledgers contained records for 42974 births in Hertfordshire between 1931 and 1939; the National Health Service Central Register (NHSCR) in Southport (now part of NHS Digital) traced 8650 men and women who were still alive in Hertfordshire in 1998. Permission to contact 6099 men and women by letter was obtained from their GPs and 3225 (53%) agreed to a home interview with a trained research nurse. 2997 (93%) men and women subsequently attended a clinic for detailed physiological investigations. The baseline HCS fieldwork was county-wide but conducted in phases by geographical area and gender. 966 (68%) of the 1412 men and women who attended clinic in East Hertfordshire also underwent a Dual-energy X-ray Absorptiometry (DXA) bone scan and knee radiography. An overview of the data collected at the HCS baseline home interviews and clinics is provided in Table 2.\n\nThe HCS baseline investigations had ethical approval from the Hertfordshire and Bedfordshire Local Research Ethics Committee and all subjects gave written informed consent. All interviews and physiological investigations were carried out according to strict protocols and studies of within- and between-observer variation were conducted at regular intervals during fieldwork to ensure comparability of measurements obtained over time. Data were collected, anonymised, processed and stored in accordance with contemporary data protection regulations. Participants are free to withdraw from the study at any time.\n\nSince their first contact in 1998–2004, HCS cohort members have generously agreed to take part in various follow-up postal questionnaires, face-to-face interviews, clinics, intervention studies, and focus groups (see Figure 2). With their consent, data on hospital admissions experienced by the cohort members across a decade have been obtained from NHS Digital10, and cohort members remain flagged on the National Health Service Central Register for ongoing notification of mortality. Ethical approval was obtained for all elements of the study at the time they were conducted.\n\nTable 3 to Table 11 describe the scientific objective and methodology of each HCS follow-up study and detail key data availability. Taken together, the HCS datasets provide a rich, detailed and longitudinal characterisation of the sociodemographic, lifestyle, medical and biological attributes of a contemporary group of community-dwelling older men and women in the UK.\n\n\nKey findings\n\nHCS is a particularly valuable resource for research that aims to identify lifecourse influences on ageing. Longitudinal measurements are available for a range of markers of disease and for potential predictor and confounding variables. The study has been pioneering in its uptake of techniques not previously used in epidemiology to characterise health (for example High Resolution peripheral Quantitative Computed Tomography (HR pQCT) scanning and muscle biopsy) and has sought to maximise the value of routinely collected data through record linkage.\n\nThe following sections outline key publications in the areas of osteoporosis, osteoarthritis, sarcopenia and diet.\n\nOsteoporosis is the most common metabolic bone disease affecting older people; almost one in two women and one in five men aged 50 years will have an osteoporotic fracture in their remaining lifetime11. The LEU is a renowned contributor to research on osteoporosis, in part because of the wealth of data collected in HCS (Table 2–Table 12). Estimates of bone mineral density (BMD) and bone mineral content (BMC) have been obtained by DXA; estimates of bone strength (including strength strain index (SSI)), cortical bone and trabecular bone by peripheral Quantitative Computed Tomography (pQCT); and ‘virtual bone biopsies’ by HR pQCT. Few other cohorts have characterised bone in such detail.\n\nPredictors of osteoporosis. The Health Visitors’ records that are uniquely available in HCS have enabled investigation of the early-life origins of osteoporosis. Studies using DXA showed an increase in adult BMC with rising weight at birth, and more strongly, with weight at one year. Models that also included weight in adulthood showed impacts accruing from each time-point, with greater contributions of early growth to BMC than to BMD12. In addition, growth during the first year was shown to alter the geometry of the adult hip13. Studies using pQCT showed that bone strength was similarly related to early weights14.\n\nRisk of osteoporosis attributable to lifestyle has also been investigated in this cohort. While no social patterning has been observed15, nor any relationship found between cigarette smoking, alcohol use, or physical activity and either BMC or BMD16, alcohol has been associated with less radial cortical and trabecular bone in men and less tibial trabecular bone in women17. Clustering of risk factors (physical activity, diet, alcohol consumption, smoking behaviour, grip strength, personal and family interactions, comorbidities) has been found to increase the risk of low BMD in women18. Also in women, pQCT scans have shown that having a ‘healthier’ diet (characterised by greater consumption of fruit, vegetables and wholegrain cereals) is associated with increased bone area on pQCT scanning19. Some lifestyle risk factors appear to interact with early life factors to influence osteoporosis risk, for example, in men, having a low birth weight and being a current smoker increased the risk of low femoral BMD16.\n\nOsteoporosis and comorbidity. HCS has provided valuable insight into the interplay between osteoporosis and other chronic conditions. For example, in the case of cardiometabolic disease, analyses have shown that: BMD is associated with serum triglycerides and HDL22; higher BMI is associated with lower 25(OH) vitamin D23; arterial calcification affects the structure of cortical and trabecular bone in women, but not in men24; hyperinsulinemia influences BMD through BMI25; and the associations between lean and fat mass and bone microarchitecture differ26. In the case of chronic inflammation in later life (inflammaging), HCS has shown that an elevated inflammatory state is associated with reduced BMD at baseline and with an accelerated rate of decline over time27. Relationships between muscle and bone health are discussed in the section on sarcopenia that follows.\n\nOutcomes of osteoporosis. The primary outcome of osteoporosis is fracture, which carries high personal and societal costs. Risk of fracture has been elucidated in HCS, showing, for example, that history of falls adds to clinical risk factors and low BMD as a risk factor for fracture28. Research using pQCT scans has shown that in women, cortical radial thickness and area, and tibial cortical area and density, are associated with adult fracture risk, and in men tibial SSI is associated with adult fracture risk29. Finally, cluster analysis of bone microarchitecture from HR pQCT has demonstrated two separate phenotypes associated with high fracture risk31.\n\nOA is the most common joint disease affecting older people32. It is an important element of the research programme of the LEU and its incorporated Arthritis Research UK MRC Centre for Musculoskeletal Health and Work. In HCS, radiographic, clinical and self-reported markers of OA have been collected at multiple time points (Table 2–Table 12).\n\nDescriptive epidemiology. HCS data have been used to show that the method of ascertainment of OA status affects both its calculated prevalence33 and the demographic characteristics with which it is associated34. For example, prevalence of knee OA ranged from 18% for clinically diagnosed disease through 21% for self-reported disease to 42% when the diagnostic criteria were radiographic; with higher specificity and lower sensitivity for clinical and self-reported OA than for radiographically determined disease33. Further work utilising data for multiple methods of ascertainment showed that in the presence of radiographically determined disease in the knee, signs and symptoms differed dependent on the bones involved, tibiofemoral OA being associated with both signs and symptoms, and patellofemoral OA only with clinical signs35.\n\nPredictors of OA. Results from this study have demonstrated that growth before birth and in the first year both affect the likelihood of developing OA in later life36. Specifically, an increase in number of osteophytes on hip radiographs has been associated with low birthweight and the number on knee radiographs with low weight at 1 year. An increased prevalence of clinical OA at the hand was also seen in those whose weight at 1 was low. Conversely, rheumatoid arthritis was not associated with early weight37. The role of biochemical markers of bone in the aetiology of OA has also been investigated, finding that CTX-II and glucosyl-galactosyl-pyridinoline were associated with both osteophyte score and joint-space narrowing, whist CTX1 and osteocalcin were not38.\n\nOutcomes of OA. The consequences of OA include an increase in the impact of objectively measured neighbourhood problems on participants’ physical activity39 and quality of life40. Pain is an important symptom of OA which, in HCS and other EPOSA cohorts, has been found to fully explain the association of OA with objectively measured (but not with self-reported) physical function41. This suggests that individuals consider factors other than pain (e.g. comorbidity) in making an assessment of their physical function.\n\nTable 2–Table 12 show that HCS has collected a wealth of markers of the mass, strength, function and morphology of muscle. Together, these have enabled a rich body of research on the predictors and consequences of sarcopenia.\n\nDescriptive epidemiology. Sarcopenia may be defined by one or more of the markers collected in HCS, inter-relationships between which have been explored in detail42,43. The prevalence of sarcopenia across HCS, estimated according to European Working Group for Sarcopenia in Older People (EWGSOP) guidelines, but with lean mass estimated by skinfold measurements, was 4.6% in men and 7.9% in women44. For subgroups of HCS in which imaging and anthropometric data coexist, estimates of prevalence have been calculated according to more than one operational definition of sarcopenia (EWGSOP; International Working Group on Sarcopenia; Foundation for the National Institutes of Health sarcopenia project; dysmobility syndrome), enabling the methods to be compared directly44,45. Late-life declines in muscle mass as well as strength have been observed using data from two of the time points shown on Figure 2 (MSFU and EPOSA); these data suggest that muscle mass declines less rapidly than strength or function46.\n\nPredictors of sarcopenia. The early-life origins of sarcopenia have been investigated using data from the Health Visitors’ records. Weights at birth and one year of age have been related to adult anthropometry, showing that whilst BMI and lean mass rise with birthweight and with weight at one, fat mass has associations only with the latter47. This suggests that whilst the prenatal environment affects lean mass, the postnatal environment is more influential in the development of adult adiposity. Direct measures of muscle size from pQCT scanning have confirmed that birthweight is associated with muscle area in the calf and forearm48, whilst biopsies showed that smaller babies have lower total myofibre scores (kg/mm2) than those who were bigger at birth49. Grip strength was more strongly associated with birth weight than with growth during the first year of life in HCS participants50. This again suggests that sarcopenia has primarily pre- rather than post-natal origins, though there is some evidence to link breast feeding during infancy with increased grip strength51. A systematic review and meta-analysis of work in HCS and other cohorts showed that the positive relationship between birth weight and muscle strength exists across the lifecourse52.\n\nThe effects of adult lifestyle on risk of sarcopenia have also been investigated. First, to examine the role of diet, associations of grip strength with specific nutrients, individual foods and dietary patterns were studied. Healthier diets were associated with higher grip strength, although the strongest association was with more frequent consumption of oily fish53. Physical activity in old age, objectively measured by accelerometry, was found to be associated with decreased risk of EWGSOP sarcopenia and improved physical function, though not with grip strength54. Grip strength was not associated with lifetime occupational exposure to physically demanding activities55, a somewhat surprising finding in the light of evidence of social patterning in sarcopenia56. The role of adult heath in causal paths to sarcopenia has also been investigated, identifying multimorbidity57, inflammaging58, and some59, but not all60, cardiovascular drugs as important influences.\n\nOutcomes of sarcopenia. Sarcopenia has profound consequences for health and wellbeing in later life and is associated in HCS with increased risk of hospital admission61. Changes in ageing muscle have been shown to contribute to the metabolic syndrome and all its components62; to self-rated health and health-related quality of life63; and to bone mineral content64. Together with strong independent relationships between muscle and bone size/strength65, this latter finding supports the existence of a muscle/bone unit and provides some insight into the mechanisms that underlie it.\n\nDeterminants of diet quality. The HCS has described marked differences in dietary choices in later life. For example, healthier diets have been found to be associated with being female, reporting a non-manual social class, and with being a non-smoker66. Psychological and social factors have important influences67, greater involvement in social activities being associated with slower declines in diet quality68. In addition, a relationship has been described between infant feeding and health behaviours in later life, such that people who were breastfed as infants were more likely to have healthier diets69.\n\nOutcomes of diet quality. Our studies have described links between differences in nutrient intake and physical functioning in older women70, and shown that there is a graded increase in the prevalence of poor physical functioning in older adults as the number of risk factors (from obesity, smoking and poor diet) increases71. Less healthy diets also contribute to risk of hospital admission72. Additionally, higher processed meat consumption and antioxidant intake have been related to poorer lung function in HCS73, while low dietary antioxidants were also associated with poorer glucose tolerance74.\n\nMethodology. Assessment of diet in older populations can be burdensome. In HCS, a short food frequency questionnaire was therefore developed and validated to assess compliance with a healthy dietary pattern in older community-dwelling adults75.\n\n\nRoutinely collected data\n\nThe Medical Research Council76 and the Wellcome Trust77 have advocated for the scientific and interdisciplinary potential of cohorts to be enhanced through linkage to routine health records and administrative datasets. Such linkage is particularly valuable in an ageing cohort like HCS, where attrition may lead to bias if data collection requires participant contact. Successful linkage has been achieved in HCS with mortality and Hospital Episode Statistics (HES) data (Table 12), though difficulties persist over data access due to constantly changing regulations.\n\nIn total, the 1911 to 1939 Hertfordshire birth cohort comprises 37 000 men and women, 7916 of whom had died by the end of 1999. Risk of death from circulatory disease was lower among men and women who were heavier at birth. Women who were heavier at birth also had lower risk of death from pneumonia, injury, diabetes, and musculoskeletal disease78.\n\nHCS was one of the first English cohorts to link with HES data and produced early evidence on the descriptive epidemiology of hospital admissions among older individuals10. Such individual level data contrast with published statistics, which are at the population level. A novel methodology was developed to explore risk factors for admission in these multiple-failure survival data79, demonstrating links between admission and grip strength61, and showing that poor lifestyle risk factors have a cumulative effect on likelihood of admission72.\n\n\nGenetics and epigenetics\n\nSince its inception, HCS has studied the genetics of musculoskeletal ageing. The scientific approach has evolved across the years, alongside dramatic changes in the technologies available to characterise the genome. Early studies of genetic influences on adult disease in HCS considered the role of single nucleotide polymorphisms (SNPs) and their specific interaction with early life phenotypes as predictors of muscle and bone health in later life80,81. Haplotype studies followed81–83, and more recently HCS has been an important contributor to many genome wide association studies conducted by worldwide consortia84,85. Finally, in the Hertfordshire Sarcopenia Study (HSS), participation in the industry funded MEMOSA (Multi-Ethnic MOlecular determinants of human SArcopenia) collaboration has permitted cutting edge scientific techniques such as deep sequencing of RNA as well as high coverage methylation arrays to be completed in a subset of participants. Furthermore, muscle cells obtained from biopsies of older men and women have been cultured in vitro providing a biobank that together with RNA and methylation data, will permit the investigation of genetic and epigenetic pathways in musculoskeletal ageing. This is an exciting area for future research in HCS\n\n\nNational and international collaborations\n\nThe HCS research group has a long and successful history of conducting collaborative research; principal local, national and international collaborators are detailed in Table 13. The nature of HCS collaborations varies widely, all are welcomed by the HCS research team. For example, bespoke results from statistical analyses and/or datasets have been provided to collaborators (ranging from PhD students to established research groups and international consortia) for inclusion in systematic reviews and meta-analyses. In other instances, students have visited the MRC LEU and worked alongside the HCS research team to conduct a specific sub-study or to run statistical analyses and draft a paper for publication. Another mode of collaboration is the provision of biological samples (e.g. plasma or DNA) to research groups and their laboratories; the information that they generate is returned to the HCS research team for inclusion in the master HCS databases. The resulting enhanced database is either analysed by HCS statisticians or released to the collaborators for them to analyse; either approach results in a jointly authored peer reviewed publication. Finally, the HCS research team may collaborate by submitting an ethics application jointly with a research group and conducting the study together; for example, the HPAT physical activity trial, the wellbeing follow-up study, and the VIBE study were conducted in collaboration with the MRC Epidemiology Unit (University of Cambridge), the MRC Unit for Lifelong Health and Ageing (at UCL) and the University of Bristol, respectively (see Figure 2 for timeline of HCS sub-studies).\n\n\nThe future\n\nHCS is a flagship cohort that has demonstrated lifecourse influences on musculoskeletal health in later life. All of the information collected to date is extensively documented and carefully curated and, as evidenced by a broad body of collaborators and over 250 peer-reviewed publications to date, the cohort will continue to be an invaluable resource for national and international research on the lifecourse determinants of musculoskeletal health in later life. Cohort members remain flagged with NHS Digital for ongoing notification of death, and further extracts of HES data would enable long-term follow-up of morbidity for the whole cohort (subject to funding and permission to access data).\n\nTwo examples of future research areas in HCS are: first, development of lifecourse interventions to promote maintenance of health, and slow loss of function, of the musculoskeletal system in later life; and second, investigation of the mechanisms that explain how early life environment influences health in later life.\n\nThe ambition to develop lifecourse interventions to maintain the health of the musculoskeletal system in later life is a natural progression from the epidemiological studies that have been conducted in HCS and other cohorts to date. Diet and physical activity are now well established as key influences on musculoskeletal health in later life, but research in HCS and elsewhere has shown that lifestyle risk factors cluster together to impact on physical function in later life71. Therefore, any lifestyle based intervention strategy would be well advised to adopt a holistic whole-person approach. Accordingly, the HCS research team’s objective is to develop and test the feasibility of a combined healthy conversation and physical activity intervention to promote muscle mass, strength and function in later life with the ultimate aim of preventing falls and fractures. If this is successful, the intervention will be implemented more widely and its efficacy evaluated.\n\nEpigenetics is an important area for research that seeks to move from epidemiological associations to identification of the mechanisms that underpin early life influences on health in later life. Epigenetics is the study of changes in organisms caused by modification of gene expression rather than alteration of the genetic code itself. For example, in HSS and HSSe participants, we are in the early phase of investigating whether the expression of genes associated with cellular pathways is altered during muscle ageing and how this contributes to sarcopenia. We are also studying genes associated with inflammation, mitochondrial function and the regulation of muscle growth, and taking the novel approach of investigating epigenetic pathways by studying DNA methylation markers in muscle tissue and also in cultured primary skeletal muscle cells. The results from this work will give us clues about how sarcopenia begins, progresses, and what can be done to prevent or treat it.\n\nEpigenetics underpins the ambition to expand HCS to an intergenerational study. The theory of transgenerational epigenetics implies that maternal environment during pregnancy transmits to offspring and affects their subsequent gene expression86. An intergenerational study in HCS is a very special opportunity to investigate how epigenetic expression transmits across generations, and indeed to explore commonalities and differences in musculoskeletal health across generations. To date, more than 1,000 members of the HCS cohort (the ‘F0 generation’) have provided us with the names and contact details of their children (the ‘F1 generation’) and grandchildren (the ‘F2 generation’). We are now in the process of recruiting these children and grandchildren to an HCS three generation cohort (HCS 3G); in the first instance we are seeking their consent to contact them in the future with invitations to participate in specific studies as they are planned. To date, more than 700 children and grandchildren have agreed to be members of the HCS 3G study. Following on in the tradition of their parents or grandparents who have participated so generously in HCS, these people will make an important contribution to our understanding of lifecourse and intergenerational influences on musculoskeletal health.\n\n\nData availability\n\nHospital Episode Statistics and mortality data were obtained from NHS Digital under Data Sharing Agreements numbered 148284 and 343023; but cannot be made openly available for ethical reasons. HCS data collected directly by the MRC LEU are accessible via collaboration with the HCS research group as described above. Initial enquires should be made to Professor Cyrus Cooper (Principal Investigator, e-mail: cc@mrc.soton.ac.uk). Potential collaborators will be sent a collaborators’ pack and asked to submit a detailed study proposal to the HCS Steering Group; applications will be reviewed at the first available steering group meeting (held three times annually).",
"appendix": "Grant information\n\nHCS is supported by the Medical Research Council University Unit Partnership grant number MRC_MC_UP_A620_1014.\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\nWe thank the men and women who have participated in the Hertfordshire studies, the Hertfordshire General Practitioners, and all the nurses and doctors who have conducted HCS fieldwork over many years.\n\nThe Hertfordshire Cohort Study research group currently comprises: Gregorio Bevilacqua; Ilse Bloom; Michael Clynes; Ken Cox; Vanessa Cox; Nicholas Fuggle; Catharine Gale; John Holloway; Karen Jameson; Camille Parsons; Harnish Patel; Kate Ward; Leo Westbury.\n\n\nReferences\n\nBarker DJ, Osmond C: Infant mortality, childhood nutrition, and ischaemic heart disease in England and Wales. Lancet. 1986; 1(8489): 1077–81. PubMed Abstract | Publisher Full Text\n\nOsmond C, Barker DJ, Winter PD, et al.: Early growth and death from cardiovascular disease in women. BMJ. 1993; 307(6918): 1519–24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFall CH, Osmond C, Barker DJ, et al.: Fetal and infant growth and cardiovascular risk factors in women. BMJ. 1995; 310(6977): 428–32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHales CN, Barker DJ, Clark PM, et al.: Fetal and infant growth and impaired glucose tolerance at age 64. BMJ. 1991; 303(6809): 1019–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBarker DJ, Hales CN, Fall CH, et al.: Type 2 (non-insulin-dependent) diabetes mellitus, hypertension and hyperlipidaemia (syndrome X): relation to reduced fetal growth. Diabetologia. 1993; 36(1): 62–7. PubMed Abstract | Publisher Full Text\n\nCooper C, Fall C, Egger P, et al.: Growth in infancy and bone mass in later life. Ann Rheum Dis. 1997; 56(1): 17–21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSayer AA, Cooper C, Evans JR, et al.: Are rates of ageing determined in utero? Age Ageing. 1998; 27(5): 579–83. PubMed Abstract | Publisher Full Text\n\nBarker DJP: Mothers, babies and health in later life. 2 ed. Edinburgh: Churchill Livingstone; 1998. Reference Source\n\nSyddall HE, Aihie Sayer A, Dennison EM, et al.: Cohort profile: the Hertfordshire cohort study. Int J Epidemiol. 2005; 34(6): 1234–42. PubMed Abstract | Publisher Full Text\n\nSimmonds SJ, Syddall HE, Walsh B, et al.: Understanding NHS hospital admissions in England: linkage of Hospital Episode Statistics to the Hertfordshire Cohort Study. Age Ageing. 2014; 43(5): 653–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Staa TP, Dennison EM, Leufkens HG, et al.: Epidemiology of fractures in England and Wales. Bone. 2001; 29(6): 517–22. PubMed Abstract | Publisher Full Text\n\nDennison EM, Syddall HE, Sayer AA, et al.: Birth weight and weight at 1 year are independent determinants of bone mass in the seventh decade: the Hertfordshire cohort study. Pediatr Res. 2005; 57(4): 582–6. PubMed Abstract | Publisher Full Text\n\nJavaid MK, Lekamwasam S, Clark J, et al.: Infant growth influences proximal femoral geometry in adulthood. J Bone Miner Res. 2006; 21(4): 508–12. PubMed Abstract | Publisher Full Text\n\nOliver H, Jameson KA, Sayer AA, et al.: Growth in early life predicts bone strength in late adulthood: the Hertfordshire Cohort Study. Bone. 2007; 41(3): 400–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSyddall HE, Evandrou M, Dennison EM, et al.: Social inequalities in osteoporosis and fracture among community-dwelling older men and women: findings from the Hertfordshire Cohort Study. Arch Osteoporos. 2012; 7(1–2): 37–48. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMoinuddin MM, Jameson KA, Syddall HE, et al.: Cigarette smoking, birthweight and osteoporosis in adulthood: results from the hertfordshire cohort study. Open Rheumatol J. 2008; 2: 33–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaccou J, Edwards MH, Ward K, et al.: Relationships between bone geometry, volumetric bone mineral density and bone microarchitecture of the distal radius and tibia with alcohol consumption. Bone. 2015; 78: 122–9. PubMed Abstract | Publisher Full Text\n\nZhang J, Jameson K, Sayer AA, et al.: Accumulation of risk factors associated with poor bone health in older adults. Arch Osteoporos. 2016; 11: 3. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShaw SC, Parsons CM, Fuggle NR, et al.: Diet Quality and Bone Measurements Using HRpQCT and pQCT in Older Community-Dwelling Adults from the Hertfordshire Cohort Study. Calcif Tissue Int. 2018; 103(5): 494–500. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFinucane FM, Horton J, Purslow LR, et al.: Randomized controlled trial of the efficacy of aerobic exercise in reducing metabolic risk in healthy older people: The Hertfordshire Physical Activity Trial. BMC Endocr Disord. 2009; 9: 15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatel HP, Syddall HE, Martin HJ, et al.: Hertfordshire sarcopenia study: design and methods. BMC Geriatr. 2010; 10: 43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDennison EM, Syddall HE, Aihie Sayer A, et al.: Lipid profile, obesity and bone mineral density: the Hertfordshire Cohort Study. QJM. 2007; 100(5): 297–303. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVimaleswaran KS, Berry DJ, Lu C, et al.: Causal relationship between obesity and vitamin D status: bi-directional Mendelian randomization analysis of multiple cohorts. PLoS Med. 2013; 10(2): e1001383. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPaccou J, Edwards MH, Patsch JM, et al.: Lower leg arterial calcification assessed by high-resolution peripheral quantitative computed tomography is associated with bone microstructure abnormalities in women. Osteoporos Int. 2016; 27(11): 3279–87. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDennison EM, Syddall HE, Aihie Sayer A, et al.: Type 2 diabetes mellitus is associated with increased axial bone density in men and women from the Hertfordshire Cohort Study: evidence for an indirect effect of insulin resistance? Diabetologia. 2004; 47(11): 1963–8. PubMed Abstract | Publisher Full Text\n\nEdwards MH, Ward KA, Ntani G, et al.: Lean mass and fat mass have differing associations with bone microarchitecture assessed by high resolution peripheral quantitative computed tomography in men and women from the Hertfordshire Cohort Study. Bone. 2015; 81: 145–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFuggle NR, Westbury LD, Syddall HE, et al.: Relationships between markers of inflammation and bone density: findings from the Hertfordshire Cohort Study. Osteoporos Int. 2018; 29(7): 1581–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdwards MH, Jameson K, Denison H, et al.: Clinical risk factors, bone density and fall history in the prediction of incident fracture among men and women. Bone. 2013; 52(2): 541–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDennison EM, Jameson KA, Edwards MH, et al.: Peripheral quantitative computed tomography measures are associated with adult fracture risk: the Hertfordshire Cohort Study. Bone. 2014; 64: 13–7. PubMed Abstract | Publisher Full Text\n\nvan der Pas S, Castell MV, Cooper C, et al.: European project on osteoarthritis: design of a six-cohort study on the personal and societal burden of osteoarthritis in an older European population. BMC Musculoskelet Disord. 2013; 14: 138. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdwards MH, Robinson DE, Ward KA, et al.: Cluster analysis of bone microarchitecture from high resolution peripheral quantitative computed tomography demonstrates two separate phenotypes associated with high fracture risk in men and women. Bone. 2016; 88: 131–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFelson DT, Naimark A, Anderson J, et al.: The prevalence of knee osteoarthritis in the elderly. The Framingham Osteoarthritis Study. Arthritis Rheum. 1987; 30(8): 914–8. PubMed Abstract | Publisher Full Text\n\nParsons C, Clynes M, Syddall H, et al.: How well do radiographic, clinical and self-reported diagnoses of knee osteoarthritis agree? Findings from the Hertfordshire cohort study. Springerplus. 2015; 4: 177. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKerkhof HJ, Meulenbelt I, Akune T, et al.: Recommendations for standardization and phenotype definitions in genetic studies of osteoarthritis: the TREAT-OA consortium. Osteoarthritis Cartilage. 2011; 19(3): 254–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nParsons C, Fuggle NR, Edwards MH, et al.: Concordance between clinical and radiographic evaluations of knee osteoarthritis. Aging Clin Exp Res. 2018; 30(1): 17–25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nClynes MA, Parsons C, Edwards MH, et al.: Further evidence of the developmental origins of osteoarthritis: results from the Hertfordshire Cohort Study. J Dev Orig Health Dis. 2014; 5(6): 453–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdwards CJ, Goswami R, Goswami P, et al.: Growth and infectious exposure during infancy and the risk of rheumatoid factor in adult life. Ann Rheum Dis. 2006; 65(3): 401–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJordan KM, Syddall HE, Garnero P, et al.: Urinary CTX-II and glucosyl-galactosyl-pyridinoline are associated with the presence and severity of radiographic knee osteoarthritis in men. Ann Rheum Dis. 2006; 65(7): 871–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTimmermans EJ, van der Pas S, Cooper C, et al.: The neighbourhood environment and use of neighbourhood resources in older adults with and without lower limb osteoarthritis: results from the Hertfordshire Cohort Study. Clin Rheumatol. 2016; 35(11): 2797–805. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTimmermans EJ, van der Pas S, Schaap LA, et al.: Associations between perceived neighbourhood problems and quality of life in older adults with and without osteoarthritis: Results from the Hertfordshire Cohort Study. Health Place. 2017; 43: 144–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZambon S, Siviero P, Denkinger M, et al.: Role of Osteoarthritis, Comorbidity, and Pain in Determining Functional Limitations in Older Populations: European Project on Osteoarthritis. Arthritis Care Res. 2016; 68(6): 801–10. PubMed Abstract | Publisher Full Text\n\nStevens PJ, Syddall HE, Patel HP: Is grip strength a good marker of physical performance among community-dwelling older people? J Nutr Health Aging. 2012; 16(9): 769–74. PubMed Abstract | Publisher Full Text\n\nHardy R, Cooper R, Aihie Sayer A, et al.: Body mass index, muscle strength and physical performance in older adults from eight cohort studies: the HALCyon programme. PLoS One. 2013; 8(2): e56483. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatel HP, Syddall HE, Jameson K, et al.: Prevalence of sarcopenia in community-dwelling older people in the UK using the European Working Group on Sarcopenia in Older People (EWGSOP) definition: findings from the Hertfordshire Cohort Study (HCS). Age Ageing. 2013; 42(3): 378–84. PubMed Abstract | Publisher Full Text | Free Full Text\n\nClynes MA, Edwards MH, Buehring B, et al.: Definitions of Sarcopenia: Associations with Previous Falls and Fracture in a Population Sample. Calcif Tissue Int. 2015; 97(5): 445–52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatel A, Edwards MH, Jameson KA, et al.: Longitudinal Change in Peripheral Quantitative Computed Tomography Assessment in Older Adults: The Hertfordshire Cohort Study. Calcif Tissue Int. 2018; 103(5): 476–482. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSayer AA, Syddall HE, Dennison EM, et al.: Birth weight, weight at 1 y of age, and body composition in older men: findings from the Hertfordshire Cohort Study. Am J Clin Nutr. 2004; 80(1): 199–203. PubMed Abstract | Publisher Full Text\n\nSayer AA, Dennison EM, Syddall HE, et al.: The developmental origins of sarcopenia: using peripheral quantitative computed tomography to assess muscle size in older people. J Gerontol A Biol Sci Med Sci. 2008; 63(8): 835–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPatel HP, Jameson KA, Syddall HE, et al.: Developmental influences, muscle morphology, and sarcopenia in community-dwelling older men. J Gerontol A Biol Sci Med Sci. 2012; 67(1): 82–7. PubMed Abstract | Publisher Full Text\n\nSayer AA, Syddall HE, Gilbody HJ, et al.: Does sarcopenia originate in early life? Findings from the Hertfordshire cohort study. J Gerontol A Biol Sci Med Sci. 2004; 59(9): M930–M4. PubMed Abstract | Publisher Full Text\n\nRobinson SM, Simmonds SJ, Jameson KA, et al.: Muscle strength in older community-dwelling men is related to type of milk feeding in infancy. J Gerontol A Biol Sci Med Sci. 2012; 67(9): 990–6. PubMed Abstract | Publisher Full Text\n\nDodds R, Denison HJ, Ntani G, et al.: Birth weight and muscle strength: a systematic review and meta-analysis. J Nutr Health Aging. 2012; 16(7): 609–15. PubMed Abstract | Publisher Full Text\n\nRobinson SM, Jameson KA, Batelaan SF, et al.: Diet and its relationship with grip strength in community-dwelling older men and women: the Hertfordshire cohort study. J Am Geriatr Soc. 2008; 56(1): 84–90. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWestbury LD, Dodds RM, Syddall HE, et al.: Associations Between Objectively Measured Physical Activity, Body Composition and Sarcopenia: Findings from the Hertfordshire Sarcopenia Study (HSS). Calcif Tissue Int. 2018; 103(3): 237–245. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWalker-Bone K, D'Angelo S, Syddall HE, et al.: Heavy manual work throughout the working lifetime and muscle strength among men at retirement age. Occup Environ Med. 2016; 73(4): 284–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSyddall H, Evandrou M, Cooper C, et al.: Social inequalities in grip strength, physical function, and falls among community dwelling older men and women: findings from the Hertfordshire Cohort Study. J Aging Health. 2009; 21(6): 913–39. PubMed Abstract | Publisher Full Text\n\nSyddall HE, Westbury LD, Shaw SC, et al.: Correlates of Level and Loss of Grip Strength in Later Life: Findings from the English Longitudinal Study of Ageing and the Hertfordshire Cohort Study. Calcif Tissue Int. 2018; 102(1): 53–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWestbury LD, Fuggle NR, Syddall HE, et al.: Relationships Between Markers of Inflammation and Muscle Mass, Strength and Function: Findings from the Hertfordshire Cohort Study. Calcif Tissue Int. 2018; 102(3): 287–95. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAshfield TA, Syddall HE, Martin HJ, et al.: Grip strength and cardiovascular drug use in older people: findings from the Hertfordshire Cohort Study. Age Ageing. 2010; 39(2): 185–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWitham MD, Syddall HE, Dennison E, et al.: ACE inhibitors, statins and thiazides: no association with change in grip strength among community dwelling older men and women from the Hertfordshire Cohort Study. Age Ageing. 2014; 43(5): 661–6. PubMed Abstract | Publisher Full Text\n\nSimmonds SJ, Syddall HE, Westbury LD, et al.: Grip strength among community-dwelling older people predicts hospital admission during the following decade. Age Ageing. 2015; 44(6): 954–9. PubMed Abstract | Publisher Full Text\n\nSayer AA, Syddall HE, Dennison EM, et al.: Grip strength and the metabolic syndrome: findings from the Hertfordshire Cohort Study. QJM. 2007; 100(11): 707–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSayer AA, Syddall HE, Martin HJ, et al.: Is grip strength associated with health-related quality of life? Findings from the Hertfordshire Cohort Study. Age Ageing. 2006; 35(4): 409–15. PubMed Abstract | Publisher Full Text\n\nPatel HP, Dawson A, Westbury LD, et al.: Muscle Mass, Muscle Morphology and Bone Health Among Community-Dwelling Older Men: Findings from the Hertfordshire Sarcopenia Study (HSS). Calcif Tissue Int. 2018; 103(1): 35–43. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEdwards MH, Gregson CL, Patel HP, et al.: Muscle size, strength, and physical performance and their associations with bone structure in the Hertfordshire Cohort Study. J Bone Miner Res. 2013; 28(11): 2295–304. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobinson S, Syddall H, Jameson K, et al.: Current patterns of diet in community-dwelling older men and women: results from the Hertfordshire Cohort Study. Age Ageing. 2009; 38(5): 594–9. PubMed Abstract | Publisher Full Text\n\nBloom I, Lawrence W, Barker M, et al.: What influences diet quality in older people? A qualitative study among community-dwelling older adults from the Hertfordshire Cohort Study, UK. Public Health Nutr. 2017; 20(15): 2685–93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBloom I, Edwards M, Jameson KA, et al.: Influences on diet quality in older age: the importance of social factors. Age Ageing. 2017; 46(2): 277–83. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobinson S, Ntani G, Simmonds S, et al.: Type of milk feeding in infancy and health behaviours in adult life: findings from the Hertfordshire Cohort Study. Br J Nutr. 2013; 109(6): 1114–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartin H, Aihie Sayer A, Jameson K, et al.: Does diet influence physical performance in community-dwelling older people? Findings from the Hertfordshire Cohort Study. Age Ageing. 2011; 40(2): 181–6. PubMed Abstract | Publisher Full Text\n\nRobinson SM, Jameson KA, Syddall HE, et al.: Clustering of lifestyle risk factors and poor physical function in older adults: the Hertfordshire cohort study. J Am Geriatr Soc. 2013; 61(10): 1684–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSyddall HE, Westbury LD, Simmonds SJ, et al.: Understanding poor health behaviours as predictors of different types of hospital admission in older people: findings from the Hertfordshire Cohort Study. J Epidemiol Community Health. 2016; 70(3): 292–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOkubo H, Shaheen SO, Ntani G, et al.: Processed meat consumption and lung function: modification by antioxidants and smoking. Eur Respir J. 2014; 43(4): 972–82. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOkubo H, Syddall HE, Phillips DI, et al.: Dietary total antioxidant capacity is related to glucose tolerance in older people: the Hertfordshire Cohort Study. Nutr Metab Cardiovasc Dis. 2014; 24(3): 301–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobinson SM, Jameson KA, Bloom I, et al.: Development of a Short Questionnaire to Assess Diet Quality among Older Community-Dwelling Adults. J Nutr Health Aging. 2017; 21(3): 247–53. PubMed Abstract | Publisher Full Text\n\nMedical Research Council: Maximising the value of UK population cohorts. MRC strategic review of the largest UK population cohort studies. London; 2014. Reference Source\n\nWellcome Trust: Enabling linkage to maximise the value of public health research data. London: Wellcome Trust; 2015. Reference Source\n\nSyddall HE, Sayer AA, Simmonds SJ, et al.: Birth weight, infant weight gain, and cause-specific mortality: the Hertfordshire Cohort Study. Am J Epidemiol. 2005; 161(11): 1074–80. PubMed Abstract | Publisher Full Text\n\nWestbury LD, Syddall HE, Simmonds SJ, et al.: Identification of risk factors for hospital admission using multiple-failure survival models: a toolkit for researchers. BMC Med Res Methodol. 2016; 16: 46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDennison EM, Syddall HE, Jameson KA, et al.: A study of relationships between single nucleotide polymorphisms from the growth hormone-insulin-like growth factor axis and bone mass: the Hertfordshire cohort study. J Rheumatol. 2009; 36(7): 1520–6. PubMed Abstract | Publisher Full Text\n\nLips MA, Syddall HE, Gaunt TR, et al.: Interaction between birthweight and polymorphism in the calcium-sensing receptor gene in determination of adult bone mass: the Hertfordshire cohort study. J Rheumatol. 2007; 34(4): 769–75. PubMed Abstract | Free Full Text\n\nValdes AM, Styrkarsdottir U, Doherty M, et al.: Large scale replication study of the association between HLA class II/BTNL2 variants and osteoarthritis of the knee in European-descent populations. PLoS One. 2011; 6(8): e23371. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVaranasi SS, Tuck SP, Mastana SS, et al.: Lack of association of bone morphogenetic protein 2 gene haplotypes with bone mineral density, bone loss, or risk of fractures in men. J Osteoporos. 2011; 2011: 243465. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChambers JC, Zhang W, Sehmi J, et al.: Genome-wide association study identifies loci influencing concentrations of liver enzymes in plasma. Nat Genet. 2011; 43(11): 1131–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKemp JP, Morris JA, Medina-Gomez C, et al.: Identification of 153 new loci associated with heel bone mineral density and functional involvement of GPC6 in osteoporosis. Nat Genet. 2017; 49(10): 1468–75. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGrossniklaus U, Kelly WG, Kelly B, et al.: Transgenerational epigenetic inheritance: how important is it? Nat Rev Genet. 2013; 14(3): 228–35. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "43323",
"date": "12 Feb 2019",
"name": "Rebecca Reynolds",
"expertise": [
"Reviewer Expertise Developmental Origins of Health and 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 is a clearly written overview of the Hertfordshire Cohort Study. The manuscript presents the background leading to the start of the Cohort and then logically describes the study phases and methods with clear summary tables of the variables collected. The manuscript reports key outcomes in some of the major areas of research that have been conducted using the Cohort, particularly in relation to bone and musculoskeletal health and diet. Details of linkage to routinely collected data, contribution to genetic and epigenetic studies and key collaborators are listed. Plans for the future directions of the cohort, including recruitment to an intergenerational study are described. Overall, the manuscript provides a very useful reference tool and resource for parties interested in collaborating with the Southampton team. My only suggestion for improvement of the manuscript would be to add a paragraph on the limitations of the Cohort e.g. whether there are variables that are not available or with missing data, and whether the lack of these data means that there are certain study questions that cannot be readily addressed using the Cohort. For example, it does not appear that pregnancy-related variables are available in the original cohort, but I presume some may be available in the new intergenerational cohort?\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": "4436",
"date": "20 Feb 2019",
"name": "Cyrus Cooper",
"role": "Author Response",
"response": "We thank Prof Reynolds for her positive assessment of our manuscript. We have taken her useful comments and suggestions on board and will consider how best to respond to them (either by making minor changes to the text of the manuscript, or by posting specific responses to her comments via this online system)."
}
]
},
{
"id": "43324",
"date": "18 Feb 2019",
"name": "Aravinda Guntupalli",
"expertise": [
"Reviewer Expertise Ageing",
"life course perspective",
"anthropometry"
],
"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 provides an excellent overview. See below a few comments that could further improve the overview.\n\n1. Motivation and contribution of the paper\nThe Hertfordshire cohort study has resulted in several important discoveries related to the relationship between conditions in infancy and later life and this paper summarises the outcomes very well.\nHowever, I believe that it would be important to further highlight the motivation of the paper in the introduction. Most importantly, the following questions should be addressed:\nWhat can we learn from the Hertfordshire cohort study (three most prominent findings)? Can the experience in setting up and conducting the Hertfordshire cohort study be transferred to other settings (put differently are there country and context specific circumstances that need to be taken into consideration)? What are the main challenges (perhaps costs) that need to be considered?\nHence, it would be important to provide a clear motivation for this paper before discussing the historical background. Most importantly, the authors need to convince potential readers, who maybe unfamiliar with the UK, why this paper and the experience gained in conducting the Hertfordshire cohort study matters for their research agenda or potential cohort studies. Put differently, can the success of the Hertfordshire cohort study be replicated in other countries?\n2. Age, period and cohort effect\nThe early study focused on men and women born between 1911-30 with a special focus on 1920 and 1930 cohort. The 1920-30 cohort were are born at the tail end of the 'Spanish flu' and the start of the great depression. The younger cohort (the cohort of this study) comprises of 3000 men and women born between 1931 and 1939. During this period the great depression impact continued and the World War 2 started. It would be good to see the differences in findings between these two older and newer cohort studies from this context, if it is possible. Also, it would be good to say something on the socio-economic situation of the 1931-39 cohort. Would any of the historical events impact the results perhaps through survival bias? How was the infant mortality in this period? Can you describe the socio-economic status of the cohort and the differences in the ageing experiences by SES, if such information is available?\n\n3. Methods of data collection over time\nThe paper discusses various methods of data collection summarised in Table 2. It would be interesting to know whether these methods (e.g. home interviews, clinic visits) have changed over time due to improved technology (e.g. home testing kits, questionnaires via mobile phones or email links etc.). Finally, could the authors provide cost estimates per participant? This would be valuable information for other research projects that consider setting up a cohort study in other countries and settings.\n4. The ‘early studies’ versus the Hertfordshire cohort study\nThe authors highlight that the ‘early studies’ had limitations, which were subsequently addressed. It might be useful to explain in detail what the limitations were. What can readers learn from this experience?\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": "4437",
"date": "20 Feb 2019",
"name": "Cyrus Cooper",
"role": "Author Response",
"response": "We thank Dr Guntupalli for her positive assessment of our manuscript. Her comments are interesting, particularly her suggestion about placing the cohort in its broader historical and socioeconomic context. Information has previously been published (in papers and books) which does serve this purpose but it would be useful to incorporate some detail with this particular manuscript (either by making minor changes to the text of the manuscript, or by posting specific responses to her comments via this online system). We will consider how best to provide further details."
}
]
},
{
"id": "43326",
"date": "22 Feb 2019",
"name": "Cecily Kelleher",
"expertise": [
"Reviewer Expertise Cardiovascular and nutritional epidemiology and health promotion"
],
"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 report provides an overview of the Hertfordshire Cohort Study, one of the most influential of epidemiological studies which provided ground-breaking evidence of early life influences on the development of adult chronic disease and laid the foundation for a huge corpus of scientific work on the developmental origins of health and disease hypothesis. The original retrospective cohort study was established by linking contemporary health outcome data to health visitor records collected between 1911-1930 in the English county of Hertfordshire.\nThe paper concentrates on the highly innovative strategy of further follow-up examinations and data collection in the younger cohort of 3000 men and women who were born between 1931 and 1939. The objective was to evaluate interactions between intra-uterine and early postnatal environment, adult diet and lifestyle, genetics in the aetiology of chronic disorders of ageing including osteoporosis. A comprehensive battery of social, biological and health status variables was collected and the summary of studies carried out between 1998-2018 is summarised in Figure 2. A further highly original development is the inter-generational follow-up now planned and in train. To date more than 1000 members have provided contact details for their children and grandchildren, which will make this one of very few prospective studies with three generation follow-up. There is also detailed linkage to routine health records and administrative datasets, mortality and hospital episodes statistics.\nThe review provides a comprehensive summary of key publications in the areas of osteoporosis, osteoarthritis and diet and this study has greatly added to our understanding of the aetiology and natural history of these outcomes. There are also a number of detailed tables which summarise the wealth of data collected at different stages of the study.\nMy comments are all minor and primarily to do with presentation of the data in the report. Firstly, as a suggestion, I think the report would flow better if the sections on diet, routinely collected data, genetics and epigenetics and possibly National and International collaborations (currently pages 15-16) were moved back earlier to precede the outcome sections, focusing particularly on Osteoporosis, Osteoarthritis, Sarcopenia and Diet (currently pages 4-15). It might make the scope of the study clearer to the general reader. Also, the tables provide considerable detail and if the variables were sequenced in the same order and as appropriate sub-categorised, it would be easier to cross-reference what was collected and when.\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/8-82
|
https://f1000research.com/articles/8-71/v1
|
18 Jan 19
|
{
"type": "Research Article",
"title": "Progression free survival in Iraqi breast cancer patients treated with adjuvant 3D conformal radiotherapy: A cross-sectional study",
"authors": [
"Manwar Abdulelah Al-Naqqash",
"Enas Khudhair Al-Bdaer",
"Wieeam Abdulfattah Saleh Saleh",
"Ahmed Salih Al-Shewered",
"Enas Khudhair Al-Bdaer",
"Wieeam Abdulfattah Saleh Saleh",
"Ahmed Salih Al-Shewered"
],
"abstract": "Background: Breast cancer is a common malignancy in Iraq, accounting for one-third of female cancers in the Iraqi Cancer Registry. Radiotherapy confers benefits for local control and progression free survival (PFS) in patients with breast cancer. This study aimed to assess PFS in patients treated by hypofractionated three-dimensional conformal radiotherapy (3DCRT) and correlate PFS with patients' clinical and pathological profiles. Methods: We retrospectively reviewed 299 women with breast cancer treated at Baghdad Radiation Oncology Center between October 2017 and May 2018. Regarding radiotherapy, 4005 cGy in 15 fractions over 3 weeks was adopted as standard practice for patients undergoing mastectomy and 4005 cG in 15 fractions + 1000 cG in 5 fractions as a booster dose for women undergoing breast-conserving surgery. Results: Age ranged from 25 to 75 years, and the mean age was 49.9±10.99 years. The most common stage was T2 (156, 53.9%), which mostly comprised luminal A (105, 36.3%). The results showed a high frequency of N1 (109, 37.2%), with luminal A (69, 23.4%). Relapse occurred in 35/299 (11.7%) patients, and the chest wall was a common relapse site in 9 of these patients (25.9%). Conclusions: We conclude that adjuvant radiotherapy reduces locoregional recurrence, distant metastasis and mortality rate.",
"keywords": [
"3D conformal radiotherapy",
"Breast cancer",
"Progression free survival",
"Luminal subtypes",
"HER2 neu receptors"
],
"content": "Introduction\n\nBreast cancer is a common cancer in women, and it is estimated that one in eight women in the US will develop it in their life1. It is the leading cancer among women in both Europe and the US and has become an emerging disease in developing countries2,3. In Iraq, 3845 patients were estimated to have breast cancer in 20114. Therapy requires a multidisciplinary team involving surgeons, medical oncologists, radiation oncologists, pathologists, radiologists, and supportive care personnel1. Adjuvant radiotherapy decreases locoregional relapse, increases survival and palliates symptoms; after breast-conserving surgery (BCS), it is an essential component of treatment1. In oncology, progression free survival (PFS) referred to the length of time during and after the treatment of a cancer, that a patient lives with it but it does not get worse and refers to situations in which a tumor is present, as demonstrated by laboratory, radiology, and/or clinical evidence5. In developing countries, survival rates are poorer than those in developed countries6. Approximately 30% patients with early-stage cancer develop metastases, whereas metastatic breast cancer occurs in 6 – 7% of newly diagnosed patients7. The 5-year survival rate with metastasis is 25%, and the overall survival (OS) is reported to be 24 months8. Adjuvant radiotherapy reduces the incidence of locoregional recurrence from 30% to 10.5% at 20 years and decreases deaths by 5.4% at 20 years9.\n\nThis study aimed to assess PFS in patients treated by hypofractionated three-dimensional conformal radiotherapy (3DCRT) and correlate PFS with patients' clinical and pathological profiles.\n\n\nMethods\n\nWe retrospectively reviewed 299 women who were consecutively treated at Baghdad Radiation Oncology Center between October 2017 and May 2018. Patients were included if they planned to undergo adjuvant 3DCRT. Medical records were reviewed in detail after approval was obtained from our institution. All women were treated after completing chemotherapy cycles in a period not less than 13 weeks from the last cycle.\n\n1. T3, T4 stages.\n\n2. All patients with node positive status.\n\n3. All patients after breast conservative surgery.\n\n4. Multifocal breast tumors, extensive DCIS, central tumors in a small breast and incomplete excision.\n\n1. Significant pre-existing cardiac or lung disease.\n\n2. Scleroderma.\n\n3. Limited shoulder mobility.\n\n4. M1 breast cancer (metastasis).\n\nIndividual data and information on primary and advanced disease were collected. The following variables were studied as possible prognostic factors for specific clinical outcomes: age, tumor-node-metastasis staging, lymph nodes (LN), histopathology, grades, estrogen receptors, progesterone receptors, human epidermal growth factor receptor 2 (HER2) neu, surgery types, body mass index (BMI), period between the last cycle of chemotherapy and radiotherapy, date of radiotherapy, follow-up time, and metastatic or recurrent events (brain, supraclavicular LN, axillary LN, liver, chest wall scars and bones).\n\n1. 4005 cGy in 15 fractions over 3 weeks adopted in mastectomy, the indications are10–13:\n\na. Involvement of one or more axillary lymph nodes.\n\nb. T3 (>5 cm tumor), T4 (skin/chest wall invasion) or stage III tumor.\n\nc. Positive surgical margin.\n\nd. Gross residual disease.\n\ne. Multiple primary tumors (multi-centricity).\n\n2. 4005 cGy / 15 fractions / 3 weeks performed for breast conserving surgery BCS plus 1000 cGy / 5 fractions / 1 week as booster10,11,13.\n\nData were first entered into an Excel file, then transferred for statistical analysis into SPSS v22. For categorical variables, such as age, histo type, ER, PR, HER2, the frequency distribution calculated and analyzed with Fisher’s test and chi-squared. Correlation coefficient (r) was used to detect significance relations. Kaplan-Meier survival curves for PFS were used, which is a way of graphically displaying the time until the study developed an endpoint, often death, or an event such as recurrence of cancer, which was obtained during follow-up.\n\nWritten informed consent was obtained from the patients, for the use of their data in this study, and the study was conducted according to the ethical standards established by the 1964 Declaration of Helsinki. The Medical Ethical Committee of College of Medicine, Baghdad University approved this study (code: 6/2018), which covers the Baghdad Radiation Oncology Center.\n\n\nResults\n\nThe majority of patients were in the age group 46–55 years (115, 38.5%). BMI revealed that the majority of the patients were either overweight (60, 27.2%) or had moderate obesity (73, 33%) (Table 1).\n\n*Total sample sizes and numbers missing may differ due to missing values for selected variables.\n\n**Relapsing include: chest wall relapse, liver metastasis, bone secondaries, axillary LN relapse, lung metastasis and supraclavicular LN relapsed.\n\nLuminal A breast cancer type present in 197 patients (67.2%), while non-luminal A phenotypes were recorded as luminal B subtype (34, 11.6%), triple negative (33, 11.3%) and HER2 enriched (29, 9.9%) (Table 1).\n\nThe T2 stage was predominant (156, 53.9%), which mostly comprised luminal A (105, 36.3%). Other stages and subtypes presented in different proportions (Table 2). The results showed a high frequency of N1 in 109 patients (37.2%), related to molecular luminal A (69, 23.4%) (Table 2).\n\n*Total sample sizes and numbers missing may differ due to missing values for selected variables.\n\nA significant result was obtained between age, BMI, and T stage when correlated to molecular subtypes, ER and HER2 neu receptors. Age correlated to molecular subtypes (r = +1); there was good correlation (2-tailed = 0.46 and P = 0.043). BMI was significantly correlated to estrogen receptors (ER) (r = +1, 2-tailed = 0.5 and P = 0.046). Furthermore the correlation between T stages and HER2 neu receptors showed a significant value (r = +1, 2-tailed = 0.99 and P = 0.001).\n\nRelapse occurred in 35 (11.7%) women, while the reminder (264, 88.3%) were non-relapsed (Table 1). The waiting period from last cycle of chemotherapy to the date of starting radiotherapy in weeks of this study had median 17 weeks and mode about 9 weeks, which ranged from 4 to 30 weeks (Figure 1).\n\nPFS in this study presented in 264 women, while relapsing was evident in 35 cases. Chest wall relapse occurred in 9 patients (25.7%). Other relapses included liver metastasis, bone secondaries, axillary LN relapse, lung metastasis and supraclavicular LN in 8 (22.9%), 6 (17.1%), 5 (14.3%), 4 (11.4%), 3 (8.6%) patients, respectively (Figure 2).\n\nThe luminal A subtype relapsed in 69% of patients which represented major subtypes reliable for relapse, while HER2 enriched and triple negative subtypes occurred in 14% patients for both and 3% for luminal B (Figure 3).\n\nKaplan Meier survival curve estimation of PFS is shown in Figure 4. The PFS rates for relapse were 35 (11.7%) versus 264 (88.3%) for non-relapse, estimated using this curve, which indicated that when a woman received adjuvant radiotherapy, she was about nine times more likely to survive beyond the time than someone who did not receive adjuvant radiotherapy. The follow-up period was 12 months; the Kaplan Meier estimates of PFS for all the patients was censored at 9 months and the last was at 35 months. There was a trend for better PFS in patients receiving adjuvant radiotherapy, so radiotherapy was significantly associated with prolonged PFS.\n\n\nDiscussion\n\nIn the present study, the age of patients was 25–75 years with a mean± standard deviation of 49.9±10.99 years, with the majority of patients in the group of 46–55 years (115, 38.5%). This is similar to the results of previous studies conducted in Iraq: Al-Rawaq et al., 201614, Al-Khafaji et al., 201015, and Al-Naqqash et al., 200916. Age represents an important factor both for occurrence and management of breast cancer17. In our study, the median age of onset was > 10 years younger than that shown in studies in Europe and North America18. It has been proposed that these differences are due to differences in exposure to hormones, diet, physical activities, and other risk factors17. In most Arabian countries, breast cancer is commonly diagnosed in women younger than 50 years, which is consistent with the findings of our study, unlike in the US, where women aged 50 years and older are most commonly affected18.\n\nThe molecular subtype plays an important role in follow-up survival and prognosis in breast cancer. Luminal A (67.2%), luminal B (11.6%), triple-negative (11.3%), and HER2-enriched (9.9%) subtypes were observed in the present study. These results are similar to those found in Cheang’s study18 and unlike those of other studies, such as Al-Naqqash et al., 200916, Al-Sarraf et al., 201519, and El-Fatemi and Chahbounil, 201220. In the present study, luminal A relapse was recorded in 69% of patients, while the reminder were non-luminal A; HER2-enriched, triple-negative and luminal B occurred in 14%, 14%, and 3% of patients, respectively. These results were inconsistent with the results of Fitzgibbons et al., 200021 and Al-Sarraf et al., 201519. Luminal A has the most favorable prognosis, with locoregional relapse rates of only 3–8% at 10 years20. HER2-enriched groups exhibit the highest rates of local recurrence and regional relapse21.\n\nTumor size (T stage) was directly correlated with survival and metastasis in the present study. Size rank is among the strongest predictors of metastasis, disease-free survival (DFS), and OS. Although tumor size correlates strongly with the presence of axillary LNs, this is clearly an independent prognostic factor5,21. Among those with documented negative nodes, tumor size remained a strong and independent predictor5. Many studies showed a 20 year DFS of 79% for those with tumors smaller than 2 cm, compared with 64% for those with tumors larger than 2 cm5. This study demonstrated T2 stage was the most common (53.9%), followed by T3, T1 and T4; regarding the T stage related to molecular subtype, T2/luminal A was most common (36.3%). These results were similar to those of Al-Rawaq et al., 201614, Al-Khafaji et al., 201015, and Al-Naqqash et al., 200916, but different than those of Goldhirsch et al., 201322 and Cheang et al., 200918.\n\nThe status of LNs is an important prognostic factor related to survival and predictors of systemic micro-metastases22, disease recurrence, and poor prognosis23. In the present study, the N1 frequency was 37.2% followed by N2, N0, and N4. The N1/luminal A frequency was 23.5%, which was recorded commonly, whereas the N3/luminal B frequency was the lowest (0.6%). These agree with the results of Al-Rawaq et al., 201614 and Al-Khafaji et al., 201015, while they disagree with those of Goldhirsch et al., 201322 and Cheang et al., 200918.\n\nIn the present study, the PFS, which was obtained using Kaplan Meier survival curves, estimated an association between receiving adjuvant radiotherapy and/or locoregional or recurrence rates and/or distant metastasis and/or the end point (death). The estimated resampling study findings in the Early Breast Cancer Trialists' Collaborative Group (EBCTCG) in 2011, found that after BCS, radiotherapy halved the rate of recurrence and reduce the death rate by about a sixth24, which was similar to that reported in the EBCTCG study in 2014, which found that radiotherapy beyond mastectomy and axillary dissection reduced both recurrence and mortality25.\n\nIn the present study, non-relapse represented a large group (88.3%), whereas relapse was recorded in 11.7%, after a follow-up period from 6 to 12 months. Both proportional and absolute reductions in the recurrence rate are large in the first year, whereas the reduction in death becomes definite after the first few years5.\n\n\nConclusions\n\nThe fifth decade of life is the most common age of presentation for breast cancer in this Iraqi population, and the luminal A phenotype is the most common molecular subtypes. The commonest stage is T2 and N1, while the least common is T4 and N3. There was a strong correlation between age, BMI, and staging for molecular subtypes and hormonal status. Adjuvant radiotherapy treatment reduced locoregional recurrence, distant metastasis and death rates, and the period of waiting from last chemotherapy cycle to the date of radiotherapy represented a risk factor that affected the survival curve. Chest wall recurrence is most common site of disease progression.\n\n\nData availability\n\nZenodo: Progression free survival in Iraqi breast cancer patients treated with adjuvant 3D conformal radiotherapy: A cross-sectional study, http://doi.org/10.5281/zenodo.252852926.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nMurthy RK, Valero V, Buchholz TA: Breast cancer, overview. In: Gunderson and Tepper (editors). Clinical Radiation Oncology. 4th edt. Netherlands, Elsevier, Inc. 2016; 1284–1299. Reference Source\n\nSiegel R, Ma J, Zou Z, et al.: Cancer statistics, 2014. CA Cancer J Clin. 2014; 64(1): 9–29. PubMed Abstract | Publisher Full Text\n\nFerlay J, Parkin DM, Steliarova-Foucher E: Estimates of cancer incidence and mortality in Europe in 2008. Eur J Cancer. 2010; 46(4): 765–781. PubMed Abstract | Publisher Full Text\n\nIraqi Cancer Registry. Ministry Of Health, Iraqi Cancer Board, Baghdad. 2011. Reference Source\n\nDancey JE, Dodd LE, Ford R, et al.: Recommendations for the assessment of progression in randomised cancer treatment trials. Eur J Cancer. 2009; 45(2): 281–9. PubMed Abstract | Publisher Full Text\n\nCheng YC, Ueno NT: Improvement of survival and prospect of cure in patients with metastatic breast cancer. Breast Cancer. 2012; 19(3): 191–199. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDawood S, Broglio K, Gonzalez-Angulo AM, et al.: Trends in survival over the past two decades among white and black patients with newly diagnosed stage IV breast cancer. J Clin Oncol. 2008; 26(30): 4891–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKiely BE, Soon YY, Tattersall MH, et al.: How long have I got? Estimating typical, best-case, and worst-case scenarios for patients starting first-line chemotherapy for metastatic breast cancer: a systematic review of recent randomized trials. J Clin Oncol. 2011; 29(4): 456–463. PubMed Abstract | Publisher Full Text\n\nJabbari S, Park C, Fowble B: Breast Cancer. In: Hansen RK and Roach III M (editors). Handbook of Evidence-Based Radiation Oncology. 2nd ed. Springer Science+Business Media, LLC. CA. USA. 2010; 263–311. Publisher Full Text\n\nSTART Trialists' Group, Bentzen SM, Agrawal RK, et al.: The UK Standardisation of Breast Radiotherapy (START) Trial B of radiotherapy hypofractionation for treatment of early breast cancer: a randomised trial. Lancet. 2008; 371(9618): 1098–107. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSTART Trialists' Group, Bentzen SM, Agrawal RK, et al.: The UK Standardisation of Breast Radiotherapy (START) Trial A of radiotherapy hypofractionation for treatment of early breast cancer: a randomised trial. Lancet Oncol. 2008; 9(4): 331–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHaviland JS, Owen JR, Dewar JA, et al.: The UK Standardisation of Breast Radiotherapy (START) trials of radiotherapy hypofractionation for treatment of early breast cancer: 10-year follow-up results of two randomised controlled trials. Lancet Oncol. 2013; 14(11): 1086–1094. PubMed Abstract | Publisher Full Text\n\nFAST trialists group, Agrawal RK, Alhasso A, et al.: First results of the randomised UK FAST Trial of radiotherapy hypofractionation for treatment of early breast cancer (CRUKE/04/015). Radiother Oncol. 2011; 100(1): 93–100. PubMed Abstract | Publisher Full Text\n\nAl-Rawaq MK: Molecular Classification of Iraqi Breast Cancer Patients and Its Correlation with Patients’ Profile (Observational Study).Thesis. Baghdad-Iraq. Baghdad-Iraq. University of Baghdad College of Medicine. 2016. Reference Source\n\nAl-Khafaji AH: Immunohistochemical expression of Estrogen, Progesterone receptors, P53 and Ki67 in Iraqi and Syrian breast cancer patients, A clinicopathological study. Thesis. Baghdad-Iraq. Baghdad-Iraq. University of Baghdad College of Medicine. 2010.\n\nAl-Naqqash MA: The role of c-myc oncogene as a prognostic marker in breast cancer patients evaluated by immunno-histochemistry and in situ hypridization (prospective study). Thesis. Baghdad-Iraq. University of Baghdad College of Medicine. 2009.\n\nOussama MNK: Guidelines for the early detection and screening of breast cancer: EMRO.Technical Publications Series 30. WHO. 2006. Reference Source\n\nCheang MC, Chia SK, Voduc D, et al.: Ki67 index, HER2 status, and prognosis of patients with luminal B breast cancer. J Natl Cancer Inst. 2009; 101(10): 736–750. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAl-Sarraf FS: Immunohistochemical Expression of ER, PR, Her2/neu and Ki67 in Breast carcinoma. Clinico-pathological Study. Thesis. Baghdad-Iraq, University of Baghdad College of Medicine. 2015.\n\nEl-Fatemi H, Chahbounil S, Jayi S, et al.: Luminal B tumors are the most frequent molecular subtype in breast cancer of North African women: an immunohistochemical profile study from Morocco. Diagn Pathol. 2012; 7: 170. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFitzgibbons PL, Page DL, Weaver D, et al.: Prognostic factors in breast cancer. College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med. 2000: 124(7): 966–978. PubMed Abstract\n\nGoldhirsch A, Winer EP, Coates AS, et al.: Personalizing the treatment of women with early breast cancer: highlights of the St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2013. Ann Oncol. 2013; 24(9): 2206–2223. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPegram MD, Takita C, Casciato DA: In: Casciato DA and Territo MC (editors). Manual of Clinical Oncology. 7th edt. Lippincott Williams & Wilkins, a Wolters Kluwer business. USA. 2012; 285–319.\n\nEarly Breast Cancer Trialists' Collaborative Group (EBCTCG), Darby S, McGale P, et al.: Effect of radiotherapy after breast-conserving surgery on 10-year recurrence and 15-year breast cancer death: meta-analysis of individual patient data for 10,801 women in 17 randomised trials. Lancet. 2011; 378(9804): 1707–1716. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEBCTCG (Early Breast Cancer Trialists' Collaborative Group), McGale P, Taylor C, et al.: Effect of radiotherapy after mastectomy and axillary surgery on 10-year recurrence and 20-year breast cancer mortality: meta-analysis of individual patient data for 8135 women in 22 randomised trials. Lancet. 2014; 383(9935): 2127–2135. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlnaqqash M, Abdulfattah W, Albdear E, et al.: Progression free survival in Iraqi breast cancer patients treated with adjuvant 3D conformal radiotherapy: A cross-sectional study. 2018. http://www.doi.org/10.5281/zenodo.2528529"
}
|
[
{
"id": "45102",
"date": "12 Mar 2019",
"name": "Angel Montero-Luis",
"expertise": [
"Reviewer Expertise Clinical oncology",
"radiation oncology",
"brachytherapy"
],
"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 a very interesting job. The authors have conceived a serious and exhaustive analysis of their own data. The study is well justified, and the characteristics of the patients and the intervention are clear and concise, both in the text and in the tables. The results are in agreement with the existing data in the current scientific literature.\nHowever, the authors state that the observed progression-free survival data allows them to argue that adjuvant radiotherapy reduces the risk of recurrence and significantly increases PFS, which is a gratuitous statement since it has not been compared to any control arm without radiotherapy and, therefore, the conclusion is only an extrapolation of the data of a particular series.\nIt would be much more correct for the authors to limit their conclusions to presenting their own data, very satisfactory and consistent with the evidence, rather than maintaining statements that cannot be substantiated scientifically in the analyzed data. Finally, the article is well structured and written and the bibliography is consistent and updated.\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? Partly",
"responses": []
},
{
"id": "64749",
"date": "23 Jul 2020",
"name": "Julie A. Bradley",
"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\nOutcomes of women with breast cancer in Iraq is an important, and the authors have done substantial work collecting and analysing the data.\nThe primary area that is unclear is whether all patients in this cohort had radiation, or some did not. If all had radiation, then it cannot be concluded from this data that radiation improves disease outcomes, as there is no control group for comparison. The outcomes of this series can be compared to historical published series of no radiation. This statement is not substantiated by the data presented: “The PFS for relapse were 35 vs 264 with non-relapse… which indicated that when a woman received radiation therapy, she was 9 times more likely to survive… than someone who did not receive adjuvant radiotherapy.” If all patients received radiation, then this comparison cannot be performed. If some did not receive radiation, then the numbers for those who received and did not receive radiation need to be provided.\nWere recurrences more common in those with positive lymph nodes? What percent of patients had recurrence after having undergone lumpectomy versus mastectomy?\n\nDid patients receive adjuvant endocrine therapy if they were estrogen/progesterone receptor positive?\n\nSpecifics of the radiation techniques would be very helpful. Was CT simulation used? Were contours drawn? What energy photons? Was bolus used for the chest wall? What nodal levels were treated (ie were internal mammary nodes included)? What were the indications for nodal radiation?\n\nWhat type of chemotherapy was used? What % of patients received chemotherapy? How was the axilla addressed – what % of patients had sentinel lymph node biopsy versus axillary lymph node dissection?\n\nThis data is interesting and important, and can be strengthened by consideration of the above points.\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": []
}
] | 1
|
https://f1000research.com/articles/8-71
|
https://f1000research.com/articles/8-68/v1
|
17 Jan 19
|
{
"type": "Research Note",
"title": "Exposure to nickel from the metal equipment in the gym",
"authors": [
"Denis Vinnikov",
"Zhangir Tulekov",
"Anar Dushpanova",
"Zhanna Romanova",
"Aleksandr Sokolov",
"Valeriya Sokolova",
"Tair Nurpeisov",
"Zhangir Tulekov",
"Anar Dushpanova",
"Zhanna Romanova",
"Aleksandr Sokolov",
"Valeriya Sokolova",
"Tair Nurpeisov"
],
"abstract": "It remains unclear whether gym customers are exposed to any nickel from the metal equipment and if the exposure is associated with the duration of contact. Therefore, the aim of this study was to ascertain exposure to nickel measured through nickel concentration in the hair in those exercising in a fitness gym. We enrolled 100 amateur athletes in one of the gyms in Almaty, Kazakhstan (all men, median age 30 (interquartile range (IQR) 10) years), exercising from 2 to 7 days a week for 40 to 180 minutes and their age- and sex-matched controls who did not exercise. All subjects filled in the questionnaires on the exercising patterns, smoking and occupational exposure and then donated 0.25 g of head hair, in which nickel was measured using atomic absorption spectrophotometry. Hair nickel concentration ranged from 0 to 8.5 µg/g with notable left-skewness towards low concentrations in both groups. Hair nickel concentration was not associated with age, smoking or occupation, but was significantly lower in amateur athletes compared to controls (median 0 (IQR 0.5) vs. 0.9 (IQR 1.4) µg/g). More days a week in a gym, longer workout history, longer workout duration or supplements use did not increase the probability of being stratified in a high-exposure subgroup (defined as 75th percentile of hair nickel concentration and higher); however, there were more smokers in a low-exposure group (p<0.05). With the mixed pattern of exposure, gym goers may be unlikely exposed to more nickel from the metal equipment in the gym, however the exposure may depend on the specific alloy composition.",
"keywords": [
"gym",
"nickel",
"spectrophotometry",
"exposure"
],
"content": "Introduction\n\nExercising in a gym has become quite a prevalent leisure activity in adults, given significant benefit of regular exercising on cardiovascular health, mood and probably self-confidence. Little is known, however, about the adverse effects of attending the gym with the corresponding exposures. Gym goers may be exposed to a variety of chemicals inside gyms including metal bars and barbells. Those are usually produced of stainless steel, and the chemical composition of the latter may vary; however, selected metals in the steel, such as nickel, may be associated with adverse health effects in environmental and occupational studies. Thus, nickel is a known carcinogen1,2 and may also cause allergic dermatitis.\n\nVery few studies, however, assess exposure of those exercising in the gyms to nickel. There is only one report with a small sample to show higher concentrations of nickel from the contact with bars compared to non-exercising individuals3. In this presentation, they found nickel both on the bars and the skin of exercising individuals using acid test. A case of allergic dermatitis was also reported in a regularly exercising individual4, but no studies with larger samples have approached the issue of nickel exposure and its adverse health effects in the gyms. Whether nickel found in the palm skin of gym customers may lead to higher nickel blood concentrations, therefore causing systemic effects, remains unknown. In professional sportsmen, nickel blood concentrations were found to be higher compared to controls5, and given that nutritional intake of nickel in these two groups did not differ, such findings should raise some concern as to whether such elevated concentration is associated with exposure in the gyms and has any negative impact on health.\n\nWith all that scarce evidence, it remains unclear whether gym customers are exposed to any nickel from the metal equipment and if the exposure is associated with the duration of contact. Therefore, the aim of this study was to ascertain exposure to nickel measured through nickel concentration in the hair in those exercising in a fitness gym compared to non-exercising controls.\n\n\nMethods\n\nThis study was approved by the Committee on Bioethics of al-Farabi Kazakh National University. All subjects in this study provided written informed consent to participate and donate head hair sample for nickel analysis.\n\nWe enrolled 100 customers from two gyms in of one the popular chains in Almaty, Kazakhstan. Anyone willing to participate and regularly exercising in a gym could be included in the study. The only exclusion criterion was female sex, as there were very few women in the gym. Subjects were invited to participate by authors DV, ZhT or AD in a random fashion, thus, reducing selection bias. Data were collected in July and August 2018. Sample size calculation with a given statistical power did not seem feasible for this study, as we could not find any other similar analysis of this kind in the literature; therefore, we set the sample size of 100 subjects. We also enrolled sex- and age-matched controls who were their friends or acquaintances to ensure comparable lifestyle, eating habits and general interests to control for confounding. Controls should not have exercised in a gym for at least 2 years prior to the enrollment in the study. All subjects were asked to fill in a questionnaire6, which consisted of the demographic part, followed by detailed section in the exercising pattern, smoking, occupational exposure and the use of supplements. We asked the respondents how many days they normally attended the gym, what the duration of the usual workout was in minutes, how long was the gym exposure history, whether gloves were used in the gym, whether a contact allergy to metals in the gym was ever experienced, and whether any fitness tracker or supplement was used.\n\nWe then detailed smoking history with a series of questions and stratified all subjects into never, former or daily smokers and ascertained the number of smoked cigarettes a day along with the smoking durations in months or years. Occupational history section contained a series of questions whether a subject was a student at a time of the survey, had any employment in the office or had any occupational exposure with metal.\n\nWe measured hair nickel concentrations in all subjects and treated the concentration as a marker of exposure to nickel. Hair (at least 0.25 g) was cut in from the occipital. Hair samples were then washed using non-ionized surface-active solution, then acetone and then with non-ionized water. Weighed samples were treated with nitric acid (67%) and hydrogen peroxide (30%). Nickel concentrations in the samples were tested using atomic absorption spectrophotometry on Perkin Elmer AAnalyst 400 with HGA 900 (USA) and following officially approved protocol7. The lower limit of detection (LOD) in our analysis was 0.05 µg/g.\n\nHair nickel concentrations were the primary outcomes in this analysis and compared in the main and control groups using non-parametric Mann-Whitney U-test, since all concentrations were left-skewed. Demographic attributes, smoking and occupational exposure were tested as predictors in bivariate models and compared between the groups. We used NCSS 12 (Utah, USA) for all computations. P<0.05 was considered significant.\n\n\nResults\n\nHair nickel concentration ranged from 0 to 8.5 µg/g, with notable left-skewness towards low concentrations. Thus, 25th percentile was 0 µg/g; 50th, 0.38 µg/g; and 75th, 1.22 µg/g. A total of 56 (28%) subjects showed concentrations below the limit of detection (LOD); 52% of amateur athletes and 4% of those in the control group had nickel levels below LOD (p<0.05). Hair nickel concentration was not associated with age, smoking or occupation, but was significantly lower in amateur athletes than controls (Table 1). Raw information of hair nickel concentration, in addition to all questionnaire answers, are available on OSF6.\n\n*P<0.05, gym-goers compared to control.\n\nIn the main group, the gym attendance frequency ranged from 2 to 7 days a week; however, most subjects did so 3 times a week (67%). Workout duration ranged from 40 to 180 minutes; median, 90 (IQR 37.5) minutes. The overall gym exposure was from 1 month to 30 years, with the median 2 (IQR 4.4.) years. Only 17% of those in the gym used gloves for weightlifting on a regular basis, and 4% ever had dermatitis that they associated with the gym equipment use. A total of 16% use fitness tracker in the gym on a regular basis, and 56% use any sort of supplements to attain more visible results in the gym; there was no statistically significant correlation between these two variables.\n\nWhen stratified by the 75th percentile of nickel hair concentration (0.505 µg/g) into low- and high-exposure gym customers only, we found no difference in age, the number of workouts days a week, workout duration, overall exposure to gym equipment in years or supplement use (Table 2). Surprisingly, there were significantly more subjects wearing gloves, believed to protect the skin from contact with metal, in the high-exposure group. Similarly, the latter group had fewer smokers compared to those with lower nickel concentrations.\n\n*P<0.05, high-exposure compared to low-exposure group.\n\n\nDiscussion\n\nTo our knowledge, this is the first report on hair nickel concentrations in those attending the gym compared to controls, in which we could not confirm higher exposure to nickel in amateur athletes. Guided by the pilot presentations that exposure to metal equipment in the gym may result in greater nickel absorption, we compared hair nickel concentrations in regular exercisers compared to those abstaining from the gym, but found higher hair nickel concentrations in the latter group. We conclude that it was not dermal contact with metal equipment in the gym, but fewer smokers or specific nutritional habits in the gym goers group that could explain their lower hair nickel concentrations.\n\nThe sources of nickel in the population may range from absorption to food, smoking, place of residence, lifestyle habits, such as exposure, to diverse occupational exposures. Despite some likelihood of exposure to nickel in those exercising in the gym, we could not find similar reports in the literature and could not compare the concentrations we found with other settings. However, there are plenty of other environmental and occupational publications with reported hair nickel concentrations.\n\nThe most surprising finding of this analysis was nickel concentration in controls. Although we deliberately matched controls with exercisers to ensure similar eating patterns, their hair nickel concentrations were quite high and even exceeded the concentrations in occupationally exposed industrial workers8. In order to allow for comparison between those exercising in the gym and controls, nutritional nickel consumption should be equal in both groups. Direct assessment of the amount of consumed nickel does not seem feasible in a regular setting; therefore, computational methods are often used in studies of athletes5. However, such methods yield more approximation than accuracy and therefore will lead to a notable exposure classification bias. Hence, in our study, we preferred to enroll controls from friends, matched for age and sex, to allow for comparable nickel consumption in the main group with controls.\n\nThe limitations of this analysis originate from its cross-sectional design. The overall sample size of 200 subjects may also limit statistical power. Another limitation is the use of matching rather than a detailed questionnaire on eating habits and computational method to ascertain food nickel consumption. Finally, we could not obtain detailed information on the metal composition of the steel used for a particular brand of metal equipment in the chain of gym under study. Guided by anecdotal reports in non-professional literature, stainless steel for metal equipment in the gym is very likely produced of steel with some nickel content, but we could not confirm whether the given equipment had any nickel in it, either from the original documentation or, alternatively, using acid nickel testing.\n\nTo conclude, this pilot study of nickel exposure measured through hair nickel concentration in those contacting metal equipment in the gym failed to demonstrate greater hair nickel concentration in the latter compared to their non-exercising friends.\n\n\nData availability\n\nRaw data for this study, including basic demographic information, answers to the questionnaire and hair nickel levels, are available on OSF. DOI: https://doi.org/10.17605/OSF.IO/RQJ3Z6.\n\nThe questionnaire in the original (Russian) and in English are available on OSF. DOI: https://doi.org/10.17605/OSF.IO/RQJ3Z6.",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nSeilkop SK, Lightfoot NE, Berriault CJ, et al.: Respiratory cancer mortality and incidence in an updated cohort of Canadian nickel production workers. Arch Environ Occup Health. 2017; 72(4): 204–219. PubMed Abstract | Publisher Full Text\n\nGrimsrud TK, Berge SR, Haldorsen T, et al.: Exposure to different forms of nickel and risk of lung cancer. Am J Epidemiol. 2002; 156(12): 1123–1132. PubMed Abstract | Publisher Full Text\n\nGumulka M, Matura M, Lidén C, et al.: Nickel exposure when working out in the gym. Acta Derm Venereol. 2015; 95(2): 247–249. PubMed Abstract | Publisher Full Text\n\nLedon JA, Tosti A: CrossFit-Associated Allergic Contact Dermatitis. Dermatitis. 2017; 28(6): 368. PubMed Abstract | Publisher Full Text\n\nMaynar M, Llerena F, Bartolomé I, et al.: Seric concentrations of copper, chromium, manganesum, nickel and selenium in aerobic, anaerobic and mixed professional sportsmen. J Int Soc Sports Nutr. 2018; 15: 8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVinnikov D: Hair nickel concentration in gym goers. 2019. http://www.doi.org/10.17605/OSF.IO/RQJ3Z\n\nIron, zinc, nickel, copper and chrome determination in hair using atomic absorbtion. Methodological recommentations. МУК 4.1.776-99. Ministry of Health of Russian Federation; 1999.\n\nVinnikov D, Semizhon S, Rybina T, et al.: Occupational exposure to metals and other elements in the tractor production. PloS One. 2018; 13(12): e0208932. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "69535",
"date": "23 Sep 2020",
"name": "Maria Jose Gonzalez-Munoz",
"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\nIt is a very novel and interesting work, but a bit poor in its methodology and conclusions. I send you the doubts:\nMaterials and methods Subjects were invited to participate by authors DV, ZhT or AD in a random fashion, thus, reducing selection bias. I do not understand\n\nDetermination of Zn With what instrument was the sample collection carried out? Scissors? What material? Was it taken into account if the hair was dyed or not? It could indicate the analytical parameters used for the determination of Zn by AAS. The determinations were made in duplicate? Was a standard reference used? Which one?\n\nResults I don't understand the Tables, what do the numbers refer to? ¿ The authors send the data on the raw Zn levels in hair and the questionnaire used to another publication. I think it is not very comfortable to understand.\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? Partly\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? Yes",
"responses": []
},
{
"id": "71260",
"date": "28 Sep 2020",
"name": "Muhammad Abdul Qayyum",
"expertise": [
"Reviewer Expertise Metal toxicology on carcinogenics/diseases",
"Phytochemistry"
],
"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\nReviewer comments: Reference to the Manuscript Number: F1KR00CDE F1R-VER19451-R (end code), the authors describes Ni metal and its exposure to people in the gym and in relation to controls. The authors did good efforts to perform this kind of work which may be new one in their country/region. There are some minor revisions which ought to be correct before indexing.\nAbstract Line….measured using atomic absorption spectrophotometry…. (Should add). The hair samples were digested/mineralized with the mixture of HNO3 and H2O2 prior to analysis. Hair nickel concentration ranged from 0 to 8.5 μg/g….\nIntroduction The authors should write/add at least one paragraph which is comprised of hazards or toxicity of Nickel on human beings with references.\nMethods The authors should mention full name of city and country as in this sentence ….This study was approved by the Committee on Bioethics of al-Farabi Kazakh National University……. What this line explain…….participate by authors DV, ZhT or AD in a random fashion? The authors indicate occupational exposure. I suggest authors should shown in name of occupational exposure as exposed by participants and also shown by duration in the table 1. In statistical analysis p< 0.05, p should be in italic.\n\nHair nickel concentration measurement The authors should write as followed or may your own style/words as well the method they adopted. This is a method for reference. In the present study, scalp hair specimens (at least 0.25 g) were collected from the occipital region of the head with a pair of stainless-steel scissors. The collected hair samples were put into small polyethylene bags, labelled with relevant codes and stored at room temperature until digestion and analysis of metals were performed. The hair samples were thoroughly washed to offer an accurate assessment of endogenous metal contents. Before washing, the samples were cut into small pieces (approximately 0.5 cm) and mixed to make a representative sample. Afterwards, each hair sample was washed in series with 5% detergent solution, 0.5% Triton X-100 solution and deionized water. First of all, the scalp hair sample was taken in a conical flask containing 50 mL of 5% detergent solution and mixed well. The flask contents were then shaken on an auto-shaker at 320 vibrations per minute for about 30 min. After leaving it at room temperature for at least 2 h, it was washed with plentiful water. Then, 30 mL of non-ionic detergent Triton X-100 (0.5% v/v) solution was added to each flask and again placed on the auto-shaker for 30 min. The samples were then washed with deionized water followed by drying in an electric oven overnight at 70 °C (Reference). Similarly explain, the authors should write the complete process of digestion/mineralization e.g. heating, heating source and from what temperature used to complete process, duration of time as well as reference.\n\nResults Explain Table 1 completely in results section. For example and for reference.\n\nCharacteristics of the Study Subjects The demographic parameters related to the stomach cancer patients and healthy donors are displayed in Table 1. Stomach cancermalignancy was confirmed histopathologically along with clinical examination. The age of the stomach cancer patients ranged from 17 to 63 years with a mean value of 45 years while for healthy donors, it ranged from 15 to 65 years with a mean value of 42 years. Majority of the participants (> 50%) in both groups were vegetarians and 63%of the patients and 64% of the healthy donors resided in the rural areas. More than 50% of the patients and healthy subjects were not addicted to tobacco (smoking). Most of the patients (68%) were suffering from adenocarcinoma. Based on the division of histopathological stage, 32% of the patients were diagnosed at stage II, 25% at stage I, and 23% at stage III while 20% at stage IVof stomach cancer in the present study.\n\nAlso demonstrate table 2 in results more clearly. In Table 1. Demographic profile and nickel concentrations in the amateur athletes and controls. Please correct. Age, years 29 59 (8.8) 30 (10) 29 (9)\nSummary Giving conclusions, the article has clear objective, approach is appropriate, However, the introduction does not provide any background on Ni toxicity/hazards/disease. The experiments and analyses performed with technical rigor to allow confidence in the results. But should be explain step by step. E.g., collection and processing of hair samples. Use of measuring unit is a good step and its name should be shown in the tables as well as in the context if needed. Some global points of view, at least to discuss the results, should be highlighted from relevant published reports. The variables shown in Tables are extracted conclusive information. Occupational exposure name with durations should be added in table 1. It seems number of samples are limited which we did not drive any clinically conclusion. However, after analyzing the data, results/tables are convincing especially elemental concentration. I suggest this manuscript is suitable for indexing.\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": []
}
] | 1
|
https://f1000research.com/articles/8-68
|
https://f1000research.com/articles/8-63/v1
|
16 Jan 19
|
{
"type": "Research Article",
"title": "Antimicrobial of nyirih (Xylocarpus granatum) against pathogens of tiger shrimp post-larvae",
"authors": [
"Gina Saptiani",
"A. Syafei Sidik",
"Fikri Ardhani",
"A. Syafei Sidik"
],
"abstract": "Background: Xylocarpus granatum has been used as a medicinal plant by coastal communities, which may indicate that this plant is a potential source of pharmaceuticals. Methods: Xylocarpus granatum leaf extract was tested as an antimicrobial agent for pathogens infecting tiger shrimp post-larvae. Of the treatments applied to the post-larvae, 25 were crudely extracted with ethanol, distilled water, and seawater solvent given by immersion. Vibrio harveyi and Saprolegnia sp. were microbial species used for the test. Results: X. granatum extract had the potential to inhibit V. harveyi and Saprolegnia sp., reducing infection and improving the survival of shrimp. Shrimp soaked with X. granatum extract had a total Vibrio count ranging from 14.67x103 to 22.67x103 CFU/ml. The survival rate of shrimp was recorded as 53.33% to 78.67% and 54.67% to 76.00% due to V. harveyi, and Saprolegnia sp infection, respectively. The relative percentage of the survival of shrimp protected from V. harveyi and Saprolegnia sp infection in treatments compared to negative controls ranged from 40.61% to 72.89% and 35.84% to 66.12%, respectively. Conclusions: Leaf extracts of X. granatum, which might have better antimicrobial activities to prevent tiger shrimp from pathogenetic infection, were consecutively extracted ethanol at 800-1,000 ppm, distilled water at 800-1,000 ppm, and seawater at 1,000 ppm.",
"keywords": [
"antimicrobial",
"pathogen",
"tiger shrimp",
"Xylocarpus granatum"
],
"content": "Introduction\n\nShrimp pond culture in East Kalimantan, Indonesia, is generally conducted in a traditionally extensive system, and the filling and draining of ponds is fully dependent upon tides. Wild biota transported into ponds along with the tide water during filling might carry pathogens. Vibriosis and fungal often affect larvae and post-larvae of tiger shrimp in hatcheries and ponds in East Kalimantan1. The use of antibiotics to prevent diseases is a problem, since uncontrolled use may bring about resistance and toxicity2. As an alternative, to minimize the use of antibiotics, various plant extracts have been reported to have the ability to reduce microbial attacks in aquaculture thus reducing the risk of death3–5.\n\nVarious species of mangrove are found in the Mahakam Delta6 and are from six important genera: Pedada (Sonneratia), Api-api (Avicennia), Bakau (Rhizopora), Tancang (Bruguiera), Nyirih (Xylocarpus), and Nipah (Nypa)7. Mangroves have long been used for traditional medicine by coastal communities. Several studies have been published concerning mangrove activity on pathogens in humans, animals, and plants8.\n\nResearch using mangrove extract as a source of pharmaceutical ingredients and drugs, or as an antibacterial for tackling diseases in shrimp culture, has shown positive developments3. The use of plant extracts has been reported by many authors, proving that they are able to be utilized as an antibacterial and antifungal or as immunostimulants without causing resistance1,4,9. In this experiment, leaf extract of Xylocarpus granatum was examined as an antibacterial and antifungal material to maintain the health and survival of tiger shrimp post-larvae in captivity.\n\n\nMethods\n\nLeaves of X. granatum were obtained from a shrimp pond area in the Muara Badak Subdistrict of the Kutai Kartanegara District in East Kalimantan, Indonesia. Leaves were washed and drained until there was no water and, after being dried to around 40% of their original weight, were chopped and wind-dried in a room not exposed to direct sunlight. After about 50 days, the leaves were then macerated with three different solvents namely 80% ethanol, distilled water, and 22‰ seawater for 24 hours. The ratio between leaves and solvents was 300 g of leaves in 2,100 ml solvent. The maceration product was extracted by evaporation, and the extraction products were heated over the bath until the ethanol solvent was evaporated to obtain crude. Extractions with distilled water and seawater were stopped when the solvent reached 10% of the initial volume.\n\nMicrobes used for the challenge tests were Vibrio harveyi and the fungi Saprolegnia sp., supplied by the Laboratory of Aquatic Microbiology, Faculty of Fisheries and Marine Science, Mulawarman University, Samarinda. Before use, the pathogenicity of V. harveyi was tested by intramuscularly injecting 0.05 ml of the bacteria at a dose of 12.4×104 CFU/ml into five 3-g tiger shrimp. After 5 days, when the shrimp showing signs of redness V. harveyi was then isolated from the hepatopancreas. Furthermore, V. harveyi was isolated and cultured on Thiosulfate Citrate Bile Salt Sucrose Agar medium (Merck KGaA. 1.10263.0500) and incubated for 24 hours at 33°C. Saprolegnia sp. was rejuvenated by culturing it on Potato Dextro Agar medium (Himedia REF M096-500G) incubated at 33°C for 24 hours.\n\nSeawater with 22‰ salinity, and 28°C temperature was used as the culture medium for shrimp, and confirmed to be free from pathogens by isolating and identifying Vibrio sp and Saprolegnia sp. The seawater was deposited in a tank for 2 days and then flowed to another tank and aerated. Each aquarium was filled with 5 l sea water and aerated.\n\nThe PL-25 tiger shrimp came from a brood stock and controlled hatchery which never applies chemicals or drugs and were confirmed to be Vibrio-free after sampling for isolation and bacterial culture in the medium. Each aquarium was stocked with 25 shrimp.\n\nThe treatments were leaf extract, extracted with ethanol solvent, distilled water or seawater solvent, and each leaf extract treated to the shrimp had concentrations of 1,000, 800, 600, or 400 ppm (where 1,000 ppm is 1 ml of extract in 1,000 ml of water). Control treatments consisted of a positive control, the antibiotic erythromycin 500 mg/1,000 ml, and a negative control, NaCl 0.85%. Assessments were carried out every 6 hours.\n\nAs many as 25 shrimp of PL-25 were soaked in each extract solution at each indicated concentration in each aquarium, and each treatment was replicated three times. The challenge test with each of V. harveyi and Saprolegnia sp. in each aquarium was performed 6 days after extract-soaking by immersing a concentration of 10.6×105 CFU/ml microbes into the shrimp culture medium. The shrimp was reared for 28 days for the challenge test. Clinical symptoms, including changes in activity, appetite, and reflexes, were observed every day. Clinical symptoms were analysed by calculating the percentage of shrimp in an aquarium showing inactive (passive) motion, decreasing appetite, and weakening response of reflex. Observation of pathological anatomy (PA) was performed on dead shrimp and at the end of experiment based on changes in colour, shape, and other abnormalities in the shrimps’ bodies and organs. PA observation included the body condition, carapace, legs, uropod, hepatopancreas, and abdomen (Table 1).\n\nDead shrimp were recorded each day in order to calculate shrimp survival rate. In addition, the total content of V. harveyi proliferating in each shrimp was calculated using total plate count (TPC) method. The shrimp bacteria were isolated and cultured on Thiosulfate Citrate Bile Salt Sucrose Agar medium, after incubation for 24 hours at 33°C the number of growing V. harveyi colonies was counted. The relative percentage of survival (RPS) was also applied to know the effectiveness of leaf extract X. granatum in preventing shrimp from V. harveyi infection using the following formula:\n\nRPS = (1- (mortality of treatment shrimp) × (mortality of control shrimp)-1) × 100%\n\n\nResults and discussion\n\nThe body colour of shrimp changed darker and shifted back to normal colour again, after 24 to 36 hours of extract soaking. The change of colour indicated the intrusion of extract into the body fluids of the shrimp as osmoregulatory processes try to balance the osmotic pressure of body fluids with its surrounding environment. The extract-penetrating body fluids will stimulate the immunity mechanism of shrimp. Discolouration in the shrimp is a sign that foreign substances are penetrating the body fluids, influencing the chromatophore, which is part of a shrimp’s immunity system1,4. Colour change in crustaceans can be stimulated by many factors as a behavioural form for adaptation and protection10–12.\n\nAt 2 days after the challenge test with V. harveyi, the activities and appetites of the shrimp were decreasing. Clinical symptoms on day 14 were increasingly apparent in negative controls. Clinical symptoms of the shrimp incubated with seawater extract were more apparent than other treatments, especially at concentrations of 400 ppm and 600 ppm. Those incubated with ethanol and distilled water extract appeared to be better than the positive control. Ethanol extract 800 ppm and 1,000 ppm and distilled water extract 1,000 ppm resulted in a better average activity of shrimp. Better appetite was also shown by shrimp in the culture medium with ethanol and distilled water extract 1,000 ppm, and the reflex response of shrimp was found better in the ethanol extract 1,000 ppm. However, all clinical symptoms in the shrimp were showing improvement until day 28, except for negative controls.\n\nClinical symptoms on shrimp challenged with Saprolegnia sp. began to appear on day 6, and shrimp appeared healthier than shrimp challenged with V. harveyi. In the negative control, legs were visibly dirty, a symptom that did not appear in treatments challenged with V. harveyi. Clinical symptoms on shrimp on days 14 and 28 in all extract treatments appeared better than antibiotic-added positive controls. Clinical symptoms that appeared on shrimp for all treatments are presented in Table 2. Shrimp infected with of V. harveyi exhibited symptoms of decreasing activity, weakening reflex responses and loss of appetite1,13. Raw data behind each table is available on OSF14.\n\nBetter anatomical pathology was shown by shrimp in all extract treatments when compared to negative controls. Some individual shrimp in negative controls challenged with V. harveyi experienced anatomical pathology issues, such as: incomplete organs; reddish carapace; broken rostrum; reddish and broken legs; swollen and gripped uropod; brownish, reduced, and disturbed hepatopancreas; and brownish, hard abdomen. Shrimp in negative controls challenged with Saprolegnia sp., showed a lighter anatomical pathology than those challenged with V. harveyi. The anatomical pathology of shrimp was indicated with reddish appearance in the body, legs, and uropod or with incomplete carapace, legs, and uropod. The hepatopancreas softened, and both it and the stomach changed brownish. Vibrio sp. causes gills to turn dull pale and reddish yellow and carapace dark reddish, pleopods and uropod to break, and hepatopancreas slightly reddish to dark red15. Vibriosis brings reddish black colour on shrimp, red spots on legs and uropod, haemorrhage in the body, and deformity and moulting failure13.\n\nBased on clinical symptoms and anatomical pathology, shrimp culture in a medium with extract treatments showed a better physiological condition and resistance against for V. harveyi and Saprolegnia sp. infection, compared to negative controls. The positive control gave almost the same result as the ethanol and distilled water extract treatments. This evidence indicated that X. granatum leaf extract is supposed to be effective at preventing tiger shrimp from both V. harveyi and Saprolegnia sp. infection. Vibrio infection may occur in all phases of shrimp development, from egg to broodstock16. Vibriosis, especially caused by the luminous Vibrio, often brings about serious losses in shrimp hatcheries17. Vibrio species are abundant in a seawater environment, and some opportunistic pathogenic strains are associated with the immunity of cultured shrimp in unfavourable environmental conditions18.\n\nThe colonial density of V. harveyi (TVC) in shrimp soaked with leaf extract on the last day of the experiment ranged from 14.67×103 to 22.67×103 CFU/ml. This bacterial count indicated that leaf extract for all solvents at concentrations ranging from 400 to 1,000 ppm could inhibit the proliferation of V. harveyi better than negative controls with TVC of 25.67×103 CFU/ml. The TVC value in the positive controls included in the value range of leaf extract treatments, which was 16.00×103 CFU/ml. The strongest inhibition of extract against V. harveyi, consecutively, was distilled water extract at 1,000 ppm, distilled water extract at 800 ppm, ethanol extract of 800 ppm, and ethanol extract of 1,000 ppm, as presented in Table 3. The bacterial density of V. harveyi in water, sediment, and shrimp samples varied between 6.0×103 to 88×103 CFU/ml,1,200×103 to 10,400×103 CFU/g, and 5.0×103 to 73 ×103 CFU/ml, respectively19. Bacterial density of V. harveyi in hepatopancreas of 1.5-month-old tiger shrimp, 14 days after the challenge test with 105 CFU/ml, was about 14.67 ×103 CFU/ml.\n\nThe above facts indicated that X. granatum leaves extracted with distilled water and ethanol solvents, at concentrations between 800 to 1,000 ppm, were potentially antibacterial and able to inhibit V. harveyi growth better than antibiotics. All Vibrio isolates were found to be resistant to ampicillin, gentamycin, oxytetracyclin, chloramphenicol, trimethoprim, and kanamicin—the antibiotics commonly used in aquaculture20. Crude ethanolic extract of X. granatum, in vitro at 400 ppm, could inhibit the growth of Staphylococcus epidermis, Staphylococcus aureus, Shigella boydii, Proteus spp., Escherichia coli, and Streptococcus pyogenes21.\n\nThe survival rate of shrimp incubated with extracts of X. granatum was better than the negative control, ranging from 53.33% to 78.67% following to V. harveyi infection, and 54.67% to 76.00% following to Saprolegnia sp. infection. The survival rates in the positive control was recorded as 78.67% and 77.33%, and in the negative control were 21.33%–29.33% (Table 4). The RPS of shrimp soaked with extract, in the negative controls, and in the positive controls ranged from 40.61% to 72.89%, from 35.84% to 66.12%, and from 67.97% to 72.98%, respectively. The highest RPS against V. harveyi infection was 72.89% obtained in treatments of ethanol extract at 1,000 ppm, distilled water extract of 1,000 ppm, and positive control, followed consecutively, by distilled water extract of 800 ppm 71.14%, ethanol extract of 800 ppm 69.39%, and seawater extract of 1,000 ppm 67.81%. The highest RPS against Saprolegnia sp. infection was in ethanol extract of 1,000 ppm (66.12%), followed by ethanol extract of 800 ppm (66.01%), distilled water extract of 1,000 ppm (65,90%), and seawater extract (64.27%) (Table 5).\n\nThe above results indicated that X. granatum extract had the potential to protect shrimp from both V. harveyi and Saprolegnia sp. infection, and, thus, may be applied to improve the survival rate of shrimp in captivity. Vibrio attacks may cause shrimp death from the larva to adult stage if reared in ponds22. The causative vibriosis generally is V. harveyi leading to mortality range of around 40–65%, while the causative agent for the failure of larval development in hatcheries is commonly Saprolegnia sp.\n\nX. granatum has been utilized by the local community for food, animal feed, food preservatives, and traditional medicine. Mangroves are the best choice to isolate natural or bioactive products to challenge against bacteria and fungi23. Substances extracted from mangrove may function as herbal remedies to treat various biological dysfunctions with minimal side effects but with maximum healing potential24. Mangroves provide rich secondary metabolites, such as alkaloids, flavonoids, phenolics, steroids, and terpenoids. Natural phenols, alkaloids, and flavonoids have antioxidant, antibacterial, anti-tumour, and anti-viral properties25–27. Flavonoids are synthesized by plants to respond to microbial infections, and, in vitro, these metabolites are effective antimicrobial substances against microorganisms extensively28. Flavonoid and phenolic compounds from natural sources are known to be associated with a variety of biological activities, such as antioxidant properties, anti-inflammatory actions, and anticancer activities23.\n\n\nConclusions\n\nLeaf extracts of X. granatum had the antimicrobial potential to inhibit the infection of tiger shrimp by V. harveyi and Saprolegnia sp. The ethanol and distilled water leaf extract at concentrations of 800 and 1,000 ppm exhibited higher activity in inhibiting and reducing the infection of V. harveyi and Saprolegnia sp. than antibiotics.\n\n\nData availability\n\nRaw data for tables can be accesed on OSF, DOI: https://doi.org/10.17605/OSF.IO/349SK14.\n\nData are available under the terms of the Creative Commons Zero “No rights reserned” data waiver (CC0 1.0 Public domain dedication).",
"appendix": "Grant information\n\nThis research is part of the Higher Education Research Grant for the 2018 fiscal year, funded by the Four University Development as The Centre of Excellent for Nation Competitiveness, IDB Project for the DIPA 042.05.2.401435/2018. Special gratitude is presented to the Director General of Higher Education and the Rector of Mulawarman University, who made it possible for us to obtain this grant.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nAcknowledgments\n\nThe authors thank to Mr. Basuki, the owner of Windu Permata Hatchery Muara Badak East Kalimantan, Indonesia, and Laboratory of Aquatic Microbiology, Fishery and Marine Science Faculty, Mulawarman University, Indonesia. Our appreciation goes to all of ouer students who helped the authors during the trial in the field.\n\n\nReferences\n\nSaptiani G, Asikin AN, Ardhani F, et al.: Tanaman bakau api-api putih (Avicenia marina) berpotensi menghambat mikrob patogen dan melindungi post larva udang windu (The potential of Avicennia marina to inhibits pathogen microbes and protects the post larva of tiger prawn). J Veteriner. 2018; 19(1): 45–54. Publisher Full Text\n\nSaptiani G, Hardi EH, Pebrianto CA, et al.: Alpinia galanga extracts for improving egg hatchability and larval viability of catfish. In: AIP Conference Proceedings. 2016; 1755: 140002-1–140002-5. Publisher Full Text\n\nRamesh K, Natarajan M, Sridhar H, et al.: Anti-Vibrio activity of mangrove and mangrove associates on shrimp pathogen, Vibrio harveyi VSH5. Global Veterinaria. 2014; 12(2): 270–276. Reference Source\n\nSaptiani G, Prayitno SB, Anggoro S: The effectiveness of Acanthus Ilicifolius in protecting tiger prawn (Penaeus monodon F.) from Vibrio harveyi infection. J Coast Dev. 2012; 15(2): 217–224. Reference Source\n\nSaptiani G, Hardi EH, Pebrianto CA, et al.: Ekstrak Daun Pepaya dan Kangkung untuk meningkatkan daya tetas telur dan kelangsungan hidup larva lele (Extracts of Carica papaya and Ipomoea Aquatica for improving egg hatchability and larval viability of Catfish). J Veteriner. 2016; 17(2): 285–291. Publisher Full Text\n\nSaptiani G, Asikin AN, Ardhani F, et al.: Mangrove plants species from Delta Mahakam, Indonesia with antimicrobial potency. Biodiversitas. 2018; 19(2): 516–521. Publisher Full Text\n\nSidik AS: The changes of mangrove ecosystem in Mahakam Delta, Indonesia: A complex social environmental pattern of linkages in resources utilization. Paper presented at the South China Sea Conference 2008. The South China Sea: Sustaining ocean productivities, Maritime Communities and the Climate. Kuantan, Malaysia 25-29th November. 2008. Reference Source\n\nSahoo G, Mulla NSS, Ansari ZA, et al.: Antibacterial Activity of Mangrove Leaf Extracts against Human Pathogens. Indian J Pharm Sci. 2012; 74(4): 348–351. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSaptiani G, Handayani Hardi E, Agus Pebrianto C, et al.: Antimicrobial potential of Carica papaya, Ipomoea aquatica, Alpinia galanga and Piper betle against the aquatic microbials. Nusantara Bioscience. 2016; 8(2): 252–257. Publisher Full Text\n\nWade NM, Budd A, Irvin S, et al.: The combined effects of diet, environment and genetics on pigmentation in the Giant Tiger Prawn, Penaeus monodon. Aquaculture. 2015; 449(1): 78–86. Publisher Full Text\n\nWade NM, Paulo C, Goodall J, et al.: Quantitative methods to measure pigmentation variation in farmed Giant Tiger Prawns, Penaeus monodon, and the effects of different harvest methods on cooked colour. Aquaculture. 2014; 433: 513–519. Publisher Full Text\n\nBernal Rodríguez CE, Carvajal García A, Ponce-Palafox JT: The color of marine shrimps and its role in the aquaculture. Int J Aquac Fish Sci. 2017; 3: 62–65. Publisher Full Text\n\nSaptiani G, Prayitno SB, Anggoro S, et al.: The influence of Acanthus ilicifolius extracts to histopathological on hepatopancreas of tiger shrimp (Penaeus monodon F.). IJMARCC. 2017; 2(10): 1–6. Reference Source\n\nSaptiani G: Antimicrobial of Nyirih (Xylocarpus Granatum) against Pathogens of Tiger Shrimp Post-Larvae. OSF. 2018. http://www.doi.org/10.17605/OSF.IO/349SK\n\nEl Far SAH, Khalil RH, Saad TT, et al.: Occurrence, characterization and antibiotic resistance patterns of bacterial communities encountered in mass kills of pond cultured Indian prawn (Fenneropenaeus indicus) at Damietta governorate, Egypt. Int J Fish Aquat Stud. 2015; 2(4): 271–276. Reference Source\n\nLomelí-Ortega CO, Martínez-Díaz SF: Phage therapy against Vibrio parahaemolyticus infection in the whiteleg shrimp (Litopenaeus vannamei) larvae. Aquaculture. 2014; 434: 208–211. Publisher Full Text\n\nMirbakhsh M, Akhavan Sepahy A, Afsharnasab M, et al.: Molecular identification of Vibrio harveyi from larval stage of Pacific White Shrimp (Litopenaeus vannamei) Boone (Crustacea:Decapoda)By polymerase chain reaction and 16S rDNA sequencing. Iran J Fish Sci. 2014; 13(2): 384–393. Reference Source\n\nCadiz RE, Traifalgar RFM, Sanares RC, et al.: Comparative efficacies of tilapia green water and biofloc technology (BFT) in suppressing population growth of green vibrios and vibrio parahaemolyticus in the intensive tank culture of Penaeus vannamei. AACL Bioflux. 2016; 9(2): 195–203. Reference Source\n\nKannapiran E, Ravindran J, Chandrasekar R, et al.: Studies on luminous, Vibrio harveyi associated with shrimp culture system rearing Penaeus monodon. J Environ Biol. 2009; 30(5 suppl): 791–795. PubMed Abstract\n\nHeenatigala PPM, Fernando MUL: Occurrence of bacteria species responsible for vibriosis in shrimp pond culture systems in Sri Lanka and assessment of the suitable control measures. Sri Lanka J Aquat Sci. 2016; 21(1): 1–17. Publisher Full Text\n\nAlam MA, Sarder M, Awal MA, et al.: Antibacterial activity of the crude ethanolic extract of Xylocarpus granatum stem barks. Bangl J Vet Med. 2006; 4(1): 69–72. Publisher Full Text\n\nKumaravel K, Ravichandran S, Sritama S: In vitro antimicrobial activity of shrimps haemolymph on clinical pathogens. African J Microbiol Res. 2010; 4(23): 2592–2596. Reference Source\n\nVeni PS, Sunita S, Srinivasulu A: Antibacterial and phytochemical screening of Xylocarpus moluccensis leaf and stem on selected drug resistant and sensitive bacteria. Int J Microbiol Res. 2014; 5(1): 30–34. Reference Source\n\nPrabhu VV, Guruvayoorappan C: Phytochemical screening of methanolic extract of mangrove Avicennia marina (Forssk.) Vierh. Pelagia Res Libr Des Phamacia Sin. 2012; 3(1): 64–70. Reference Source\n\nBaba S, Chan HT, Kainuma M, et al.: Botany, uses, chemistry and bioactivities of mangrove plants III: Xylocarpus granatum. 2016; 14(1): 1–4. Reference Source\n\nSoonthornchareonnon N, Wiwat C, Chuakul W: Biological activities of medicinal plants from mangrove and beach forests. Mahidol Univ J Pharm Sci. 2012; 39(1): 9–18. Reference Source\n\nSharief Md N, Srinivasulu A, Satya Veni P, et al.: Quantification of phytochemicals and antibacterial activity of fruit extract of Avicennia officinalis. Asian J Pharm Clin Res. 2014; 7(2): 127–130. Reference Source\n\nKumar S, Pandey AK: Chemistry and biological activities of flavonoids: an overview. ScientificWorldJournal. 2013; 2013: 162750. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "49679",
"date": "13 Jun 2019",
"name": "Ramalingam Ananda Raja",
"expertise": [
"Reviewer Expertise Vibriosis in shrimp"
],
"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 give the following suggestions on this manuscript:\n1. The Results and Conclusion in the starting of this manuscript itself are confusing and not so clear.\n2. Vibrio spp. and Saprolegnia spp. need NCBI accession number for authentication.\n3. How did the authors determine the initial doses for these bacteria and fungi?\n4. The Introduction is not elaborate and objectives are not specifically mentioned.\n5. The English language needs to be improved, especially tenses.\n6. The RPS formula itself is confusing.\n7. “Mayor” or “Major”; “damaged” or “damage”?\n8. Values should be with S.E or S.D.\n9. Please refer to Ananda Raja et al. (20171).\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": []
},
{
"id": "53934",
"date": "07 Oct 2019",
"name": "Chinmayi Joshi",
"expertise": [
"Reviewer Expertise Antiinfectives from plants"
],
"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 is about the antimicrobial potential of Xylocarpus granatum leaf extracts against tiger shrimp post-larvae pathogens i.e. Vibrio harveyi and Saprolegnia sp.\nComments:\nAuthenticity of test organisms should be confirmed.\n\nDid the authors perform any additional experiments to decide the concentration of bacteria for infection?\n\nIn methods, (‘Treatment’ section) the sentence “The treatments were leaf extract, extracted with ethanol solvent, distilled water or seawater solvent” is confusing.\n\nTables 2-5 contain various values, which should be presented with standard deviation.\n\nThe authors concluded that “the ethanol and distilled water leaf extract at concentrations of 800 and 1,000 ppm exhibited higher activity in inhibiting and reducing the infection of V. harveyi and Saprolegnia sp. than antibiotics.” Effective concentration of antibiotic should also be mentioned here.\n\nIf possible, provide the picture showing the difference between infected and extract-treated shrimp.\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": []
}
] | 1
|
https://f1000research.com/articles/8-63
|
https://f1000research.com/articles/8-62/v1
|
16 Jan 19
|
{
"type": "Case Report",
"title": "Case Report: Basal cell adenoma of esophagus, a rare tumor at an atypical site",
"authors": [
"Sandeep Pandey",
"Yuanjun Gao",
"Dan Li",
"Shengbao Li",
"Guojian Xie",
"Huihui Lü",
"Yan Guo",
"Yu’e Lei",
"Dan Li",
"Shengbao Li",
"Guojian Xie",
"Huihui Lü",
"Yan Guo",
"Yu’e Lei"
],
"abstract": "Basal cell adenomas are rare, salivary gland tumors. These are a class of non-malignant tumors that have glandular epithelial tissue origin. Often described to have a disputed recurrence rate after excision, the prognosis of basal cell adenomas is generally good. However, their rarity makes their continuous monitoring vital. Basal cell adenomas normally cause problems exerting a pressure effect on the surrounding organs or to the originating area. Often due to its location and effect, differentiation between other tumors, e.g. basal cell carcinoma, adenocarcinoma and cystic carcinoma, must be done. We were able to treat a patient with complaints of sensation of regurgitation that was diagnosed as an esophageal mass on esophagogastroduodenoscopy and later as basal cell adenoma on histopathological examination. The patient was treated with endoscopic submucosal dissection successfully and has had no fresh complaints since. Here, we present a case of basal cell adenoma in the esophagus, a rare entity in itself.",
"keywords": [
"basal cell adenoma",
"rare case",
"case report"
],
"content": "Introduction\n\nThe National Cancer Institute, USA, describes adenoma as “a tumor that is not cancer. It starts in gland-like cells of the epithelial tissue (thin layer of tissue that covers organs, glands, and other structures within the body)”1. Basal cell adenomas are a rare type of benign glandular tumors, salivary in origin. Accounting for less than 1-2% of salivary gland tumors2, they are located usually superficial within the glandular body, and often a brownish appearance is observed3. First described by Kleinasser and Klein in 1967, they are termed as rare benign tumors with a high recurrence rate and, in general, good prognosis and recognized as an independent entity in the Second Edition of the Salivary Gland Tumors Classification of the World Health Organization4. Normally indigenous to (parotid) salivary glands, basal cell adenomas have also been reported in the buccal mucosa, palate and nasal septum. They are often seen in patients in their fifth-seventh decade of life, a contrast to other benign tumors5. Basal cell adenoma can be divided into four subtypes: solid, trabecular, tubular and membranous6. Histologically characterized by the presence of uniform and regular basaloid cells having two differenced morphologies, the tumor cells are intermingled. Among the two groups, one has characteristic constitution of small cells with little cytoplasm and intense basaloid nuclei located near the tumoral nests while another is made up of large cells with abundant cytoplasm and pale nuclei in the centre of tumoral nests. The tumoral nests are separated from the surrounding connective tissues by a basal membrane like structure, which surrounds them5. Immunologically speaking, they strongly express PAN CK and CK 5/6, along with Calponin, P-63 and weak staining of Ki-67. Vimentin, actin also stain positive in basal cell adenomas, along with S-100 (alpha subunit) in ductal cells and S-100 (beta subunit) in basaloid cells.\n\n\nCase report\n\nA 63-year-old woman of Chinese (Han) origin was referred to Taihe Hospital, Department of Digestive Medicine after esophagogastroduodenoscopy (EGD) in her county hospital due to a protuberant mass in the esophagus, 35 cm from the incisors. The patient had initially complained of sensation of regurgitation and was a non-smoker and non-drinker. There was no history of heartburn, chest pain, nausea, vomiting or diarrhea. She also had no complaints of hematemesis, hematochezia or melena. The patient was non-diabetic, non-hypertensive, with no history of renal disorders, cardiac diseases or tuberculosis. She had no remarkable family history (medically relevant) and no signs of genetic disorders. The patient had undergone appendectomy 13 years back and had been diagnosed and treated for intestinal obstruction 4 years previously. Laboratory investigation reports were normal. Upper abdominal and chest contrast-enhanced computer tomography (CECT) showed: lower esophagus and cardiac region of stomach wall hypertrophy (highly indicates presence of a tumor); right hepatic lobe diffuse (indicates small cyst present); B/L lung field showed some scattered fibrotic and calcified lesions.\n\nEGD was repeated and showed a protuberant lesion in the esophagus, 3cm in size, 35 cm from the incisors (Figure 1). It also showed another protuberant lesion 0.7 cm in size in the gastric cardia. Multiple yellow spots (gastric xanthoma) were seen in the stomach. Pylorus, pyloric sphincter and duodenum were normal. Endoscopic ultrasound (EUS) of the protuberance in the esophagus showed that the lesion as hypoechoic, an originated from and limited to the submucosal layer with an area of more than 3 cm (Figure 1). No lymphadenopathy was found around the lesion. Radiofrequency ablation was done on the yellow spots.\n\n(EGD by OLYMPUS EVIS LUCERA CV-260SL; EUS by OLYMPUS MAJ-935.)\n\nWith the working diagnosis of esophageal leiomyoma/polyp, the patient was scheduled for surgery. Endoscopic submucosal dissection (ESD) was done under general anesthesia (Figure 2). The surgery went well without complications. The protuberant mass was then sent for histopathological examination and the patient was kept for observation.\n\nRight panel: Closure of the defect by hemoclips. (ESD by single-use Olympus DualknifeTM, tip thickness-0.3mm.)\n\nImmunohistology of the sample showed Calponin (+), CD117 (+), CD 56 partially (+), CKP (+), GATA-3(-), CK5/6 (+), CK7 (+), SOX 10(+), P63 (+), INI-1 (+), and Vimentin (+) (Figure 3). It also showed P63 (+), S-100(+), KI-67 about 2%. Light microscopy revealed the submucosal layer of esophagus showing glandular duct like structure also with double layer epithelium. Around the gland duct there was transparent substance partially solid in nature. The sample also showed two types of epithelium-muscular epithelium and glandular epithelium, which corresponds with the diagnosis of basal cell adenoma, a type of salivary gland tumor. The patient was then discharged seven days after surgery with PPI (Cap. Lansoprazole 30mg BD for 45 days), Itopride (50mg TDS for 45 days) and Hydrotalcite (1gm TDS for 7 days) as discharge medication. Follow up was scheduled two months later.\n\nFollow-up EGD of the patient was performed two months later and showed no recurrence of the tumor. The patient has had no fresh complaints since the ESD procedure (Figure 4).\n\nThe scar from ESD procedure is indicated by the red arrow.\n\n\nDiscussion\n\nBasal cell adenomas are uncommon variants of benign salivary gland tumor with varying recurrence rates. There have been very few reported cases of basal cell adenomas in ectopic sites. Its rarity makes comparison between two similar cases very difficult7. Basal cell adenomas are generally characterized by no regional node involvement, no calcification and no cystic component within the tumor8.\n\nThis case posed a challenge to our team since ectopic occurrences of basal cell adenoma have been reported very rarely in literature and is not a very common occurrence in itself. The prognosis of basal cell adenoma, when found at ectopic sites is not clear and thus poses a challenge to predict the outcome even after intervention.\n\nDifferentiation between benign and malignant basal cell adenoma is vital since they share the same site and characteristics. However, the mechanism and rate of progression is vastly different. Histology is not alone enough to differentiate or predict the progression, hence immunohistology plays a vital role in differentiation. Immunohistological staining with CD5/6, Keratin, Alpha-1 antichymotrypsin, CEA, S-100, Vimentin, CD117, Actin must be done to differentiate it from other benign and malignant tumors.\n\nWhenever encountered and suspected, differentials must be established in favor of basal cell adenoma. Pleomorphic adenoma, adenoid cystic carcinoma, basaloid squamous carcinoma must be excluded first with the help of gross microscopic examination and immunohistological staining.\n\nSurgical excision remains the first and foremost approach of choice for treatment of basal cell adenoma. However, during removal, it is mandatory not to disturb the tumor capsule to prevent recurrence. Moreover, the patient must be kept under regular follow-up to monitor malignant transformation as well as recurrence.\n\n\nConsent\n\nWritten informed consent was obtained from the patient for the publication of this case report and any associated 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\nDefinition of Adenoma- NCI Dictionary of Cancer Terms. Assessed 9 January 2019. Reference Source\n\nNeville BW, Damm DD, Allen CA, et al.: Salivary gland pathology. In: Oral & maxillofacial pathology. W.B. Saunders, Philadelphia, 2009; 484. Reference Source\n\nBernacki EG, Batsakis JG, Johns ME: Basal cell adenoma. Distinctive tumor of salivary glands. Arch Otolaryngol. 1974; 99(2): 84–87. PubMed Abstract | Publisher Full Text\n\nSeifert G, Sobin LH: The World Health Organization's Histological Classification of Salivary Gland Tumors. A commentary on the second edition. Cancer. 1992; 70(2): 379–85. PubMed Abstract | Publisher Full Text\n\nNagao K, Matsuzaki O, Saiga H, et al.: Histopathologic studies of basal cell adenoma of the parotid gland. Cancer. 1982; 50(4): 736–45. PubMed Abstract | Publisher Full Text\n\nGonzález-García R, Nam-Cha SH, Muñoz-Guerra MF, et al.: Basal cell adenoma of the parotid gland. Case report and review of the literature. Med Oral Patol Oral Cir Bucal. 2006; 11(2): E206–9. PubMed Abstract\n\nKatsuno S, Ishii K, Otsuka A, et al.: Bilateral basal-cell adenomas in the parotid glands. J Laryngol Otol. 2000; 114(1): 83–5. PubMed Abstract | Publisher Full Text\n\nTakeshita T, Tanaka H, Harasawa A, et al.: CT and MR findings of basal cell adenoma of the parotid gland. Radiat Med. 2004; 22(4): 260–4. PubMed Abstract"
}
|
[
{
"id": "49887",
"date": "20 Jun 2019",
"name": "Adriana Handra-Luca",
"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 a case of basal cell adenoma of the esophagus.\nThe authors could focus on benign tumors of the esophagus, as well rare epithelial tumors. The authors could add information on the 2 other lesions observed. (Biopsied or not biopsied?) The standard microscopy feautures should be noted before the IHC features.\n\nWhat do the authors mean by “muscular epithelium”? Indicators (arrows) could be added to photos.\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": []
},
{
"id": "59685",
"date": "06 Mar 2020",
"name": "Sadhna Dhingra",
"expertise": [
"Reviewer Expertise Gastrointestinal and Liver Pathology"
],
"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 an interesting and unique case of submucosal basal cell adenoma in the esophagus presenting as a symptomatic esophageal nodule, and diagnosed when resected by endoscopic submucosal dissection.\n\nEndoscopic submucosal dissection is an innovative minimally invasive technique for resection of early cancers of gastrointestinal tract. It became popular in Japan in early 2000 for treatment of superficial gastric cancers and esophageal squamous cell carcinoma. It has gained momentum in other parts of the world for treatment of early superficial neoplastic lesions of the gastrointestinal tract. A byproduct is incidental detection of rare and unique submucosal lesions at unusual sites that present as a nodule and are amenable to endoscopic resection, such as this case.\n\nThe diagnosis rests on the histopathological description and supportive images. The histopathological description and the images provided in the case report to support the diagnosis are not very clear. Good quality microscopic description accompanied with quality images will enhance the quality of the submission. Only one hematoxylin and eosin (HE) stained image is provided. Addition of a couple of more HE stained images that depict clear unequivocal pathology of this lesion as basal cell adenoma will help improve the case report.\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/8-62
|
https://f1000research.com/articles/7-1154/v1
|
30 Jul 18
|
{
"type": "Case Report",
"title": "Case Report: Recurrent hypokalemic periodic paralysis associated with distal renal tubular acidosis (type 1) and hypothyroidism secondary to thyroiditis",
"authors": [
"E. Dante Meregildo-Rodríguez",
"Virgilio E. Failoc-Rojas",
"E. Dante Meregildo-Rodríguez"
],
"abstract": "Background: Hypokalemic periodic paralysis (HypoKPP) is characterized by transient episodes of flaccid muscle weakness. We describe the case of a teenaged boy with HypoKPP and hyperthyroidism due to Hashimoto's thyroiditis with initial manifestation of renal tubular acidosis. This combination is rare and little described previously in men. Case presentation: A 17-year-old boy was admitted after three days of muscular weakness and paresthesia in the lower limbs with an ascending evolution, leading to prostration. Decreased strength was found in the lower limbs without a defined sensory level, reduced patellar and ankle reflexes. Positive antithyroid antibodies were found. He received hydration treatment, IV potassium and levothyroxine, with which there was a clinical improvement. Other examinations led to the diagnosis of type 1 renal tubular acidosis. Conclusion: HypoKPP is a rare disorder characterized by acute episodes of muscle weakness. Type 1 renal tubular acidosis can occur as a consequence of thyroiditis, which is explained by the loss of potassium. This combination is unusually rare, and has not been described before in men. The etiopathogenesis of the disease as well as a dynamic explanation of what happened with the patient are discussed in this report.",
"keywords": [
"hypokalemic periodic paralysis",
"distal renal tubular acidosis (type 1)",
"hypothyroidism",
"Hashimoto Disease"
],
"content": "Abbreviations:\n\nHypoKPP: hypokalemic periodic paralysis; RTA: renal tubular acidosis; DRTA: distal renal tubular acidosis\n\n\nIntroduction\n\nHypokalemic periodic paralysis (HypoKPP) is an uncommon neuromuscular disorder characterized by transient episodes of flaccid muscle weakness, and exceptionally, respiratory failure and death1,2. In addition, HypoKPP is the most common type of periodic paralysis, and may be primary (familial or idiopathic) or secondary (acquired)2–4. Primary HypoKPP occurs when the channelopathies produce potassium intracellular translocation. Secondary HypoKPP is caused by the loss of potassium from kidneys, gastrointestinal tract or skin1,5–9. HypoKPP cases related to thyroid disorders, more frequently thyrotoxicosis, and several autoimmune diseases have been previously reported2,3,5.\n\nRenal tubular acidosis (RTA) is defined as the failure of kidneys to acidify the urine when the glomerular filtration rate is normal or almost normal3,9,10,11.\n\nWe report the case of a teenaged boy with HypoKPP and hyperthyroidism due to Hashimoto's thyroiditis with initial manifestation of RTA. This combination is rare and previously has only been reported in women3–9,12.\n\n\nCase report\n\nA 17-year-old boy, from the region of Cajamarca, high Andean area of Peru, without any relevant medical history, was admitted to the Lambayeque Regional Hospital in April, 2017. For three days the patient had muscle weakness and paresthesia in the lower limbs with an ascending evolution and proximal predominance that made his condition worse, leading to prostration and arrival by emergency route. The patient arrived at the hospital awake, hemodynamically stable, with 24 rpm tachypnea. A neurological physical examination showed weakness in the lower limbs without a defined sensory level, and reduced patellar and ankle reflexes. There was no evidence of bulbar muscle, respiratory and sphincter involvement.\n\nRegarding the ancillary examinations upon admission: hematology tests were within the normal range; normal biochemistry values; elevated thyroid stimulating hormone (TSH) of 5.5 mU/ml (normal values [VN]: 0.27–4.2 mU/ml); low free T4 of 0.78 ng/dl (VN: 0.9–1.7 ng/dl). Regarding serum electrolytes upon admission, they showed hypokalemia (1.44 mmol/L [VN: 3.5–4.5 ng/dl]) without sodium, chloride or calcium alterations.\n\nThe patient was evaluated by the Department of Nephrology, Endocrinology and Neurology and diagnosed with hypothyroidism and hypokalemia. He received replacement treatment with normal saline solution, IV potassium and levothyroxine (T4) 25ug/day. On the fourth day of the treatment, he showed normal potassium values (3.7 mmol/L). After the patient’s clinical condition improved, one week after his admission to the hospital, he was discharged with diagnoses of hypothyroidism (etiology to be determined) and hypokalemia resolved.\n\nAround five weeks after the patient was discharged, he was examined in the endocrinology office and did not show any symptoms. He was indicated to continue with T4 at 25ug/d. Glucose, urea, creatinine, prolactin, morning serum cortisol, testosterone, follicular stimulating hormone, TSH, and free T4 were determined in serum; all of them within normal range. Antithyroid antibodies were positive: anti-thyroglobulin 445.5 UI/ml, anti-TPO 48.20 UI/ml. The following elements were analyzed in 24 hour urine sample: sodium 255.66 mEq., chloride 55.1 mEq., and potassium 89 mEq. Urine test: leukocytes 1–3/field, red blood cells 1–3/field, density 100.5, pH 8, glucose, blood, proteins and leukocyte esterase, all of them negative. Kidney echography: Kidneys maintained their size with multiple images compatible with nephrocalcinosis, bilateral and vesical renal lithiasis. Contrast-enhanced CT of sella turcica was normal. Electrocardiogram: signs compatible with hypokalemia (ST-segment descent, prominent U waves and pseudo-prolongation of the QT interval).\n\nA left-sided X-ray (Figure 1) shows the bone age corresponding to 13 years and 6 months according to the Greulich and Pyle method13. From this, it can be concluded that the diagnosis is secondary hyperthyroidism to thyroiditis and distal RTA.\n\nIt showed a bone at the age of 13 years and 6 months.\n\nThree months after the first episode, over a three week period, the patient stopped the intake of levothyroxine. Around one week later, a new episode of progressive muscle weakness occurred again, which was similar to the previous one. Two days later, the patient was evaluated at a private consultation. The administration of T4 at 50 ug/day was recommend. The symptoms continued progressing, and 5 days later, the patient was admitted again to our hospital by emergency presenting with flaccid quadriparesis with predominance in the lower limbs.\n\nLaboratory tests on second admission: normal hematology values; normal biochemistry values. Urine tests: leukocytes 10–12/field, red blood cells 1–3/field, density 100.5, pH 8, glucose, blood, proteins, and leukocyte esterase, all of them negative. No provocation tests or genetic studies were performed in search of channelopathies. A urine culture and a thorax x-ray were conducted, and both of them showed a normal outcome. Blood gases and serial electrolytes tests shown in Table 1 were also performed. Periodic hypokalemic paralysis was diagnosed in addition to the previous diagnoses.\n\nBGA: Blood Gas Analysis\n\nHe received treatment with IV potassium, IV sodium bicarbonate and T4 25ug/day. The patient improve clinically and was discharged at day five, with T4 25 ug/day. Follow up is every three months with nephrology and internal medicine departments.\n\n\nDiscussion and conclusions\n\nThyrotoxicosis is the most common secondary cause of HypoKPP; while hypothyroidism is an extremely rare cause of HypoKPP6,8.\n\nRTA is a rare cause of HypoKPP6,9,10–12. Table 5 in the paper by Rodríguez Soriano (2002)14 summarizes the different characteristics of types of RTA, leading us to conclude that our patient had distal RTA (DRTA). DRTA is the most common form of RTA. In adults, the most common causes of DRTA are autoimmune disorders such as hyperthyroidism, hypothyroidism, Hashimoto’s thyroiditis, rheumatoid arthritis, Sjögren syndrome, systemic lupus erythematosus, and type 1 diabetes mellitus3–5. However, autoimmune diseases rarely cause DRTA with severe hypokalemia2–4, but this was observed in the case we report here.\n\nThe excretion of H+ is carried out by the intercalated alpha cells (type A) of the distal nephron (distal convoluted tubule and cortical collecting tubule) by means of H+-ATPase luminal pumps, and to a lesser extent, H+-K+-ATPase pumps. The secretion of K+, is carried out fundamentally, by the main cells in the collecting tubules10,11. In hypothyroidism, the number and functioning of these pumps are reduced, which causes a reduction of the excretion of H+, exacerbating the acidosis produced by RTA. The treatment with thyroid hormone increases the activity of these cellular pumps4,12. Since Hashimoto’s thyroiditis is the most common cause of hypothyroidism, DRTA could be an underdiagnosed condition associated12. DRTA associated with non-autoimmune hyperthyroidism has been also described3.\n\nIt is most common that the hereditary forms of DTAR produce bone demineralization, which can cause osteoporosis and osteomalacia in adults, and rickets and growth delay in children3,4,10,12. A bone age retardation of >2 years, in the absence of endocrine deficiency, suggests a constitutional growth delay; while a bone age retardation >3 years is considered pathological11,15. Our patient showed a bone age >3 years.\n\nDTAR is associated with hypercalciuria, hypocitraturia, nephrolithiasis, nephrocalcinosis, chronic interstitial nephritis, and progressive renal failure3–5,10,12. Severe complications of chronic acidosis like myocardial failure, lethargy and coma are rare9. Although HypoKPP associated with P(proximal)TAR have been previously described, the HypoKPP associated with DTAR is more common, more abrupt and more severe7.\n\nIn HypoKPP, whether by potassium movement into cells or by potassium loss, the resulting hypokalemia reduces the resting membrane potential and blocks the action potential1,2. The paralytic attacks occur suddenly with localized or generalized weakness, and can last from one hour to days1,6. Muscle weakness is predominantly proximal and more in the legs than arms, and hyporeflexia is typical. Sensitivity and consciousness are preserved and it is uncommon that extra-ocular, facial, bulbar and sphincter muscles are involved1,2,4.\n\nECGs can show signs of hypokalemia, including ST depression, decrease of the amplitude of the T-wave and an increase of the amplitude of the U-waves; but arrhythmias such as atrial fibrillation, supraventricular paroxysmal tachycardia, or ventricular fibrillation are not common1,6. Our patient showed pseudo-QT prolongation due to the presence of U-wave in the QT segment. However, when QT is measured in aVL, where the U-wave is less prominent, the actual QTc value is obtained and it was normal.\n\nIn the presents case, we didn’t find any other causes of hypokalemia, apart from the DRTA. The possibility of hypokalemic tubulopathies losing salt (Gitelman syndrome and Bartter syndrome) are ruled out since they present with metabolic alkalosis, hypomagnesemia, hyponatremia and elevated potassium levels. Other rare causes of periodic paralysis that occur with hypokalemia, such as thyrotoxic periodic paralysis and Andersen's syndrome, are unlikely given the case presentation. The first of these disorders present with thyrotoxicosis and the second presents with dysmorphic features, ventricular arrhythmias, and a long QT interval with normal, high, or low serum potassium1.\n\nIn our patient, two secondary HypoKPP causes that were extremely rare were found, the secondary hypothyroidism to Hashimoto’s thyroiditis and distal RTA. HypoKPP and DRTA diagnosis was established for that patient with a clinical condition associated with severe hypokalemia. The DTAR diagnosis is fully supported by the presence of hyperchloremic metabolic acidosis, severe hypokalemia, alkaline urine, positive urine anion gap, nephrolithiasis, nephrocalcinosis and polyuria, with normal renal function. In addition, according to the anthropometry of the patient, he failed to thrive and had a bone age retardation of 3.5 years, both conditions typically associated with DRTA. The diagnosis of autoimmune hypothyroidism is evident due to the results of the thyroid profile.\n\nDRTA requires the administration of alkaline salts to correct the acidosis. Sodium bicarbonate should be administered again (1–2 mEq/kg/d) to help to satisfy the alkali requirements and compensate the bicarbonate loss4,5. In addition, oral and IV potassium must be administered again. Our patient responded well to the oral and IV potassium, IV sodium bicarbonate and thyroid hormone replacement treatment. The different differential diagnoses in patients with muscle weakness episodes should be taken into account. Despite not having provocation tests or genetic tests, the clinic and laboratory led to the success of the treatment\n\n\nConsent\n\nWritten informed consent was obtained from the patient for publication of this case report and any 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": "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\nTo Alvaro Taype-Rondan for comments on the article\n\n\nReferences\n\nWilterdink Jl: Hypokalemic periodic paralysis. In: UpToDate, Post TW (Ed), UpToDate, Waltham, MA.\n\nVelarde-Mejía Y, Gamboa-Cárdenas R, Ugarte-Gil M, et al.: Hypokalemic Paralysis: A Hidden Card of Several Autoimmune Diseases. Clin Med Insights Arthritis Musculoskelet Disord. 2017; 10: 1179544117722763. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKoul PA, Wahid A: Distal Renal Tubular Acidosis and Hypokalemic Paralysis in a Patient with Hypothyroidism. Saudi J Kidney Dis Transpl. 2011; 22(5): 1014–1016. PubMed Abstract\n\nBasak RC, Sharkawi KM, Rahman MM, et al.: Distal renal tubular acidosis, hypokalemic paralysis, nephrocalcinosis, primary hypothyroidism, growth retardation, osteomalacia and osteoporosis leading to pathological fracture: a case report. Oman Med J. 2011; 26(4): 271–274. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoyal G, Kant R, Vajpayee A: Hypokalemic respiratory paralysis due to distal renal tubular acidosis as the presenting manifestation of Sjögren's syndrome. Journal of Acute Medicine. 2014; 4(1); 49–52. Publisher Full Text\n\nSinha U, Sengupta N, Sinharay K, et al.: Recurrent hypokalemic paralysis: An atypical presentation of hypothyroidism. Indian J Endocrinol Metab. 2013; 17(1): 174–176. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSingh ThK, Sangme N, Singh LN, et al.: Hypokalaemic Periodic Paralysis Associated with Hypothyroidism. IJPMR. 2012; 23(2): 79–81.\n\nGündüz E, Zengin Y, Dursun R, et al.: Hypokalemic Periodic Paralysis Due To Distal Renal Tubular Acidosis. Eur J Gen Med. 2015; 12(2): 164–166. Publisher Full Text\n\nFinn BC, Young P, Bruetman JE, et al.: [Hypokalemia, distal renal tubular acidosis, and Hashimoto's thyroiditis]. Nefrologia. 2008; 28(5): 569–70. PubMed Abstract\n\nEmmett M, Palmer BF, Sterns RH, et al.: Etiology and diagnosis of distal (type 1) and proximal (type 2) renal tubular acidosis. In: UpToDate, Post TW (Ed), UpToDate, Waltham, MA. Reference Source\n\nEscobar L, Mejía N, Gil H, et al.: Distal renal tubular acidosis: a hereditary disease with an inadequate urinary H⁺ excretion. Nefrologia. 2013; 33(3): 289–96. PubMed Abstract | Publisher Full Text\n\nSuzanne M, Martin D, Eddine TA, et al.: Hypokalemia paralyzing revealing a rare association of autoimmune diseases: type 1 diabetes, thyroiditis and tubulopathy about a case. IOSR Journal Of Pharmacy. 2015; 5(8): 5–7. Reference Source\n\nGreulich WW, Pyle SI: Radiographic atlas of skeletal development of the hand and wrist. 2nd edn. Stanford University Press, Stanford. 1959. Reference Source\n\nRodríguez Soriano J: Renal tubular acidosis: the clinical entity. J Am Soc Nephrol. 2002; 13(8): 2160–70. PubMed Abstract | Publisher Full Text\n\nDahlberg PS, Mosdøl A, Ding KY, et al.: Agreement Between Chronological Age and Bone Age Based on the Greulich and Pyle Atlas for Age Estimation: A Systematic Review [Internet]. Knowledge Centre for the Health Services at The Norwegian Institute of Public Health (NIPH): NIPH Systematic Reviews: Executive Summaries. 2017. PubMed Abstract"
}
|
[
{
"id": "37480",
"date": "31 Oct 2018",
"name": "Cristian Diaz Veliz",
"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\n1. The case report is well written, light and fast reading. 2. The values of potassium are displayed in moles and the normal value in another units, use the same units 3. Specify the importance of bone age for the Case Report 4. Try to explain the presence mostly in women 5. Develop a diagram \"decision making\" with the presumptions diagnostic\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": "4307",
"date": "03 Jan 2019",
"name": "Virgilio E Failoc-Rojas",
"role": "Author Response",
"response": "Dear Dr Cristian. Answering your questions.1) For a new version, the reference values will be added and expressed in mmol/ L2) The delays in bone maturation, evaluable by bone age and the failure of bone growth are typical alterations of distal renal tubular acidosis. This is explained in the penultimate paragraph.3) Most autoimmune disorders, including are more common in women."
}
]
},
{
"id": "41009",
"date": "26 Nov 2018",
"name": "Stalin Viswanathan",
"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\nIn the introduction, 3rd paragraph: “… HypoKPP and hyperthyroidism due to Hashimoto…\" is written instead of “hypothyroidism”.\nIn the same paragraph: “This combination is rare and previously has only been reported in women”. Such a combination has been reported at least once previously in 20171. Since I am one coauthor of this paper, I am not sure whether I can insist that this should be quoted.\nIn the Case report section 2nd paragraph:\nIt would have been preferable to mention the hypokalemia first and then mention secondary investigations such as thyroid profile and magnesium/calcium. The TSH of 5.5 is normal for many labs. If it was slightly higher than your lab limits, it would be preferable to label it subclinical hypothyroidism, since TSH is <10. Family history of thyroid disease or other autoimmune diseases can be mentioned. Also, did the patient have a goitre? Did the ECG also show low voltage complexes or bradycardia suggestive of hypothyroidism? It is not mentioned whether he was treated with bicarbonate supplements on an outpatient basis when a diagnosis of RTA was made. Was TTKG considered if the facilities for urine osmolality were available?\n\nIn Table 1:\nMagnesium appears low in the table, but in page4/7 last paragraph, it is mentioned that the patient did not have hypomagnesemia and hence Gitelman was not considered. It is also mentioned that Gitelman will have elevated K levels which is untrue. Since the patient had high urine K loss and hypomagnesemia, Gitleman was considered except that ABG showed metabolic acidosis.\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": [
{
"c_id": "4308",
"date": "03 Jan 2019",
"name": "Virgilio E Failoc-Rojas",
"role": "Author Response",
"response": "Dear Dr Stalin.The editing corrections are corrected in the new version that will be sent.In addition, the article that suggests will be cited.There was no family or personal background. This additional information is added to the Case Report.The patient did not have a goiter. This additional information is added to the Case Report.The EKC showed bradycardia, but not low-voltage complexes. This additional information is added in the Case Report.During follow-up after discharge, the patient maintained treatment with T4 and it was not necessary to administer oral bicarbonate.It was considered, but it was not possible to analyze TTKG (transtubular potassium gradient) since this test is not available in our hospital. This additional information is added to the Case Report.the result of the serum magnesium was 0.54 and 0.51 mmol / L (our laboratory considers a reference range of 0.65-1.05 mmol / L). Therefore, the patient did present hypomagnesemia.The reviewer is right in stating that Gitelman syndrome cannot be ruled out, although it is true that the patient did not present with metabolic alkalosis, hyponatremia, and hyperkalemia if he presented hypomagnesemia and it is possible that due to the high losses of potassium in urine The patient will not have hyperkalemia. Therefore, the statement is corrected and reformulated: In the presents case, we didn’t find any other obvious cause of hypokalemia, apart from the DRTA. The possibility of hypokalemic tubulopathies losing salt (Gitelman syndrome and Bartter syndrome) are less probable since they typically present with metabolic alkalosis, hyponatremia, and elevated potassium levels; and also because there was no known familial history of any kidney disorder”."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1154
|
https://f1000research.com/articles/7-565/v1
|
10 May 18
|
{
"type": "Research Article",
"title": "Cigarette smoking and metabolic syndrome components: a cross-sectional study from Maracaibo City, Venezuela",
"authors": [
"Valmore Bermudez",
"Luis Carlos Olivar",
"Wheeler Torres",
"Carla Navarro",
"Robys Gonzalez",
"Cristobal Espinoza",
"Alicia Morocho",
"Andres Mindiola",
"Maricarmen Chacin",
"Victor Arias",
"Roberto Añez",
"Juan Salazar",
"Manuel Riaño-Garzon",
"Edgar Diaz-Camargo",
"Maria Judith Bautista",
"Joselyn Rojas",
"Luis Carlos Olivar",
"Wheeler Torres",
"Carla Navarro",
"Robys Gonzalez",
"Cristobal Espinoza",
"Alicia Morocho",
"Andres Mindiola",
"Maricarmen Chacin",
"Victor Arias",
"Roberto Añez",
"Manuel Riaño-Garzon",
"Edgar Diaz-Camargo",
"Maria Judith Bautista",
"Joselyn Rojas"
],
"abstract": "Background: A growing body of evidence suggests that cigarette smoking can cause the onset of metabolic syndrome prior to cardiovascular diseases. Therefore, the objective of this study was to evaluate the relationship between smoking habit and metabolic syndrome components in an adult population from Maracaibo city, Venezuela. Methods: The Maracaibo City Metabolic Syndrome Prevalence Study is a descriptive, cross-sectional study with random and multi-stage sampling. In this sub-study, 2212 adults from both genders were selected. On the basis of their medical background, they were classified as smokers, non-smokers and former smokers. Metabolic syndrome was defined according to Harmonizing 2009 criteria, using population-specific abdominal circumference cut-off points. The association between risk factors was evaluated using a logistic regression model. Results: In the studied population, 14.8% were smokers, 15.4% were former smokers. In the multivariate analysis, the presence of metabolic syndrome (smokers: OR, 1.54; 95% CI, 1.11–2.14; p=0.010) and its components were related to cigarette smoking, with the exception of hyperglycemia. High blood pressure was inversely associated with current smoking status (smokers: OR, 0.70 (0.51–0.95); p=0.025). Conclusion: Cigarette smoking represents an independent risk factor for the development of metabolic syndrome, being associated with low high-density lipoprotein-cholesterol, increased abdominal circumference and elevated triacylglyceride levels. Former smokers did not present a greater risk for developing this metabolic disease when compared to non-smokers. The effect of avoiding this habit should be evaluated in future studies in our population.",
"keywords": [
"smoking habit",
"metabolic syndrome",
"smokers",
"hypertension",
"cardiovascular risk."
],
"content": "Introduction\n\nSmoking is one the main causes of morbidity and mortality in the working-age population; it is responsible for approximately 7.2 million deaths per year1. This constitutes a major public health issue. Almost one-third of the world population older than 15 years of age smokes2, with a global prevalence of 21.2% in developing countries3. In the Americas, the prevalence in the general adult population is 17.1%4; however, this varies among different countries, with Chile having the highest (38.9%) and Panama the lowest (7.4%) rates4. Venezuela is a country with one of the highest prevalence (33.9%)5, with a frequency of 14.8% in Maracaibo City in recent studies6.\n\nThese percentages are important because smoking habit is a major modifiable risk factor for developing non-communicable diseases2, including cardiovascular disease (CVD) and type 2 diabetes mellitus (DM2)7. A growing body of evidence suggests that before the onset of these two diseases, cigarette smoking favors the appearance of metabolic syndrome (MS)8,9 high blood pressure, dyslipidemia, obesity and high blood glucose10–14. Main contributors for this association include the presence of dyslipidemia and central obesity15.\n\nThe physiopathology of the relationship between cigarette smoking and MS comes from a decrease in peripheral insulin sensitivity, lipoproteins metabolism alterations and endothelial dysfunction, all present in smoking individuals16. Until now, epidemiological published results are not definitive in showing the association between cigarette smoking and MS. On the other hand, it is not certain whether this association is caused by other behavioral patterns and unhealthy habits of patients with cardiometabolic diseases17. Thus, the aim of this study was to evaluate the relationship between smoking habit and MS components in the adult population from Maracaibo City, Venezuela.\n\n\nMethods\n\nThe Maracaibo city MS prevalence study (MMSPS) was a cross-sectional, descriptive study performed in Maracaibo, Venezuela. It was designed to provide estimations about the presence of MS and associated cardiovascular risk factors in the adult population during the period between May 2007 and December 2009. The study method was reported previously18. The most important aspects of the protocol are presented here. Maracaibo city was divided into parishes, which were sampled proportionally through a multistage random sampling, defining conglomerates in two phases: In the first phase, the conglomerates represented the sectors of the 18 parishes, selecting 4 areas per parish by means of simple random sampling; in the second phase, the conglomerates were represented by the neighborhood of each chosen area, to which a random number was assigned. To evaluate smoking habit in this sub-study, 18 subjects were excluded: 9 because of inconsistencies about when they started smoking; the other 9 subjects because they smoked cigar types that were different to cigarettes. Finally, a total of 2212 subjects were evaluated. The study was approved by the Bioethics Committee of the Endocrine and Metabolic Research Center – University of Zulia (approval number: BEC-006-0305). This ethical approval included all future studies that used the data from the MMSPS. All participants signed an informed consent form before being questioned and physically examined by a trained team.\n\nEvery subject in the study underwent a medical examination performed by trained personnel to obtain a full medical history. During the anamnesis, past medical and family history of endocrine and metabolic disorders was collected; including age, race, marriage status, education and socioeconomic status. The latter was measured using the Graffar scale modified by Mendez-Castellano and De Mendez19.\n\nThe auscultatory method was performed to measure arterial pressure, using an adequate calibrated and validated sphygmomanometer. Korotkoff phases I and V were used to measure systolic and diastolic pressures, respectively. Subjects remained sitting still for 15 minutes before assessment, with both feet on the ground. A total of 3 measurements per day were taken in 15 minute intervals, for 2 days consecutively. The employed criteria to classify the population came from the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC7)20.\n\nAnthropometric measures were taken using a height rod that had been previously calibrated and placed on a flat surface. Weight was measured using a digital weighing scale (Tanita, TBF-310 GS Body Composition Analyzer; Tokyo, Japan), with the patient wearing light clothes and no shoes. The body mass index (BMI) was calculated applying the Quetelec formula (weight/height2), and classified according to the WHO classification21, as follows: normal weight (<25 Kg/m2), overweight (25.0–29.9 Kg/m2), obese (≥30.0 Kg/m2). Abdominal circumference was measured using a plastic measuring tape, graded in centimeters and millimeters, in a spot equidistant to the lower ribcage and the anterior-superior iliac spine, according to the United States National Institute of Health protocol22.\n\nSubjects were asked about smoking habit presence and duration, being categorized as: a) current smoker, any subject who had smoked more than 100 cigarettes in his/her lifetime, is currently smoking, or less than 1 year had passed after he/she quit smoking; b) Former smoker: any subject who has quit smoking for more than 1 year; c) non-smoker, any subject who has never smoked or had smoked less than 100 cigarettes in his/her lifetime6. Smoking intensity was assessed posteriorly, according to number of cigarettes per day. It was divided in the following tertiles: T1 <3 cigarettes/day; T2 = 3–9 cigarettes/day; and T3 ≥10 cigarettes/day.\n\nPhysical activity was evaluated using the International Physical Activity Questionnaire23. It takes into account four elements of evaluation: physical activity in transport, work, domestic and gardening, and leisure time. To quantify time investment on each element, subjects were classified in quintiles. The final scoring was reported using metabolic equivalents (METs)-min/week on each item; any subject with 0 METs was considered as physically inactive.\n\nSubjects with ≥1 MET were classified in quintiles according to gender, resulting in six categories for physical activity: physical inactivity (MET = 0), very low (Q1), low (Q2), moderate (Q3), high (Q4), and very high (Q5) physical activity. Leisure time was classified as: a) Q1 or very low physical activity, <296.999 METs for men and <230.999 METs for women; b) Q2 or low physical activity, 297.000–791.999 METs for men and 231.000-445.499 METs for women; Q3 or moderate physical activity, 792.000–1532.399 METs for men and 445.500-742.499 METs for women; Q4 or high physical activity, 1532.400–2879.999 MET for men and 742,500–1798.499 METs for women; and e) Q5 or very high physical activity, ≥2879.000 METs for men and ≥1798.500 METs for women. For alcohol consumption, any subject that drinks ≥1 gram daily was considered as a “drinker”24.\n\nAfter 8 hours of fasting, a blood sample was taken from the cubital vein, and was centrifuged to obtain the serum. Serum levels of glucose (catalog number REF-10123), total cholesterol (catalog number REF-10015) and triacylglycerides (TAG) (catalog number REF-10163) were determined using enzymatic–colorimetric kits (Human Gesellschaft für Biochemica und Diagnostica mbH) and a specialized computer system. Glycemic status was classified according to ADA 2017 criteria in normal glucose (basal glucose, <100 mg/dl), impaired fasting glucose (basal glucose, 100–125 mg/dl) and DM2 (≥126 mg/dl)25. Serum hs-C reactive protein (hs-CRP) levels were quantified using immunoturbodimetric assays (Human Gesellschaft für Biochemica und Diagnostica mbH. (catalog number REF-11544), setting the cutoff point at ≥0,765 mg/l26.\n\nBasal insulin serum levels were determined using a commercial kit (catalog number EIA-2935) based on the ELISA method (DRG International, Inc.), with a detection limit of <1 mU/l. Insulin resistance (IR) was calculated using software (HOMA-Calculator v2.2.2) supplied by the Oxford Centre for Diabetes, Endocrinology and Metabolism; the cutoff-point for HOMA2-IR was 2.0027.\n\nMS diagnosis was made using the proposed criteria from the IDF and AHA/NHLBI in 200928. It requires three or more of the following components to achieve a diagnosis: 1) TAG ≥150 mg/dl; 2) high-density lipoprotein–cholesterol (HDL-C) <40 mg/dl for men or <50 mg/dl for women; 3) basal glucose levels ≥100 mg/dl, or a previous diagnosis of DM2 or use of an antidiabetic drug; 4) arterial pressure ≥130/85 mmHg, or a previous diagnosis of hypertension or use of an antihypertensive drug; 5) abdominal circumference with cutoff points adapted for our population, which are ≥91 for women and ≥98 cm for men29.\n\nQualitative variables were expressed in absolute and relative frequencies. The relationship between these was examined with a χ2 test and the difference in proportions using a Z-test. Quantitative variables were expressed in arithmetic means ± standard deviations, with prior analysis using Geary’s test. Variables without a normal distribution were submitted to logarithmic transformation with posterior normality test. Multiple logistic regression models were made to estimate odds ratios (OR) and 95% confidence intervals (95% CI); they were used for the presence of MS and each of its components, adjusted for gender, age, ethnic group, marital status, education level, socioeconomic status, working status, alcohol consumption, BMI categories, insulin resistance and smoking habit. On another model, smoking intensity was assessed dividing consumption in tertiles (T1 <3 cigarettes/day; T2 = 3–9 cigarettes/day; and T3 ≥10 cigarettes/day). Data were analyzed by using SPSS v.21 for Windows (IBM SPSS), and considering statistically significant results when p<0,05.\n\n\nResults\n\nA total of 2212 individuals were studied, of whom 52.7% (n=1166) were women. The mean age ± SD was 39.27±15.38 years, and the most frequently occurring age group was 30–49 years (38.5%; n=851). For smoking habit, 14.8% were smokers (n=328), 15.4% former smokers (n=340) and 69.8% were non-smokers (n=1544). The prevalence of MS was 35.7% (n=935) in the sample. The most frequent MS components were low HDL-C (57.6%; n=1275) and abdominal obesity (48.5%; n=1072). Other general characteristics can be found in Table 1.\n\nWC, waist circumference; HDL, high-density lipoprotein; TAG, triacylglycerides. †According to IDF and AHA/NHLBI-ISO-2009 Consensus28. ‡ According to The Maracaibo City Metabolic Syndrome Study29: WC ≥98cm for men; ≥91cm for women.\n\nSmoking habit in accordance to MS components could be seen in Table 2. It shows a statistically significant association between cigarette smoking and having MS (χ2=39.285; p<0.001) with a greater percentage of individuals with MS in former smokers (47.9%) and current smokers (42.1%) than in non-smokers (31.6%), p<0.05.\n\nMaracaibo, Venezuela.\n\nMS, Metabolic Syndrome; TAG, triacylglycerides; HDL-C, high-density lipoprotein-cholesterol; BP, blood pressure.\n\nMS: χ2= 39.285 (p<0.001); High TAG: χ2= 41.886 (p<0.001); Abdominal Obesity: χ2= 40.039 (p<0.001); Hyperglycemia: χ2= 10.759 (p=0.005); Low HDL-C levels: χ2= 2.160 (p=0.340); High Blood Pressure: χ2= 16.883 (p<0.001). †According to IDF and AHA-NHLBI-ISO 200928. ‡ According to The Maracaibo City Metabolic Syndrome Study29: WC ≥98cm for men; ≥91cm for women.\n\nEach component of the MS was analyzed in relation to smoking. A higher percentage of individuals with high TAG were former (37.1%) and current (36.9%) smokers, compared with non-smokers (23.6%) (χ2=41.886; p<0.001). The same happened for abdominal obesity in former smokers (62.9%) and current smokers (52.1%) (χ2=40.039, p<0.001). A high percentage of former smokers presented hyperglycemia (33.5% vs 25.9%; χ2=10.759; p<0.005) and high blood pressure (48.5 vs 36.9%; χ2=16.88; p<0.001) in comparison to nonsmokers. No statistical association was found between low HDL-C and smoking status.\n\nComparing smoking habit with the number of MS criteria (Figure 1), a statistically significant association was observed (χ2=49.249, p<0.001). The highest percentages were with nonsmokers who met 0 criteria (76.31%) and 1 criterion (76.77%). However, the greatest prevalence of smokers was observed in subjects who met 4 (16.96%) and 5 (20%) criteria.\n\nIn Table 3, models of multivariate logistic regression are shown for the diagnosis of MS and its components. An association between current smoking and increased risk of presenting with MS could be observed (OR, 1.54; 95% CI, 1.11–2.14; p=0.010); the same was true of high TAG serum levels (OR, 1.66; 95% CI, 1.23–2.23; p<0.001); abdominal obesity (OR, 1.54; 95% CI, 1.05–2.28; p=0.027) and low HDL-C levels (OR, 1.32; 95% CI, 1.01–1.74; p=0.046).\n\nTAG, triacylglycerides; HDL-C, high-density lipoprotein-cholesterol; BP, blood pressure; OR, odds ratio; CI, confidence interval. †According to IDF and AHA/NHLBI-ISO 200928. ‡ According to The Maracaibo City Metabolic Syndrome Study29: WC ≥98cm for men; ≥91cm for women.\n\nOn the other hand, by assessing smoking intensity according to number of cigarettes per day (Table 4), an association between the consumption tertile and high serum TAG levels was observed (T3: OR, 1.51; 95% CI, 1.03–2.22; p=0.036). Also, an association was observed between smoking intensity and abdominal obesity (T3: OR, 2.05; 95% CI, 1.15–3.64; p=0.015). By contrast, an inverse relationship was observed with high blood pressure (T3: OR, 0.66; 95% CI, 0.44–0.99; p=0.045).\n\nMS, metabolic syndrome; TAG, triacylglycerides; HDL-C, high-density lipoprotein-cholesterol; BP, blood pressure. †According to IDF and AHA/NHLBI-ISO 200928. ‡According to The Maracaibo City Metabolic Syndrome Study29: WC ≥98cm for men; ≥91cm for women.\n\n\nDiscussion\n\nCigarettes are composed of more than 1000 toxic and carcinogenic elements. Nicotine is the main alkaloid in tobacco; it constitutes 1.5% of the commercial cigarette weight and 95% of the total alkaloids present31. Despite its effects, cigarette smoking has spread all over the globe, becoming a leading cause of chronic and degenerative pathologies. In Latin America, its use has markedly increased since 1950, and is now considered the second most common cardiovascular risk factor, following high blood pressure32. This has led to an increase in cancer deaths and a drop in life expectation of 2–6 years33. In Maracaibo, high prevalence of cigarette smoking and MS has been observed, which may suggest an existing relationship between these variables6,34.\n\nIn this study, MS prevalence in current smokers was 42.1% in both genders and a greater probability of having MS than in nonsmokers was observed. Kang and Song in the Korea National Health and Nutrition Examination Survey (KHANES) reported similar results with a cross-sectional study with 11559 subjects. They evaluated smoking habit by looking for nicotine in urine samples; a greater risk for developing MS was observed in those subjects35. Likewise, Slagter et al.11 conducted a study in the Netherlands which included 59,467 subjects from both sexes. In that study, a higher prevalence of MS was observed in smokers (a dose-dependent relationship), and increased the risk of MS depended neither on BMI nor gender.\n\nOn the other hand, Sun et al.8 conducted a meta-analysis from multiple cohort studies and included 13 articles. In total, 56,691 subjects and 8688 cases from Asia, Europe and North America were included. They found that cigarette smoking actively increases the risk of having MS. The effects of smoking on the cardiovascular system could be caused by increased action of nicotinic receptors. Activation of nicotinic receptor could promote the release of neutransmitters and hormones such as vasopressin, CRH, ACTH, growth hormone, dopamine, serotonin, glutamate and GABA in the central nervous system, acetylcholine in the peripheral nervous system, and catecholamine and cortisol from the adrenal glands. All of these molecules affect metabolism and appetite regulation36.\n\nThe CKB cohort study37 included 487,527 adult subjects and reported that regular cigarette smoking was associated with a decrease in BMI and an increase in abdominal circumference in both men and women (they used an adjusted model for BMI). Similar results were reported from the FINRISK study38, which included 5817 Finnish adults; greater abdominal circumference was observed in overweight and obese women who smoke. Clair et al.39, in a cross-sectional study that included 6123 adult Caucasians from Switzerland, reported that both sexes had an increased risk of obesity according to the number of cigarettes they smoke per day. These results resemble those from the present study, where cigarette smoking was associated with increased abdominal obesity. This epidemiological behavior could be explained by the recent hypothesis of the association between cigarette smoking and a decrease in body weight, using the CHRNA3 genetic variant (rs1051730); establishing that smoking does not affect body fat distribution and the increase in localized visceral fat and in abdominal obesity are due to high cortisol plasma levels and insulin resistance, respectively40–42.\n\nAccording to the results of our study, cigarette smoking represents a risk factor for having high TAG levels. Similar behavior was seen in the ICMR-INDIAB cross-sectional study43 of 16,607 adult individuals, which showed a positive correlation between high TAG levels and smoking. Ueyama et al.44 reported there was a positive association between smoking and high TAG levels in a study of 5959 Japanese individuals. This phenomenon could be explained by the fact that stimulation of the sympathetic nervous system produces the release of insulin antagonists. These antagonists, such as cortisol and growth hormone, increase lipolysis, leading to an elevation of free fatty acids in the blood8,42.\n\nLow HDL-C levels were observed more frequent in smokers than nonsmokers in our study. Sun et al.45, in a cross-sectional study of 11,956 Chinese individuals, reported similar results by showing that current smokers had an increased risk of having low HDL-C levels. Takata et al.46, showed that in 32 individuals who were participating in an anti-smoking program using varenicline or transdermal nicotine patches, HDL-C levels, apolipoprotein AI and HDL subfractions did not change significantly according to therapeutic strategy used. In the same study, cholesterol efflux capacity and HDL inflammatory index improved significantly with the anti-smoking program (baseline cholesterol efflux capacity: 14.15±2.46% vs after smoking cessation cholesterol efflux capacity: 14.83±2.35%; p=0.01; baseline HDL inflammatory index: 1.13±0.31 vs after smoking cessation HDL inflammatory index: 0.98±0.18 %; p=0.01).\n\nThe inverse relationship between current cigarette smoking and high blood pressure observed in the present study is noteworthy. However, three decades ago cigarette smoking was globally reported as acutely increasing blood pressure, heart frequency and myocardial contractility47. This was thought to be caused by increased nicotinic activity on the sympathetic nervous system. Despite this, epidemiological evidence could not confirm the role of cigarette smoking in the development of elevated blood pressure48. On the other hand, diverse evidence suggests an inverse relationship between these factors. Kaneko et al.49, in a recent study of 1297 Japanese individuals without any history of high blood pressure, showed that cigarette smoking appeared to be a protective factor against blood pressure elevation. Onat et al.50 observed a similar pattern in a Turkish population. The inverse association between blood pressure and smoking habit could be related to the cigarette effect on weight loss, since obesity is associated with a high incidence of high blood pressure; explaining the rebound effect on blood pressure in obese subjects who stop smoking50; however, this study evidenced that smokers presented with more abdominal obesity than nonsmokers. Another theory to explain this behavior suggests that smokers show less response to psychological stress: many of them report a decrease in anxiety and stress when smoking a cigarette51. This may come from modifications to adrenal and cardiovascular responses to external stimuli caused by cigarette smoking; thus, stopping smoking would increase blood pressure52.\n\nLeone53 reports a two-phase effect of cigarette smoking on arterial pressure: the first phase, without a determined duration, when there is a decrease in blood pressure; and the second phase, when the smoker develops elevated blood pressure from the toxic effects of carbon monoxide53. This finding shows the importance in chronologically assessing smoking habit duration. Despite this, smoking does not benefit to cardiovascular health, but increases the risk of cardiovascular disease, especially in men50.\n\nSimilarly, with the assessment of smoking intensity according to number of cigarettes per day, a direct relationship was found between the number of cigarettes smoked and an increased risk of high serum TAG levels and abdominal obesity, and an inverse relationship with hypertension; this was seen especially in heavy smokers (≥10 cigarette daily). In this sense, in a study performed by Chen et al.10, 1146 individuals showed a significant dose-response relationship between the number of cigarettes per day and high TAG levels. Data analysis from the KHANES study revealed an increased risk of obesity and central obesity with an increase in smoking habit intensity54. This relationship could be caused by the dose-dependent effect of nicotine on fatty acid metabolism and catecholamine release; also inducing increase in lipolysis, free fatty acids, VLDL, LDL levels, and visceral adipose tissue independent of weight gain or loss55.\n\nIn the present study, former smokers did not exhibit an increased risk of developing MS or its components when compared with non-smokers. Similar results were reported in Korea by Oh et al.56. The benefits to cardiovascular health from stopping cigarette smoking seem to depend on the following variables: first, the time since the subject stopped; and second, the length of time for which he/she was smoking and the quantity of cigarettes. A previous study showed that smoking 20 cigarettes daily increased the risk of developing MS for the next 10 years, whereas smoking 40 cigarettes daily increased the risk for the next 20 years57. This is why in the Maracaibo population it is necessary to conduct a cohort study on subjects who stopped smoking to evaluate the long term effects on cardiometabolic health.\n\nRegarding the limitations of this study, its cross-sectional design makes it incapable of determining causality; it is also influenced by the subjectivity of its participants regarding the intensity and duration of their smoking habit. All of this should be considered in future studies.\n\nIn conclusion, the present study showed that smoking in our population represent an independent risk factor to develop MS, and is individually associated with low HDL-C levels, increased abdominal circumference and high TAG levels. Former smokers did not show any increase in risk of developing MS relative to non-smokers; despite this, future research studies should be conducted to evaluate how stopping cigarette smoking decreases cardiometabolic risk. Prevention measures focused on patients who smoke, especially anti-smoking counseling from medical personnel, could help to decrease any cigarette cardiometabolic consequences in the Maracaibo City population.\n\n\nData availability\n\nDataset 1. Cigarette smoking and MS MMSPS dataset. DOI: 10.5256/f1000research.14571.d20185130",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nThis study was funded by the Council of Scientific, Humanistic and Technological Development under No. CC-0437-10-21-09-10 and Fundacite-Zulia under No. FZ-0058-2007.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nGBD 2015 Risk Factors Collaborators: Global, regional, and national comparative risk assessment of 79 behavioural, environmental and occupational, and metabolic risks or clusters of risks, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016; 388(10053): 1659–1724. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorld Health Organization: WHO global report on trends in prevalence of tobacco smoking 2015. Reference Source\n\nOrganización Mundial de la Salud: Informe OMS sobre la epidemia mundial de tabaquismo: aumentar los impuestos al tabaco. Ginebra: OMS; 2015. Reference Source\n\nCayon A: Regional Report on Tobacco Control, 2016. Pan American Health Organization / World Health Organization. 2017 [cited 25 August 2017]. Reference Source\n\nBlanco A, Caixeta R, Sandoval R, et al.: Informe sobre Control del Tabaco en Venezuela. Organización Panamericana de Salud. 2011. Reference Source\n\nBermúdez V, Miquilena E, Salazar J, et al.: Smoking Habit in Adult Population from Maracaibo City, Venezuela. Int J Respir Pulm Med. 2016; 3: 061. Publisher Full Text\n\nWorld Health Organization: WHO Global Status Report on Noncommunicable Diseases. Attaining the nine global noncommunicable diseases targets: a shared responsibility. Geneva: WHO. 2014. Reference Source\n\nSun K, Liu J, Ning G: Active smoking and risk of metabolic syndrome: a meta-analysis of prospective studies. PLoS One. 2012; 7(10): e47791. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAl-khalifa I, Mohammed S, Ali Z: Cigarette Smoking as a Relative Risk Factor for Metabolic Syndrome. J Endocrinol Metab. 2016; 6(6): 178–182. Publisher Full Text\n\nChen CC, Li TC, Chang PC, et al.: Association among cigarette smoking, metabolic syndrome, and its individual components: the metabolic syndrome study in Taiwan. Metabolism. 2008; 57(4): 544–548. PubMed Abstract | Publisher Full Text\n\nSlagter SN, van Vliet-Ostaptchouk JV, Vonk JM, et al.: Associations between smoking, components of metabolic syndrome and lipoprotein particle size. BMC Med. 2013; 11(1): 195. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDedinska I, Laca Ľ, Miklušica J, et al.: Czypalenietytoniu jest czynnikiemryzykazespołumetabolicznego? Clinical Diabetology. 2014; 3(4): 136–143.\n\nTsai JS, Chen SC, Huang KC, et al.: Plasma zinc α2-glycoprotein levels are elevated in smokers and correlated with metabolic syndrome. Eur J Clin Invest. 2015; 45(5): 452–459. PubMed Abstract | Publisher Full Text\n\nYankey BN, Strasser S, Okosun IS: A cross-sectional analysis of the association between marijuana and cigarette smoking with metabolic syndrome among adults in the United States. Diabetes Metab Syndr. 2016; 10(2 Suppl 1): S89–S95. PubMed Abstract | Publisher Full Text\n\nCena H, Fonte ML, Turconi G: Relationship between smoking and metabolic syndrome. Nutr Rev. 2011; 69(12): 745–753. PubMed Abstract | Publisher Full Text\n\nJia WP: The impact of cigarette smoking on metabolic syndrome. Biomed Environ Sci. 2013; 26(12): 947–952. PubMed Abstract | Publisher Full Text\n\nRabaeus M, Salen P, de Lorgeril M: Is it smoking or related lifestyle variables that increase metabolic syndrome risk? BMC Med. 2013; 11(1): 196. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBermúdez V, Marcano RP, Cano C, et al.: The Maracaibo city metabolic syndrome prevalence study: design and scope. Am J Ther. 2010; 17(3): 288–294. PubMed Abstract | Publisher Full Text\n\nMéndez-Castellano H, De Méndez M: Estratificación social y biología humana: método de Graffar modificado. Arch Ven Pueric Pediatr. 1986; 49: 93–104.\n\nChobanian AV, Bakris GL, Black HR, et al.: The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA. 2003; 289(19): 2560–72. PubMed Abstract | Publisher Full Text\n\nWorld Health Organization: The World Health Report 2003. Reference Source\n\nHealth Statistics: NHANES III reference manuals and reports (CDROM). Hyattsville, MD:Centers for Disease Control and Prevention, 1996. Reference Source\n\nGuidelines for Data Processing and Analysis of the International Physical Activity Questionnaire (IPAQ). Reference Source\n\nBermúdez V, Torres Y, Apruzzese V, et al.: Alcohol drinking patterns in the adult population from the Maracaibo municipality, Zulia – Venezuela. Revista Latinoamericana de Hipertensión. 2014; 9(3). Reference Source\n\nAmerican Diabetes Association: 2. Classification and Diagnosis of Diabetes. Diabetes Care. 2017; 40(Suppl 1): S11–S24. PubMed Abstract | Publisher Full Text\n\nBermúdez V, Cabrera M, Mendoza L, et al.: High-sensitivity c-reactive protein epidemiological behavior in adult individuals from Maracaibo, Venezuela. Revista Latinoamericana de Hipertensión. 2013; 8(1): 22–29. Reference Source\n\nBermúdez V, Rojas J, Martínez MS, et al.: Epidemiologic Behavior and Estimation of an Optimal Cut-Off Point for Homeostasis Model Assessment-2 Insulin Resistance: A Report from a Venezuelan Population. Int Sch Res Notices. 2014; 2014: 616271. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlberti KG, Eckel RH, Grundy SM, et al.: Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation. 2009; 120(16): 1640–1645. PubMed Abstract | Publisher Full Text\n\nBermúdez V, Rojas J, Salazar J, et al.: Sensitivity and Specificity Improvement in Abdominal Obesity Diagnosis Using Cluster Analysis during Waist Circumference Cut-Off Point Selection. J Diabetes Res. 2015; 2015: 750265. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBermudez V, Olivar LC, Torres W, et al.: Dataset 1 in: Cigarette smoking and metabolic syndrome components: a cross-sectional study from Maracaibo City, Venezuela. F1000Research. 2018. Data Source\n\nBenowitz NL, Hukkanen J, Jacob P 3rd: Nicotine chemistry, metabolism, kinetics and biomarkers. Handb Exp Pharmacol. 2009; (192): 29–60. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPereira J, Peñaranda D, Reyes A, et al.: Prevalencia de factores de riesgo cardiovascular en América Latina: una revisión de la evidencia publicada de 2010 a 2015. Revista mexicana de cardiología. 2015; 26(3): 125–139. Reference Source\n\nPalloni A, Novak B, Pinto-Aguirre G: The enduring effects of smoking in Latin America. Am J Public Health. 2015; 105(6): 1246–1253. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBermudez V, Rojas J, Salazar J, et al.: The Maracaibo city metabolic syndrome prevalence study: primary results and agreement level of 3 diagnostic criteria. Revista Latinoamericana de Hipertensión. 2014; 9: 1–15. Reference Source\n\nKang JH, Song YM: Association between cotinine-verified smoking status and metabolic syndrome: analyses of Korean National Health and Nutrition Examination Surveys 2008-2010. Metab Syndr Relat Disord. 2015; 13(3): 140–148. PubMed Abstract | Publisher Full Text\n\nYildiz D: Nicotine, its metabolism and an overview of its biological effects. Toxicon. 2004; 43(6): 619–632. PubMed Abstract | Publisher Full Text\n\nLv J, Chen W, Sun D, et al.: Gender-specific association between tobacco smoking and central obesity among 0.5 million Chinese people: the China Kadoorie Biobank Study. PLoS One. 2015; 10(4): e0124586. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTuovinen E, Saarni S, Männistö S, et al.: Smoking status and abdominal obesity among normal- and overweight/obese adults: Population-based FINRISK study. Prev Med Rep. 2016; 4: 324–330. PubMed Abstract | Publisher Full Text | Free Full Text\n\nClair C, Chiolero A, Faeh D, et al.: Dose-dependent positive association between cigarette smoking, abdominal obesity and body fat: cross-sectional data from a population-based survey. BMC Public Health. 2011; 11(1): 23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWinsløw UC, Rode L, Nordestgaard BG: High tobacco consumption lowers body weight: a Mendelian randomization study of the Copenhagen General Population Study. Int J Epidemiol. 2015; 44(2): 540–550. PubMed Abstract | Publisher Full Text\n\nFujiyoshi A, Miura K, Kadowaki S, et al.: Lifetime cigarette smoking is associated with abdominal obesity in a community-based sample of Japanese men: The Shiga Epidemiological Study of Subclinical Atherosclerosis (SESSA). Prev Med Rep. 2016; 4: 225–232. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWilkins JN, Carlson HE, Van Vunakis H, et al.: Nicotine from cigarette smoking increases circulating levels of cortisol, growth hormone, and prolactin in male chronic smokers. Psychopharmacology (Berl). 1982; 78(4): 305–308. PubMed Abstract | Publisher Full Text\n\nJoshi SR, Anjana RM, Deepa M, et al.: Prevalence of dyslipidemia in urban and rural India: the ICMR-INDIAB study. PLoS One. 2014; 9(5): e96808. PubMed Abstract | Publisher Full Text | Free Full Text\n\nUeyama C, Horibe H, Yamase Y, et al.: Association of smoking with prevalence of common diseases and metabolic abnormalities in community-dwelling Japanese individuals. Biomed Rep. 2017; 7(5): 429–438. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSun G, Li Z, Guo L, et al.: High prevalence of dyslipidemia and associated risk factors among rural Chinese adults. Lipids Health Dis. 2014; 13(1): 189. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTakata K, Imaizumi S, Kawachi E, et al.: Impact of cigarette smoking cessation on high-density lipoprotein functionality. Circ J. 2014; 78(12): 2955–2962. PubMed Abstract | Publisher Full Text\n\nRobertson D, Tseng CJ, Appalsamy M: Smoking and mechanisms of cardiovascular control. Am Heart J. 1988; 115(1 Pt 2): 258–63. PubMed Abstract | Publisher Full Text\n\nGreen MS, Jucha E, Luz Y: Blood pressure in smokers and nonsmokers: epidemiologic findings. Am Heart J. 1986; 111(5): 932–940. PubMed Abstract | Publisher Full Text\n\nKaneko M, Oda E, Kayamori H, et al.: Smoking was a Possible Negative Predictor of Incident Hypertension After a Five-Year Follow-up Among a General Japanese Population. Cardiol Res. 2012; 3(2): 87–93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOnat A, Uğur M, Hergenç G, et al.: Lifestyle and metabolic determinants of incident hypertension, with special reference to cigarette smoking: a longitudinal population-based study. Am J Hypertens. 2009; 22(2): 156–162. PubMed Abstract | Publisher Full Text\n\nMcEwen A, West R, McRobbie H: Motives for smoking and their correlates in clients attending Stop Smoking treatment services. Nicotine Tob Res. 2008; 10(5): 843–50. PubMed Abstract | Publisher Full Text\n\nal'Absi M, Wittmers LE, Erickson J, et al.: Attenuated adrenocortical and blood pressure responses to psychological stress in ad libitum and abstinent smokers. Pharmacol Biochem Behav. 2003; 74(2): 401–10. PubMed Abstract | Publisher Full Text\n\nLeone A: Smoking and Hypertension. J Cardiol Curr Res. 2015; 2(2): 00057. Reference Source\n\nKim Y, Jeong SM, Yoo B, et al.: Associations of smoking with overall obesity, and central obesity: a cross-sectional study from the Korea National Health and Nutrition Examination Survey (2010-2013). Epidemiol Health. 2016; 38: e2016020. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChelland Campbell S, Moffatt RJ, Stamford BA: Smoking and smoking cessation -- the relationship between cardiovascular disease and lipoprotein metabolism: a review. Atherosclerosis. 2008; 201(2): 225–35. PubMed Abstract | Publisher Full Text\n\nOh S, Yoon Y, Lee E, et al.: Association between cigarette smoking and metabolic syndrome: the Korea National Health and Nutrition Examination Survey. Diabetes Care. 2005; 28(8): 2064–2066. PubMed Abstract | Publisher Full Text\n\nWada T, Urashima M, Fukumoto T: Risk of metabolic syndrome persists twenty years after the cessation of smoking. Intern Med. 2007; 46(14): 1079–1082. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "33887",
"date": "15 May 2018",
"name": "Eiji Oda",
"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 investigated cross-sectional associations between smoking and metabolic syndrome (MS) and its components in 2,212 adults from Maracaibo city, Venezuela. They reported that, in the multivariate analysis, the presence of MS, abdominal obesity, hypertriglyceridemia and hypo-HDL cholesterolemia were positively associated with current smoking while high blood pressure was inversely associated with current smoking. There was no significant association between smoking and hyperglycemia. Their data and study methods are fair. However, these results are not novel but well known findings.\nThe authors stated that cigarette smoking represents an independent risk factor for the development of MS in Conclusions. However, this conclusion is incorrect because the authors stated that, regarding the limitations of this study, its cross-sectional design makes it incapable of determining causality. After all, the authors had no data regarding the association between baseline smoking status and incidence (development) of MS which can only be obtained from longitudinal studies.\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? No source data required\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-565
|
https://f1000research.com/articles/8-40/v1
|
09 Jan 19
|
{
"type": "Research Article",
"title": "Fish colonization of the artificial reef at Gusung Batu Lampe Muara Badak, Kutai Kartanegara, East Kalimantan, Indonesia: a preliminary study",
"authors": [
"Iwan Suyatna",
"Nova Andika Saptura",
"Ristiana Eryati",
"Adnan Adnan",
"Muchlis Effendi",
"Tedy Hanjoko",
"Achmad Syafei Sidik",
"Nova Andika Saptura",
"Ristiana Eryati",
"Adnan Adnan",
"Muchlis Effendi",
"Tedy Hanjoko",
"Achmad Syafei Sidik"
],
"abstract": "Background: Artificial reefs can be utilized as new habitats to restore fish populations in degraded coral reef environments. This study evaluated the restoring ability of the artificial reef set at Gusung Batu Lampe should be investigated by observing the fish population colonizing the reef. Methods: A fish visual census was applied to identify the species and count their numbers at the site. Underwaterline transect was used to limit the observation area to 200 m2 measured with 50m roll meter. The reef structure comprised 34 tetrahedral concrete blocks arranged in a row, which was 60 cm x 60 cm at the base, 35 cm x 35 cm on top and 60 cm high, and has four holes. Two scuba divers descended, one to ensure the safety of the divers and the other to observe the fish. Each dive took place over 30–45 minutes, and was conducted at 09:00, 13:00 and 17:00. Surveys were performed four times: in November and December 2016, and in January and July 2017. Results: Results showed that fish colonization composed of 180 to 283 individual number with taxa between 13 and 25 species. Number of fish in colonization was found significant difference among observation times, and fish species from the family Chaetodontidae were observed in every survey. Conclusions: A total of 38 fish species from 16 families were identified colonizing the reef during surveys, and in four month observation the fish colonization size increased from the fish density of 1/4.444 m2 to 1/1.481 m2, confirming the ability of the reef to restore fish population.",
"keywords": [
"Line transect",
"fish visual census",
"unit of artificial reef",
"fish species",
"Chaetodontidae",
"Gusung Batu Lampe",
"Mahakam Delta",
"Kutai Kartanegara"
],
"content": "Introduction\n\nArtificial reefs, especially in East Kalimantan, are rarely utilized despite their ability to create new habitats. Nowadays, artificial reefs had been tried by the province of East Kalimantan at a particular area to rehabilitate fish habitat degradation, such as at Gusung Batu Lampe in Pangempang water, Kutai Kartanegara district. The degradation of coral reefs in the area was primarily caused by the practices of bottom trawling and blast fishing. In 2011 and 2012, the condition of coral reefs in Pangempang water was reported for the first time by Suyatna et al. (2017a), and convinced that the live coral (LC) form was recorded in only 21.0% of the area of 41.84 ha. Pangempang water belongs to the Mahakam delta which homes many of fish species. Suyatna et al. (2010) identified 43,340 fish caught by minitrawl and noticed 131 species from 87 genera, 61 families and 10 orders. A total of 13 species were observed in Mahakam river 40 km behind the river mouth (Suyatna et al., 2017b), and among of them, the longfin anchovy Setipinna sp., was found 230 km from the coastline (Suyatna et al., 2017c). Fish found in the Mahakam delta were also recognized along the coastal waters of East Kalimantan. At least 22 families from 8,291 fish were also observed at the industrial estate in Bontang, 75 km north of the delta (Suyatna et al., 2016) and 29 species in the coastal water of East Kutai district (Juliani & Suyatna, 2014).\n\nHowever, after artificial reef construction, scientific information describing fish colonization factors associated with the reef, such as fish species and fish richness remains unstudied. From the beginning, the reef is expected to restore fish populations in the area, and therefore a preliminary study is urged to be carried out for a certain duration. For this reason, this study was performed to emphasize the importance and benefits of constructing artificial reef and to identify and analyze the fish community structure associated with the reef.\n\n\nMethods\n\nAn underwater line transect and fish visual census (FVC) were applied in this survey. FVC method is commonly used at site (Lim, 2017) and line transect limits the observation area to only 200 m2 (20 m long and 10 m wide, 5 m to the right and 5 m to the left), measured using a 50 m roll meter. The distance of the observation area was around 5 km from the coast at the coordinate of 0° 13’02.1” S 117° 29’35.1” E and a 40 HP speed boat was used to transport divers to the site. The artificial reef structure comprised 34 units of tetrahedral concrete block, sizing 60 cm x 60 cm at base, 60 cm high and 35 cm x 35 cm wide on top, with four holes in each block, were placed in a row. Two SCUBA divers descended to the bottom; one ensured the safety of the divers and the other one took photographs while counting the number of fish. Fish photographs were realized with Action Kogan 4K and Nikon AW 130 underwater cameras. These cameras were also used to record the vicinity condition such as a unit of reef, and a GPS device (Garmin) was used to determine the observation site. All observational data were recorded using underwater diving slates. Observation took place for only 30 to 45 minutes in the morning at 09:00, in the early afternoon at 13:00 and at the late afternoon at 17:00, with the total duration of 90 to 135 minutes for one survey. Surveys were performed four times: in November and December 2016, and in January and July 2017.\n\nRecording and taking photographs of fish started about 20 minutes after transect was laid to provide time for fish to react to the natural habitat. Fish species identification was conducted with reference to Allen (2000); Anam & Mostarda (2012); Masuda et al. (1975); Peristiwadi (2006).\n\nWater physical parameters such as transparency, salinity and temperature were measured with water quality meter AZ 8603 (Shenzhen Hengkaituo Sci-Tech Co., Ltd, Guandong, China). Water velocity was assessed by drifting on the surface a small buoy for 10 m distance and noting the time to pass was using a stopwatch. Water depth was measured with a measuring rope.\n\nPalaeontological Statistics version 3.20 (Hammer et al., 2001) was used to analyze the diversity indices, namely dominance, Shannon’s H, evenness, and Margalef species richness index. Bray–Curtis multivariate statistics, which is known to be more effective than other analytical tools (Kurt & Merlyn, 1998) such as Euclidean, Morisita, Mahalanobis, and Jaccard statistics, was applied to analyze the similarity of fish size colonization among observation times. Fish density (individual/m2) was calculated manually.\n\n\nResults and discussion\n\nTaxonomically, 38 species belonging to 26 genera and 16 families were observed during surveys. A total of 25 species and 12 families were observed in the first survey; 22 species and 9 families in the second survey; 26 species and 11 families in the third survey; and 23 species and 13 families in the last survey (Table 1 and Table 2). All of the fish were small-bodied. In each survey, the two families Pomacentridae and Labridae were families with the most common and adaptable species as well as the most abundant in individual number, followed by the Chaetodontidae and Acanthuridae families. The remaining 12 families were only represented by one or two species. However, the four families Chaetodontidae, Siganidae, Caesionidae and Nemipteridae were present in every survey. Some families were only present in single surveys. The fish species observed at the fourth survey to represent the fish colonization of the artificial reef at Gosong Batu Lampe are shown in Figure 2.\n\n* Family shown in bold, species in italics.\n\n* Family shown in bold, species in italics.\n\nMallet et al. (2016) examined reef fish assemblages during daylight hours, 10 times a day for 34 consecutive days in a branching coral and they discovered fish abundance and taxa richness were greater in the early morning, in which the most frequent observed families was Pomacentridae and Labridae (in 100% and 99.6% of cases, respectively). In our study these two families were also encountered in every survey, and were the most specious and populous families, followed by Acanthuridae and Chaetodontidae families. Honda et al. (2013) reported that the most dominant families on the basis of species number in coral reefs were Labridae, Pomacentridae and Chaetodontidae. There were nine families identified in all surveys, but with only one species in each family, and according to Hernández-Velasco et al. (2016) the families distributed in tropical regions inhabit shallow coral reefs. Chaetodon kleinii and two other species of Chaetodontidae identified in our study indicated that Gusung Batu Lampe has potentiality to attract more coral reef fish. Chaetodon sp. are known to be obligate corallivores, which feed entirely on coral polyps (Kulbicki et al., 2011; Yusuf & Ali, 2004), and rely on reefs for breeding, nursing and shelter (Muhammad et al., 2017). In term of single species, Caesio teres, of the family Caesionidae (Fusilier), was the most abundant species and fed primarily on plankton; this species is included commercially as an important food fish. Some species of this family are distributed widely in mid-water over reefs (Bawole et al., 2014) and are utilized as bait fish to catch Tuna (Rajasuriya, 2014). Acanthurus, of the family Acanthuridae, commonly known as surgeonfish, was encountered in this study as the third major group. This group feeds upon algae.\n\nBased on the Bray–Curtis index, the number of fish at 09:00 in the morning and at 13:00 in the early afternoon was 95.0% similar, but declining less to 80.0% at 17:00 in the late afternoon (Table 3). This phenomenon of retreat of coral fish in the late afternoon could be an effect of light intensity (Rickel & Genin, 2005). Light may influence and increase water temperature (Mcleod et al., 2013), and such condition could affect the performance of coral reef fish, said otherwise they which tend to harbor or hide in reef holes and unseen (Suyatna et al., 2016). Likewise, the number of taxa between two observation times was 99.0% similar, but the similarity in the late afternoon was only 91.0%.\n\nFish density in a unit area (individual/m2) increased from the first to the last survey, from 1/4.444 m2 to 1/1.481 m2 (Table 4). This indicates that the reef at Gusung Batu Lampe potentially provide a valuable area forcoral reef fish population since the availability of refuge sites of reefs may increase fish number (Grossman et al., 1997), and reef size did not significantly affect the density of colonists (Borntrager & Farrell, 1992).\n\nThe diversity index was applied to describe regarding of how many different species are present over the surveys among the observation times. This study revealed that the number of colonists, species richness and species diversity was highest in at 09:00 in the morning and at 13:00 in the afternoon, and the highest species richness arose in the 3rd survey, particularly at 13:00 (Table 5). The other diversity indices showed a different population sizes of the colonists during this study.\n\nSome relevant water quality parameters measured at the location showed favorable water conditions for fish to live (Table 6). This is notable, since water quality factors such as current velocity may affect food intake (Belal, 2015), water transparency determines underwater visibility and fish assemblage (Enrique de Melo et al., 2009), and temperature and salinity may decrease species richness and fish abundance (Brucet et al., 2012). Fish can live in various places, but they often occupy a particular place and certain depth (Costa et al., 2013).\n\n\nConclusions\n\nOverall, 38 fish species from 16 families were successfully identified during surveys. Distribution of taxa and colonists were more frequently occurred at 09:00 and 13:00 than at 17:00. In 4 months of observation, fish density increased from 1/4.444 m2 to 1/1.481 m2 indicating the increase of fish colonization size. The most populous families were Pomascentridae and Labridae, and the obligate corallivore discovered was Chaotodon kleinii. The artificial reef should be elevated or extended to provide a more available water column for fish to be more colonized.\n\n\nData availability\n\nDataset 1. Raw water quality data for each survey. DOI: https://doi.org/10.5256/f1000research.16736.d226152 (Suyatna et al., 2018).",
"appendix": "Grant information\n\nAuthors acknowledge the Islamic Development Bank (IDB) for part financial support (the last survey; IDR. 53,000,000) as stated in the Decree of Mulawarman University Rector No 448/SK/2017, dated 16 March 2017.\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\nWe thank Faculty of Fisheries and Marine Science Mulawarman University for allowing the use of laboratory facilities, Head of the Surveillance Group of Local Community in Muara Badak: Muhammad Mansyur, Assistance of the Hydro-oceanography Laboratory Muhammad Raafi and all students involved in the field and laboratory work.\n\n\nReferences\n\nAnam R, Mostarda E: Field identification guide to the living resources of Kenya. Food and Agriculture Organization of the United Nation, Rome. 2012. Reference Source\n\nAllen G: Marine fishes of south east asia; a field guide for anglers and divers. Periplus, Singapore. 2000. Reference Source\n\nBawole R, Pattiasina TF, Kawulur EIJJ: Coral-fish association and its spatial distribution in Cenderawasih Bay National Park Papua, Indonesia. AACL Bioflux. 2014; 7(4): 248–154. Reference Source\n\nBelal IEH: Effect of water velocity on Tilapia Oreochromis niloticus Fingerlings Growth Parameters and Body Composition. J Med Biol Eng. 2015; 4(6): 457–460. Publisher Full Text\n\nBorntrager JF, Farrell TM: The effect of artificial reef size on species richness and diversity in a Florida estuary. Fla sci. 1992; 55(4): 229–235. Reference Source\n\nBrucet S, Boix D, Nathansen LW, et al.: Effects of temperature, salinity and fish in structuring the macroinvertebrate community in shallow lakes: implications for effects of climate change. PLoS One. 2012; 7(2): e30877. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCosta MR, Mattos TM, Borges JL, et al.: Habitat preferences of common native fishes in a tropical river in Southeastern Brazil. Neotrop Ichthyol. 2013; 11(4): 871–880. Publisher Full Text\n\nEnrique de Melo C, Lima JD, Francisca da Silva E: Relationships between water transparency and abundance of Cynodontidae species in the Bananal floodplain, Mato Grosso, Brazil. Neotrop Ichthyol. 2009; 7(2): 251–256. Publisher Full Text\n\nGrossman GD, Jones GP, Seaman WJ, et al.: Do Artificial Reefs Increase Regional Fish Production? A Review of Existing Data. Fisheries. Special Issue on Artificial Reef Management. 1997; 22(4): 17–23. Publisher Full Text\n\nHammer Ø, Harper DAT, Ryan PD: PAST: Paleontological statistics software package for education and data analysis. Palaeontol Electron. 2001; 4(1): 9. Reference Source\n\nHernández-Velasco A, Fernández-Rivera-Melo FJ, Melo-Merino SM, et al.: Occurrence of Holacanthus clarionensis (Pomacanthidae), Stegastes leucorus, and Stegastes acapulcoensis (Pomacentridae) at Magdalena Bay, B.C.S., Mexico. Mar Biodiver Rec. 2016; 9: 49. Publisher Full Text\n\nHonda K, Nakamura Y, Nakaoka M, et al.: Habitat use by fishes in coral reefs, seagrass beds and mangrove habitats in the Philippines. PLoS One. 2013; 8(8): e65735. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJuliani J, Suyatna I: The Stock potency of demersal fish resource at the coastal zone, East Kutai District in East Kalimantan. Int J Sci Eng. 2014; 6(2): 135–143. Publisher Full Text\n\nKulbicki M, MouTham G, Vigliola L, et al.: Major Coral Reef Fish Specie of the South Pacific with basic information on their biology and ecology. CRISP-IRD report. Noumea SPC. Banyuls/mer France. 2011. Reference Source\n\nKurt WP, Merlyn AB: Diversity and community comparison indices: Assessing macroinvertebrate recovery following a gasoline spill. Wat Res. 1998; 22(5): 619–626. Publisher Full Text\n\nLim TNS: Guideline on the assessment of coastal and marine ecosystems. Biodiversity Management Bureau. Quezon City, Philippines. 2017. Reference Source\n\nMallet D, Viliola L, Wantiez L, et al.: Diurnal temporal patterns of the diversity and the abundance of reef fishes in a branching coral patch in New Caledonia. Austral Ecol. 2016; 41(7): 733–744. Publisher Full Text\n\nMasuda H, Araga C, Yoshiro T: Coastal fishes of southern japan. Tokai Univ. Press. Japan. 1975. Reference Source\n\nMcLeod IM, Rummer JL, Clark TD, et al.: Climate change and the performance of larval coral reef fishes: the interaction between temperature and food availability. Conserv Physiol. 2013; 1(1): cot024. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMuhammad AA, Farooq S, Rabbaniha M, et al.: Occurrence of Gardiner’s Butterfly Fish, Chaetodon gardineri (Norman, 1939), (Chaetodontidae) in Coastal Waters of Pakistan. J Bas App Scie. 2017; 13: 182–184. Reference Source\n\nPeristiwadi T: Important fish species in Indonesia. Identification guidance. LIPI Press. Jakarta [Indonesia]. 2006.\n\nRajasuriya A: Field guide to reef fishes in Sri Lanka. IUCN, Sri Lanka. 2014. Reference Source\n\nRickel S, Genin A: Twilight transitions in coral reef fish: The input of light-induced changes in foraging behaviour. Anim Behav. 2005; 70(1): 133–144. Publisher Full Text\n\nSuyatna I, Bratawinata AA, Sidik AS, et al.: Demersal Fishes and their Distribution in Estuarine Waters of Mahakam Delta, East Kalimantan. Biodiversitas. 2010; 11(4): 204–210. Publisher Full Text\n\nSuyatna I, Hanjoko T, Adnan A, et al.: First record of coral reefs in the delta front of Mahakam Delta, East Kalimantan, Indonesia. AACL Bioflux. 2017a; 10(4): 687–697. Reference Source\n\nSuyatna I, Mislan, Rahman A, et al.: A Biophysical Observation of Mahakam River Around Tanjung Una of Kutai Kartanegara, Indonesia. Biodiversitas. 2017b; 18(2): 623–632. Publisher Full Text\n\nSuyatna I, Saptura NA, Eryati R, et al.: Dataset 1 in: Fish colonization of the artificial reef at Gusung Batu Lampe Muara Badak, Kutai Kartanegara, East Kalimantan, Indonesia: a preliminary study. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16736.d226152\n\nSuyatna I, Sidik AS, Almadi IF, et al.: Fish community structure in high water temperature around Bontang Industrial Estate, East Kalimantan, Indonesia. Biodiversitas. 2016; 17(2): 558–564. Publisher Full Text\n\nSuyatna I, Syahrir M, Mislan, et al.: A survey on Marine Fish Species In River of Mahakam East Kalimantan, Indonesia. Omni-Akuatika. 2017c; 13(2): 89–98. Publisher Full Text\n\nYusuf Y, Ali AB: The use of butterflyfish (Chaetodontidae) as bioindicator in coral reef ecosystems. In Phang SM and MT Brown (eds), Biomonit of trop coast eco. 2004; 175–183. Reference Source"
}
|
[
{
"id": "47332",
"date": "17 Apr 2019",
"name": "Ken Collins",
"expertise": [
"Reviewer Expertise artficial reef materials and ecology",
"seagrass",
"invasive species"
],
"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 many concerns about this paper:\nWhen was this reef deployed, and was this close to the time of the first survey?\n\nWhat is the nature of the seabed?\n\nWhat is the distance to natural reefs?\n\nHow does the survey transect relate to the artificial reefs?\n\nTables 1 & 2 are unnecessary and could simply be summarised as a single species list ranked in order of total abundance over the whole study. The raw data could be given as a separate Dataset as with the “Raw water quality data for each survey”.\n\nPage 7: “Fish colonisation size” describes numbers, not the dimensions of the fish.\n\nFigure 2 is not particularly useful, better would be a larger version the Caesio teres photographs showing the fish in relation to the reef structure.\n\nTable 3: limit similarity to 3 significant figures.\n\nTable 6: shows very similar conditions throughout and could be summarised in a single sentence.\n\nThe biggest issue, which makes the study of limited scientific value is the lack of any control data. This 1985 paper highlighted this issue (Bohnsack and Sutherland, 19851).\n\nArguably, the control is the first survey and the study tracks the fish community development over the subsequent 3 surveys. I would expect a parallel study of seabed distant from the artificial structure to show the value of creating a reef. Equally direct comparisons with natural reefs in the area would be useful.\n\nThe timescale of the whole study is very short; November 2016 to July 2017.\n\nI consider that the main value of this study is in describing the partial diurnal behaviours with the 3 set times 09:00, 13:00 and 17:00.\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": []
}
] | 1
|
https://f1000research.com/articles/8-40
|
https://f1000research.com/articles/8-39/v1
|
09 Jan 19
|
{
"type": "Research Article",
"title": "Factors influencing quality of life (QOL) amongst elderly caregivers of people living with HIV/AIDS in Phayao province, Thailand: a cross-sectional study",
"authors": [
"Pitakpong Punta",
"Ratana Somrongthong",
"Ramesh Kumar",
"Pitakpong Punta",
"Ramesh Kumar"
],
"abstract": "Background: There are many impacts on quality of life among elderly people living with HIV patients. This study aimed to assess factors influencing quality of life among elderly people living with HIV/AIDS in a northern province of Thailand. Methods: This cross-sectional study was conducted in Phayao province, Thailand. A systematic sampling technique was employed to select study participants. 152 elderly participants aged 60 years and older with a family member living with HIV/AIDS were recruited to the study. They were interviewed using the World Health Organization Quality of Life-Older Adults Module (WHOQOL-OLD) questionnaire. Stepwise multiple regression analysis was performed to determine the factors influencing quality of life among elderly people affected by family member living with HIV/AIDS. Results: The results of the study showed the mean age of elderly participants was 67.20 + 52 years, most of which were female (97 persons, 63.8%). The mean time taking care of HIV/AIDS patients was 6.61+ 4.96 years. In term of health status among the elderly participants, the majority did not have chronic diseases (61.4%), amongst those with chronic diseases (38.6%), hypertension and diabetes were the most common. The average quality of life score was at a fair level. The time taking care of HIV/AIDS patients and health status were significant predictors of quality of life among participants 8.1 % (R2=.081; p < .05). Conclusion: In order to improve quality of life among elderly caregivers to family member living with HIV/AIDS, time taking care of HIV/AIDS patients and health status should be focused on, amongst other factors. Help and support from the government, community, health organizations, academic research, and family members can help improve quality of life amongst the elderly.",
"keywords": [
"Influencing factors",
"Quality of life",
"Elderly",
"HIV/AIDS"
],
"content": "Introduction\n\nThe numbers of elderly people age 60 years and older are increasing rapidly around the world. It is expected to increase from an estimated 900 million in 2015 to nearly 2 billion in 2050. Most of aging population is found in developing countries1. Thailand is one such developing countries; and the number of elderly people is expected to rise from 13.20% of the total population in 2010, to 32.10% in 20402. Phayao province is located in the northern part of Thailand. Phayao Provincial Health office3 reported that the number of elderly people increased from 7.4% in 1992, to 18% in 2017. Many within this population are providing care for their children and grandchildren infected with HIV/AIDS. HIV/AIDS continues to be a major global public health issue, and it has claimed more than 35 million lives so far. In 2017, 940,000 people died from HIV/AIDS globally. There were approximately 36.9 million people living with HIV/AIDS; and almost 1.8 million people became newly infected worldwide in 20174\n\n440,000 HIV/AIDS infected patients have been registered with private hospitals in Thailand over the last thirty years. The majority of these patients are between 25–39 years of age and they were unemployed. When classifying the prevalence of HIV/AIDS infection by region, the majority of infected are found in the northern part of Thailand. The prevalence of HIV/AIDS in the northern region is highest amongst commercial (15.10%), and non- professional sex workers (10.20%)5. According to the Phayao Provincial Health office,3 the prevalence of HIV/AIDS infection from 2013 to 2016 was 974.89, 1052.31, 1117.15 and 1161.82 respectively, with incidence rates of 75.75, 72.83, 59.56, and 44.67. Elderly people in Phayao province are currently taking care of family members living with HIV/AIDS. The vast majority of HIV/AIDS’s parents are 50 years and over, and many of them are 60 years and above. The impact of a HIV/AIDS infected person on their caregivers can occur through numerous routes including (1) straining of caregiver and associated opportunity costs, (2) providing financial and material support, (3) raising the survival rate in their grandchildren, (4) suffering from emotional stress, and (5) losing old-age support that the child would have provided6. The study of Sung-Jae Lee, Li Li7 reported that the burden on a HIV/AIDS patient family caregiver in Thailand can be ‘moderate to severe’ or ‘severe’ burden (66.50%).\n\nThere are both direct and indirect impacts on the individuals from the spread of HIV on families and communities; for instance, a child becomes an orphan when his/her parent dies from the disease; elders has to take care of HIV/AIDS infected family member; and sometimes elderly people have to take care of the patient’s child in the case of the patient dies or has a physical disability. Attention has mainly focused on the infected group and their children but rarely considers the caregivers who are also greatly impacted8,9. Knodel, Chanpen10 mentioned that caregivers are most commonly the parents, and it is reported that a parent provides care for almost two thirds of adult who die of HIV/AIDS. Caregivers are not infected with HIV/AIDS, but they have the burden of looking after people living with the disease, often as well as taking care of the patient’s children11.\n\nTherefore, this study focused on quality of life among elderly people affected with family members living with HIV/AIDS. There are 13 districts in Phayao province. Mueang Phayao District is reported to have one of the highest prevalence’s of HIV/AIDs. The majority of elderly people living in this community provide care to HIV/AIDs infected persons, mostly their own children. Elderly people face challenges such as the need to earn extra money to support the family financially, taking care of their grandchildren, dealing with their own physical decline in old age, and chronic illnesses, which in themselves lead to a lower quality of life12,13. Quality of life can be one of indicators of a healthy life in older age. The World Health Organization (WHO) defines quality of life (QOL) as “an individual’s perception of life in the value system and context of culture in which she or he lives and relation to her or his goals, expectations, concerns and standards”. This study utilized the World Health Organization Quality of Life-Older Adults Module (WHOQOL-OLD) questionnaire to assess quality of life amongst elderly people who are caregivers to individuals infected with HIV/AIDS. The objectives of the study were to assess quality of life amongst this population, and to find factors influencing quality of life amongst elderly people affected by family member living with HIV/AIDS in Phayao province, Thailand.\n\n\nMethods\n\nThis study was a cross-sectional study of older people, aged over 60 years old, who were affected by family members living with HIV/AIDS in Phayao Province, Thailand. The study was performed from January - February 2015. Participants were screened before participating in the study with the inclusion criteria: (a) male or female aged 60 years and older; (b) providing care to family members living with HIV/AIDS disease; and (c) willing to participate in the study. The exclusion criteria included (a) infection with HIV/AIDS, (b) participant having communication problems such as hearing lost; and (c) having a physical disability.\n\nSample size calculation was done by using G power V.3.1.9.3 (α = 0.05, Effect size = 0.15), Cohen suggest that, if two groups' means don't differ by 0.2 standard deviations or more, the difference is trivial, even if it is statistically significant.\n\nThe required sample size was 138 elderly participants, this was adjusted for a 10% drop-out rate, generating a total sample size of 152 elderly participants. This research was conducted in the areas of high infection rates of HIV/AIDS. Information from the Phayao Provincial Health Office, private clinic data, and health promoting hospital showed that there are 13 sub-districts in the area of Mueang Phayao district reported high rates of HIV/AIDS infection and receiving treatment for HIV/AIDS. A simple random sampling technique was applied to select Ban Tam and Ban Tom sub-districts for this study, from the list of 13 sub-districts. Information of the respondents were taken from health centers data, and they were invited to participate in this study.\n\nData collectors were trained on how use the data collection tool, and briefed on the study prior to starting the study. The survey technique was face to face interviews. The process took approximately 30 minutes for each participant. The questionnaire included closed-ended questions, and consisted of 2 parts: Part 1. Socio-demographic characteristics questionnaire including gender, age, marital status, education, occupation, income, illness, condition of HIV/AIDS, time taking care of HIV/AIDS infected individuals, social and community activity, and leisure activity in the form of gardening. Part 2. World Health Organization Quality of Life questionnaire-version for older people (WHO QOL-OLD)14,15. The questionnaire includes 24 items using rating scales and it includes 6 facets: sensory abilities, autonomy, past-present, future activities, social participation, death and dying. Back translation was used to translate questionnaire from English to the Thai language. In terms of validity, the questionnaire was validated by three experts in the field of study. The reliability test of the questionnaire was 0.88. We have assessed the Quality of life into three levels; low, medium and high based on the mean score analyzed from the data.\n\nDescriptive statistics including frequency, percentage, mean, standard deviation, maximum, and minimum were used to describe general characteristics information of participants. Maximum, minimum, mean and standard deviation were used to display quality of life scores. Correlation coefficient and predictor were used to display relationship of general characteristics information (gender, age, marital status, education, occupation, income, health status, social activity participation, time of taking care HIV/AIDs infected person) and quality of life among elderly people. Stepwise multiple regression was used to assess factors influencing quality of life among the elderly people affected by family member living with HIV/AIDS. Data was analyzed by using SPSS version 20.\n\nAll participants received information regarding the research objectives and procedures of the study. Written informed consent was obtained from all participants. All information of participants was kept confidential. The study was approved by Ethics Committee from The College of Public Health Sciences, Chulalongkorn University (case No 193/2558).\n\n\nResults\n\nThe majority of participants were female (63.80%), aged between 60–69 years old (mean = 67.20, SD = 52), and married (63.80%). Most of them obtained education at primary school (78.90%). More than half were still working (67.76%), with an average monthly income less than 100 US dollars per month (99.30%). When classified by income, the majority had a sufficient income (67.11%). Most of the participants were free from chronic illness (61.84%), of those who did (38.16%), hypertension (20.68%) and diabetes (23.68%) were the most common. All had suffered from the chronic illness for over a year. More than half of HIV/AIDS infected persons had no disease symptoms 64 (42%). Most of the elderly participants 80 (52%) have been providing care to HIV/AIDS infected family members for 5 years, and 72 (47%) elderly people have been providing care for more than 5 years. Regarding social activity participation, 141 (92%) of the elderly participants have joined community activities and been actively involved in these social gatherings (Table-1).\n\nThe results showed that quality of life scores for the elderly participants were either fair (Score of 56-88 score), which the majority of the sample reported (134 participants, 88.20%), or low (Score of 24-55) reported by 18 participants (11.8%). The reported QOL ranged from 44 to 87 (x̄= 73.32, S.D = 10.76). When classified into each facet, it showed sensory ability (SAB) score Min= 5, Max = 17 (x̄= 10.63 S.D = 2.21), autonomy (AUT) score Min= 5, Max = 17 (x̄ = 12.27 S.D = 2.33), past, present, future activity (PPF) score Min= 7, Max = 8 (x̄= 12.99 S.D = 2.50), social participation (SOP) score Min= 8, Max = 18 (x̄ = 12.27 S.D = 2.33), death and dying (DAD) score Min= 4, Max = 17 (x̄= 10.61 S.D = 2.83), intimacy (INT) score Min= 8, Max = 20 (x̄= 13.48 S.D = 2.71) (Table -2 and Table 3).\n\nThe relationship between predictors regarding the analysis of the linear relationship (Milticolinearity) found each predicted variable had correlation less than 0.08 (Table 4).\n\n* p < .05 ** p < .01\n\nStepwise multiple regression analysis with variables including age, marital status, occupation, income, health status, social activity participation and time of providing care is presented in (Table-5). Time providing care and health status had a statistically significant relationship with quality of life among the elderly with a p-value of 0.05 (F = 6.567, p < .01), and power of prediction of 8.10 (R = .285, R Square = .081).\n\nRemarks Model 1 R = .228, R2 Square = .052, S.E = 10.514, n =152\n\nModel 2 R = .285, R2 Square = .081, S.E = 10.387, n =152\n\nConstant and regression coefficients analysis with health benefits of exercise and Quality of life is presented in (Table-6).\n\n\nDiscussion\n\nThis study focused on analyzing variables influencing quality of life among elderly people affected by family member living with HIV/AIDS. Variables including time taking care of the infected individual and health status are key variables influencing quality of life among elderly people within this study. According to previous research16 general characteristics including age, marital status, education, income, and social activity participation have a significant relationship with quality of life in elderly people; however, this study focused on elderly caregivers to HIV/AIDS patients, which may influence the results. It can be concluded that elderly people who have been proving care for family member living with HIV/AIDS disease for a long time, may have a lower quality of life. Knodel17 and Pungchompoo, Pothiban9 reported that as age declines, elderly people have an increased risk of high blood pressure, diabetes, and many health other complications leading to serious illness. Furthermore, they will experience physical decline and cognitive impairment.\n\nCaregiving to HIV/AIDS is a very challenging task; however, when they have to take care of a person with serious illness for such a long time, they may experience fatigue and exhaustion while giving care. Elderly people in this study have to provide emotional and physical support to their family member infected with HIV/AIDS, and the children of infected person; they have been caring for the HIV/AIDS patients for 1–5 years. As a result, they have no time to take care of themselves, and this can exacerbate their own health issues, particularly in those with chronic illness. This is consistent with many previous research studies that indicate elderly caregiver may experience adverse health effects including muscle strain, fatigue, exhaustion, high blood pressure, and/or arthritis when providing extreme day to day care9,10,18,19. This in turn leads to a lower quality of life8. In addition, they also have to earn extra money to support family members which increases the burden on them. Similar to previous studies, we found that the burden of care increases when they have to take care of an ill person for a long time, and they are likely to have a reduced quality of life due to being responsible for providing financial support to the family, and take care of their HIV/AIDS infected family members20,21\n\nHealth status is also one of the predictor’s of quality of life in elderly people. It is found to have a statistically significant relationship with quality of life. It can be concluded that as age increases, their physical ability declines, accompanied by an increased risk of chronic diseases such as hypertension, diabetes and high blood pressure that may lead to a lower quality of life among elderly people3,13. Previous study found being ill has a significant relationship with quality of life among the elderly (p=0.01)16 which is similar to this study. The study of Nobrega, Jaluul22 found that the QOL among elderly patients who suffered from chronic diseases can be affected by multi-morbidity in the physical domain and probably also in the psychological domain. This study was conducted in selected part of Thailand which cannot be generalized.\n\n\nConclusion\n\nIn conclusion, in order to improve quality of life among elderly caregivers to family member living with HIV/AIDS, the time taking care of HIV/AIDS patients, and health status should be focused on. Help and support from the government, community, health organizations, academic research, and family members can help improve quality of life amongst the elderly. In addition heath promoting hospitals and local government should have a home visit program regularly to ensure their needs are met.\n\n\nData availability\n\nOpen Science Framework: Factors influencing quality of life (QOL) amongst elderly caregivers of people living with HIV/AIDS in Phayao province, Thailand: a cross-sectional study, https://doi.org/10.17605/OSF.IO/N7BEK23\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\nThis work was supported Chulalongkorn University and the Higher Education Research Promotion and National Research University Project of Thailand [WCU-056 AS-57].\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\nWe would like to thank Chulalongkorn University and the Higher Education Research Promotion and National Research University Project of Thailand, Office of the Higher Education Commission for providing financial support for the study, and thanks to the participants for taking part in the study.\n\n\nReferences\n\nPreston SH, Stokes A: Sources of Population Aging in More and Less Developed Countries. Popul Dev Rev. 2012; 38(2): 221–236. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYodmai K, Phummarak S, Sirisuth JC, et al.: Quality of Life and fear of falling among an aging population in semi rural, Thailand. J Ayub Med Coll Abbottabad. 2015; 27(4): 771–74. PubMed Abstract\n\nPhayao Provincial Health office: Conclude about Budget Annual in 2017. Phayao Provincial Health office. 2017.\n\nShetty P: Grey matter: ageing in developing countries. Lancet. 2012; 379(9823): 1285–87. PubMed Abstract | Publisher Full Text\n\nGhys PD, Williams BG, Over M, et al.: Epidemiological metrics and benchmarks for a transition in the HIV epidemic. PLoS Med. 2018; 15(10): e1002678. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVanLandingham M, Knodel J, Im-Em W, et al.: The Impacts of HIV/AIDS on Older Populations in Developing Countries: Some Observations Based on the Thai Case. J Fam Issues. 2000; 21(6): 777–805. Publisher Full Text\n\nLee SJ, Li L, Jiraphongsa C, et al.: Caregiver burden of family members of persons living with HIV in Thailand. Int J Nurs Pract. 2010; 16(1): 57–63. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKnodel J: The changing impact of the AIDS epidemic on older-age parents in the era of ART: evidence from Thailand. J Cross Cult Gerontol. 2012; 27(1): 1–15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPungchompoo W, Pothiban L, Panuthai S: Needs and response to needs of older persons affected by HIV/AIDS in Chaing Mai province. Nurs J. 2015; 42(3).\n\nKnodel J, Saengtienchai C, Im-Em W, et al.: The Impact of AIDS on Parents and Families in Thailand: A Key Informant Approach. Res Aging. 2001; 23(6): 633–70. Publisher Full Text\n\nOgunmefun C, Gilbert L, Schatz E: Older female caregivers and HIV/AIDS-related secondary stigma in rural South Africa. J Cross Cult Gerontol. 2011; 26(1): 85–102. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSilverstein M, Giarrusso R: Aging and Family Life: A Decade Review. J Marriage Fam. 2010; 72(5): 1039–1058. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSomrongthong R, Hongthong D, Wongchalee S, et al.: The Influence of Chronic Illness and Lifestyle Behaviors on Quality of Life among Older Thais. Hindawi Publishing Corporation, Biomed Res Int. 2016; 2016: 1–6, 2525941. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYodmai K, Somrongthong R, Kumar R: Determinants of Quality of Life among Rural Elderly Population in Khonkean Province of Thailand. J Liaquat Uni Med Health Sci. 2018; 17(3): 180–4. Publisher Full Text\n\nPolit DF, Hungler BP: Nursing research: Principles and methods (6thed.). Philadelphia: JB Lippincott. 1999. Reference Source\n\nHongthong D, Somrongthong R, Ward P: Factors Influencing the Quality of Life (Qol) Among Thai Older People in a Rural Area of Thailand. Iran J Public Health. 2015; 44(4): 479–85. PubMed Abstract | Free Full Text\n\nKnodel J: Researching the Impact of the AIDS Epidemic on Older-age Parent in Africa: Lessons from Studies in Thailand. British of Gerontology. British Society of Gerontology. 2005; 15(2): 16–22. Reference Source\n\nAsuquo EF, Adejumo P, Etowa J, et al.: Fear of HIV Susceptibility Influencing Burden of Care among Nurses in South-East Nigeria. World J AIDS. 2013; 3: 231–8. Publisher Full Text\n\nAsuquo EF, Etowa JB, Akpan MI: Assessing Women Caregiving Role to People Living With HIV/AIDS in Nigeria, West Africa. SAGE Open. 2017; 7(1). Publisher Full Text\n\nChanprasit C, Lertmunlikaporn S, Lertpoonwilaikul W: Problem and Needs of Thai Elderly infected and Affected by HIV/AIDS: Stakeholders Perspectives. J Health Science. 2007; 16(1): 113–17.\n\nCohen I: Subjective quality of life the elderly affected by HIV/AIDS. The Faculty of Arts and Social Sciences at Stellenbosch University. 2016. Reference Source\n\nNobrega TC, Jaluul O, Machado AN, et al.: Quality of life and multimorbidity of elderly outpatients. Clinics (Sao Paulo). 2009; 64(1): 45–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKumar R: Factors Influencing Quality of Life (QOL) amongst Elderly Caregivers of People Living with HIV/AIDS in Phayao Province, Thailand: A Cross-Sectional Study. OSF. 2018. http://www.doi.org/10.17605/OSF.IO/N7BEK"
}
|
[
{
"id": "42860",
"date": "22 Mar 2019",
"name": "Faisal Abbas",
"expertise": [
"Reviewer Expertise Child Health",
"Undernutrition",
"Gender and women 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\nThere are many impacts on quality of life among elderly people living with HIV patients. This study aimed to assess factors influencing quality of life among elderly people living with HIV/AIDS in a northern province of Thailand. This cross-sectional study was conducted in Phayao province, Thailand using systematic sampling technique selecting 152 elderly participants aged 60 years and older with a family member living with HIV/AIDS were recruited to the study. The results of the study showed the mean age of elderly participants was 67.20 + 52 years, most of which were female (97 persons, 63.8%). The mean time taking care of HIV/AIDS patients was 6.61+ 4.96 years. In term of health status among the elderly participants, the majority did not have chronic diseases (61.4%), amongst those with chronic diseases (38.6%), hypertension and diabetes were the most common. The average quality of life score was at a fair level. The time taking care of HIV/AIDS patients and health status were significant predictors of quality of life among participants 8.1 % (R2=.081; p < .05). In order to improve quality of life among elderly caregivers to family member living with HIV/AIDS, time taking care of HIV/AIDS patients and health status should be focused on, amongst other factors. Help and support from the government, community, health organizations, academic research, and family members can help improve quality of life amongst the elderly.\nThis paper is interesting and written nicely. Data analyses is much improved and according to the objectives of the study. Language is also acceptable and the manuscript in my opinion need no further revision.Therefore, I would like to confirm that the manuscript is acceptable in its current form for indexing in the journal.\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": "43904",
"date": "05 Apr 2019",
"name": "Umer Farooq",
"expertise": [
"Reviewer Expertise Epidemiology",
"Health Statistics",
"Quality of Life",
"Health Systems"
],
"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 very useful article and will have policy implications especially in countries with medium to high HIV/AIDS prevalence. The results advocate the importance of the need of support from social agencies and development sectors for the patients of HIV/AIDS to improve their quality of life. I believe the results of their article can be used as a guiding principal.\nThe title needs to be rephrased. You may avoid words such as \"factors influencing\" and \"a cross sectional study\" – there is no need to put in abbreviations in the title. I believe an appropriate title would be: “Quality of life among elderly caregivers of people living with HIV/AIDS in Phayao Province, Thailand\".\nThe article is otherwise well written and the study design and the statistical methods employed are correct.\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/8-39
|
https://f1000research.com/articles/8-36/v1
|
09 Jan 19
|
{
"type": "Opinion Article",
"title": "What is reproducibility?",
"authors": [
"Gerben ter Riet",
"Bram W.C. Storosum",
"Aeilko H. Zwinderman",
"Bram W.C. Storosum",
"Aeilko H. Zwinderman"
],
"abstract": "The debate on reproducibility in biomedicine will gain precision only if we agree what reproducibility means. Importantly, reproducibility should be distinguished from validity (“truth”). We propose the application of an equivalence trials framework to clarify the concept of reproducibility by changing the (narrow) equivalence zone around a zero difference by a zone of reproducibility around (a) previous finding(s).",
"keywords": [
"reproducibility",
"replicability",
"repeatability",
"agreement",
"validation",
"truth",
"methodology",
"equivalence"
],
"content": "Introduction\n\nReproducibility is said to be a core principle of scientific progress. Nevertheless, poor reproducibility has recently been shown to haunt preclinical research1,2, translational research3, medicine4 and psychology5. False-positive initial results due to random chance or incorrect study design were among the reasons implicated, as well as data-dredging, publication bias and misconduct. Others called irreproducible results ‘biased’1 and ‘unreliable’5.\n\nComing from a background of meta-analysis with its countless examples of unexplained heterogeneity and an ingrained appreciation of sampling variability, we were surprised that these outcries cited above were not accompanied by a formal definition of the concept of reproducibility. Goodman et al. did define three types of reproducibility (methods, results, and inferences) and stated that confusion arises when, inadvertently, people use reproducibility as a synonym for “truth”6. We read their paper as being about truth although its title suggests otherwise. Our paper is about reproducibility sensu stricto and we revisit some basic definitions of reproducibility, notice that these definitions are problematic, and argue that the concept of equivalence in randomized trials may be fruitfully applied to sharpen our understanding of what we mean by reproducibility. We propose that investigators aiming to reproduce others’ findings should pay more attention to predefining a margin of (unacceptable) discordance with existing findings.\n\n\nDiscussion\n\nBox 1 shows two formal definitions of the concept of reproducibility.\n\n\n\nDefinition 1:\n\n“The value below which the absolute difference between two single test [or study, our addition] results may be expected to lie with a probability of 95%, when the results are obtained by the same method and equipment from identical test material in the same setting by the same operator within short intervals of time. A test or measurement [or study, our addition] is reproducible if the results are identical or closely similar each time it is conducted (Synonym, repeatability)”7\n\nDefinition 2:\n\n“The degree of agreement among a set of observations […] after all known sources of error are accounted for (Synonym, precision)”8\n\nNote the following differences between definitions 1 and 2:\n\n(i) In definition 1, reproducibility is taken to be a binary concept: a result is either reproduced or not. Definition 2, takes reproducibility to be a continuous concept, like a degree of concordance.\n\n(ii) Related to (i), definition 1 implies the subjective choice of a difference, δ, whose value will depend on the measurement problem at hand. Definition 2 avoids a choice of δ.\n\n(iii) Definition 1 chooses the value ‘95’ for the confidence interval to be used. Definition 2 avoids subjective choices of a particular confidence level, such as 95, 90, 68 etc.\n\n(iv) Only definition 2 emphasizes measurement that is free of bias.\n\nReproducibility studies may be seen as a type of equivalence trials (see Figure 1). Briefly, in classic superiority trials, we pose a statistical null hypothesis of no difference, which we then seek to reject to conclude that a difference exists. In equivalence trials, we define a (narrow) zone around a zero difference (between, say, our new drug and an existing one) and we establish equivalence if the entire confidence interval for the reproducibility study lies inside that zone. In this article, we propose to replace the difference of zero by the (pooled) value of (the) previous study or studies (vertical line in Figure 1). The width of the grey equivalence zone or “zone of reproducibility” is crucial and it seems sensible to define it pragmatically for each research situation separately. Without concrete ideas about the maximal width of this zone, judgments of when a result counts as a reproducibility can be quite subjective. For example, Begley and Ellis considered positive results as not reproduced if the replicate findings were not sufficiently robust to drive a drug-development program. Ioannidis considered the results of a therapeutic intervention as reproduced if the researcher’s final interpretation of the data in both studies was that the intervention was effective (or ineffective). Figure 1, however, shows that even in situations in which one has strictly defined the width of the zone and a suitable type of confidence interval, undecided outcomes may still occur (situations 5–7, Figure 1).\n\nNumbers in brackets refer to the 9 scenarios; horizontal lines are xx% confidence intervals (CI), where xx=95, 90, or 68 etc; short vertical lines depict point estimates; the grey area signifies the zone of reproducibility; delta (δ) refers to the maximal absolute value below which reproducibility (concordance with (an) existing finding(s)) is deemed present. Scenarios 1–4: reproducibility is present since the new point estimate and its entire 95%CI interval lie within the grey zone; scenarios 5–6: presence of reproducibility is uncertain since the point estimate lies inside the grey zone, but the xx%CI does not; scenario 7: presence of reproducibility is uncertain since the point estimate lies outside the grey zone, but part of its xx%CI lies inside; scenario 8–9: absence of reproducibility since point estimate and corresponding xx%CIs are outside the grey zone. Note, that two components are subjective: (1) the choice of δ, although preferably it should be chosen with a thorough understanding of theory or application of the research problem, and (2) the type of confidence interval since other choices than a 95%CI may be possible and defensible. Note also that, even after delta and the type of confidence limit have been chosen, uncertainty may persist if confidence limits overlap the boundaries of delta.\n\nReproducibility studies imply healthy scepticism: “Can we reproduce this finding?” In contrast with the comment cited above, which states that irreproducible results are biased, we emphasize that (ir)reproducibility of results says nothing about the validity of the previous nor of the current findings. For that, we need (validity) judgments about rigor of study design and execution. Meta-analyses of many small, but concordant, studies that were subsequently negated by the result of a single mega-trial (believed by many to represent the truth) illustrate this situation9.\n\nIn conclusion, the concept of reproducibility (repeatability, precision) should be distinguished from validity (“truth”). Furthermore, an equivalence trials framework can be fruitfully used to clarify the concept of reproducibility if we change the (narrow) equivalence zone around a zero difference by a zone of reproducibility around (a) previous finding(s). Care should be exercised when selecting sensible margins (delta) to decide on reproducibility of results10.\n\n\nData availability\n\nNo data is associated with this article.",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nBegley CG, Ellis LM: Drug development: Raise standards for preclinical cancer research. Nature. 2012; 483(7391): 531–533. PubMed Abstract | Publisher Full Text\n\nPrinz F, Schlange T, Asadullah K: Believe it or not: how much can we rely on published data on potential drug targets? Nat Rev Drug Discov. 2011; 10(9): 712. PubMed Abstract | Publisher Full Text\n\nHackam DG, Redelmeier DA: Translation of research evidence from animals to humans. JAMA. 2006; 296(14): 1731–1732. PubMed Abstract | Publisher Full Text\n\nIoannidis JP: Contradicted and initially stronger effects in highly cited clinical research. JAMA. 2005; 294(2): 218–228. PubMed Abstract | Publisher Full Text\n\nOpen Science Collaboration: PSYCHOLOGY. Estimating the reproducibility of psychological science. Science. 2015; 349(6251): aac4716. PubMed Abstract | Publisher Full Text\n\nGoodman SN, Fanelli D, Ioannidis JP: What does research reproducibility mean? Sci Transl Med. 2016; 8(341): 341ps312. PubMed Abstract | Publisher Full Text\n\nPorta M: A Dictionary of Epidemiology, VI ed. Oxford: Oxford University Press, 2014. Publisher Full Text\n\nMiettinen OS: Epidemiological Research: Terms and Concepts. Dordrecht: Springer, 2011. Publisher Full Text\n\nCappelleri JC, Ioannidis JP, Schmid CH, et al.: Large trials vs meta-analysis of smaller trials: how do their results compare? JAMA. 1996; 276(16): 1332–1338. PubMed Abstract | Publisher Full Text\n\nDekkers OM, Cevallos M, Buhrer J, et al.: Comparison of noninferiority margins reported in protocols and publications showed incomplete and inconsistent reporting. J Clin Epidemiol. 2015; 68(5): 510–517. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "42838",
"date": "13 Feb 2019",
"name": "Steven N. Goodman",
"expertise": [
"Reviewer Expertise Statistical inference",
"research reproducibility",
"epidemiology",
"clinical 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\nThis is a thoughtful piece that attempts to offer a construct that will help define research reproducibility. They say that their purpose is to offer an operational definition of reproducibility that they claim a previous paper entitled “What does research reproducibility mean?\" (with myself as first author) did not address:\n\n“Goodman et al. did define three types of reproducibility (methods, results, and inferences) and stated that confusion arises when, inadvertently, people use reproducibility as a synonym for “truth”6 . We read their paper as being about truth although its title suggests otherwise.”\n\nThe claim that the 2016 paper 1 was about truth and not reproducibility is not quite right. Let us see exactly what the prior paper said:\n\nResults reproducibility (previously described as replicability) refers to obtaining the same results from the conduct of an independent study whose procedures are as closely matched to the original experiment as possible. … this might be clear in principle but is operationally elusive. The problem arises in settings where there is substantial random error in any result, making unclear the criteria for considering results to be “the same.” The intuition and logic of results reproducibility are derived from systems that are deterministic or for which the signal-to error ratio is exceedingly high. But, when the same intuition and logic are applied to studies with substantive stochastic components, the paradigm of accumulating evidence might be more appropriate than any binary criteria for successful or unsuccessful replication. …. Statistical significance by itself tells very little about whether one study has “replicated” the results of another. For example, two studies that show identical 10% survival differences between the treatment and control arms would have very different degrees of statistical significance if their sample sizes were substantially different. If one was highly significant and the other far from significance, the two studies might be reported individually as supporting opposite conclusions, in spite of the fact that they are mutually corroborative. An interpretive error complementary to the one described above involves the assumption that multiple studies that fail to demonstrate statistical significance necessarily confirm the absence of an effect. ….. It is easier to statistically define non-replication than replication, through statistical tests of heterogeneity, which can evaluate whether the difference between two or more experimental results might be due to the play of chance. Two or more studies are judged to be statistically heterogeneous when the between-study variance in reported effects is substantially greater than what is expected from sampling error. Such tests, however, are greatly underpowered and therefore unreliable when comparing several studies, particularly when they are small or imprecise (17). Conversely, when there are many large studies, tests for heterogeneity might demonstrate statistical heterogeneity (and, therefore, lack of results reproducibility) even if the effect sizes of different studies are close (17) and regarded as scientifically equivalent. Therefore, a preferred way to assess the evidential meaning of two or more results with substantive stochastic variability is to evaluate the cumulative evidence they provide vis-á-vis a hypothesis of interest and not whether one contradicts or discredits the other through the lens of statistical significance. --------------------------------------------------- So it should be apparent that the 2016 paper does indeed address exactly what this paper addresses, including the notion that results can differ in statistical significance yet be regarded as \"scientifically equivalent”. That is essentially what this paper goes on to try to define, with a \"zone of equivalence” that defines \"scientifically equivalence”. But as these authors acknowledge in the legend of Figure 1, “…. even after delta and the type of confidence limit have been chosen, uncertainty may persist if confidence limits overlap the boundaries of delta.” The problem is that the confidence intervals will quite often cross the boundaries of delta, and so we are left with the same conundrum that the original paper said was inescapable.\nThe point of the original paper was that trying to define “reproducibility” was in the end not very constructive, and that if we turned our attention instead to the cumulative evidence represented by several studies, instead of whether they \"reproduced\" or not, we could avoid these distinctions, which ultimately serve little purpose. The authors here are right that I believe that the goal of science, and of scientific studies, is to move us closer to the truth. I contend that debates about which and how many studies reproduced do not, particularly when that definition is elusive. The prior paper did indeed tell us that convergence on the truth should be our lodestar, not an arbitrarily defined reproducibility criterion, which even with the improved version offered here does not provide a clear verdict in the vast majority of cases.\n\nI agree with the authors that it is helpful to have some notion of differences that make a difference, and thereby scientific equivalence. But the degree of imprecision in most health studies precludes an unambiguous conception of reproducibility even if one introduces that interval. I also agree with the authors’ conclusion that “….the concept of reproducibility (repeatability, precision) should be distinguished from validity (“truth”)”, but disagree that the purpose of assessing reproducibility is anything other than getting at the truth, and still believe that the cumulative evidence model and not the reproducibility model - which cannot be clearly defined - is what gets us there.\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": []
},
{
"id": "44139",
"date": "15 Feb 2019",
"name": "C. Glenn Begley",
"expertise": [
"Reviewer Expertise Particular interest in the area of scientific rigour and 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\nThe issue of data reproducibility is central to science and is worthy of ongoing discussion. Although the authors state at the outset that \"Reproducibility is said to be a core principle of scientific progress\", to me it IS a core principle.\nData that cannot be reproduced does not serve as a foundation upon which others can build. The authors propose that a formal definition of reproducibility be pre-defined, pre-agreed when investigators attempt to reproduce others' findings. Pre-defining criteria for failure or success is always valuable. It removes the natural bias to interpret results to suit one's prejudice post-hoc, and may also to be useful in this context.\nHowever, the paper that is cited on which I am the first author (Begley and Ellis, 2012)1 does not really support the authors' argument. The authors state that we considered results as not reproduced if findings were not sufficiently robust to drive a drug development program. That is correct: we were focused on developing new drugs, and could not justify moving forward if the results were not reproduced. But what was truly shocking, was that in the majority of cases it was the original authors themselves who were unable to reproduce their own findings. Our 'standard operating procedure' when unable to reproduce key findings was to go to the original laboratory and watch them repeat their experiments (which required a confidentiality agreement and precludes disclosure of those laboratories). Their failure to reproduce their findings certainly negated that \"research\" as being sufficiently robust to drive a drug development program.\nUsing the criteria outlined in Figure 1 of this paper, the published experiments are illustrated in Scenario 8, while the repeated (and unpublished) experiments are illustrated in Scenario 9. In our experience therefore, the pre-definition of confidence intervals appeared unnecessary. It was this experience that led us to conclude that the fundamental problem was not really one of \"reproducibility\", nor a problem of definition, it was rather a problem of cherry-picking, p-hacking, HARKING, lack of controls, lack of repeats, lack of blinding. This poor experimental methodology was employed so as to generate an initial data set that was sufficiently exciting to justify publication.\nTherefore, I do not think the issue regarding lack of reproducibility is simply one of a lack of clear definition, rather, in my view, it is systematic and driven by the perverse incentives that govern our current system. Thus focusing solely on agreeing on a definition, does not lead us toward finding a solution to a problem that is deeply embedded in our system, and in fact has been used by some to distract and argue that there isn't really an issue of irreproducibility - its simply about a definition.\nFrom my perspective, it would be valuable for these Authors to acknowledge these wide-spread scientific practices as central to the issue of \"reproducibility\".\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": []
},
{
"id": "44140",
"date": "25 Feb 2019",
"name": "Ksenija Bazdaric",
"expertise": [
"Reviewer Expertise research integrity",
"open science",
"methodology",
"plagiarism",
"publishing"
],
"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 giving me the opportunity to read this manuscript. It was very interesting. As opinion pieces are not supposed to be very long I understand that not all concepts/constructs could have been explained in detail. I think the article is about the definition of reproducibility and should be understood as such. I would advise acceptance with minor changes.\n\nComments: Introduction The aim of the article is clear but the title is not. I would advise adding a change to the title to 'What is reproducibility? - a definition proposal'. I would advise repeating at least one of the most known definitions in order to ease the reading to general audience. Readers must understand the flaws of existing definitions in order to embrace the new one(s). Perhaps this one: NSF report as “replicability,” which refers to “the ability of a researcher to duplicate the results of a prior study if the same procedures are followed but new data are collected.”1 or some other.\n\nDiscussion When you state “we were surprised that these outcries cited above were not accompanied by a formal definition of the concept of reproducibility”. I wonder what do you mean by formal, a statistical definition or a more narrow definition, or a more exact definition? Please make your statement more clear if possible. I really like Box 1 and the 2 definitions proposed. The first model might work and be valuable for life sciences and biomedicine, while the second can be more used in psychology and other social sciences. Figure 1. is very clear with clear examples. I especially like example 5 and the explanation in the discussion.\n\nConclusion Of course, the concept of reproducibility should be distinguished from validity. If a measurement is not valid there is no need for replication at all. But if you think the general audience is misunderstanding the terms and that they have to be distinguished please give a short definition of validation in brief, because it is a widely used term in psychology but not in other disciplines.\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": []
}
] | 1
|
https://f1000research.com/articles/8-36
|
https://f1000research.com/articles/7-1524/v1
|
21 Sep 18
|
{
"type": "Research Article",
"title": "Relative effectiveness of a full versus reduced version of the ‘Smoke Free’ mobile application for smoking cessation: a randomised controlled trial",
"authors": [
"David Crane",
"Harveen Kaur Ubhi",
"Jamie Brown",
"Robert West",
"David Crane",
"Harveen Kaur Ubhi",
"Jamie Brown"
],
"abstract": "Background: Smartphone applications (apps) are popular aids for smoking cessation. Smoke Free is an app that delivers behaviour change techniques used in effective face-to-face behavioural support programmes. The aim of this study was to assess whether the full version of Smoke Free is more effective than the reduced version. Methods: This was a two-arm randomised controlled trial. Smokers who downloaded Smoke Free were randomly offered the full or reduced version; 28,112 smokers aged 18+ years who set a quit date were included. The full version provided updates on benefits of abstinence, progress (days smoke free), virtual ‘badges’ and daily ‘missions’ with push notifications aimed at preventing and managing cravings. The reduced version did not include the missions. At baseline the app recorded users’: device type (iPhone or Android), age, sex, daily cigarette consumption, time to first cigarette of the day, and educational level. The primary outcome was self-reported complete abstinence from the quit date in a 3-month follow-up questionnaire delivered via the app. Analyses conducted included logistic regressions of outcome on to app version (full versus reduced) with adjustment for baseline variables using both intention-to-treat/missing-equals smoking (MES) and follow-up-only (FUO) analyses. Results: The 3-month follow-up rate was 8.5% (n=1,213) for the intervention and 6.5% (n=901) for the control. A total of 234 participants reported not smoking in the intervention versus 124 in the control, representing 1.6% versus 0.9% in the MES analysis and 19.3% versus 13.8% in the FUO analysis. Adjusted odds ratios were 1.90, 95%CI=1.53-2.37 (p<0.001) and 1.50, 95%CI=1.18-1.91 (p<0.001) in the MES and FUO analyses respectively. Conclusions: Despite very low follow-up rates using in-app follow up, both intention-to-treat/missing equals smoking and follow-up only analyses showed the full version of the Smoke Free app to result in higher self-reported 3-month continuous smoking abstinence rates than the reduced version.",
"keywords": [
"smoking cessation",
"RCT",
"smartphone application",
"smoke free"
],
"content": "Introduction\n\nSmartphone applications (apps) are used by many smokers to aid cessation but currently little evidence exists on their effectiveness. The Smoke Free app (smokefreeapp.com) is very popular worldwide, with some 4,000 new downloads per day. When it first became available, it was the subject of a trial with users finding and downloading the app from the app store being randomly assigned to a full version or a reduced version. This provided an opportunity to assess whether the full version was more effective than the reduced version in an effectiveness study closely mirroring the real-world scenario of interest. This paper reports the findings from that trial.\n\nThe Smoke Free app was developed using behaviour change techniques (BCTs) found in effective behavioural support programmes for smoking cessation1. A description of the app is given in Supplementary File 1. Behavioural support delivered via a smartphone could help smokers to stop. Internet-based support has been found in some cases to aid cessation2 and smartphone apps can provide this functionality with the added advantage of being readily accessible at almost any time. Two prospective studies of users of smoking cessation apps3,4, a randomised controlled trial (RCT) comparing an app with a text messaging intervention5 and two RCTs comparing mindfulness-based apps with other apps6,7 found self-reported success rates that were higher than would be expected from unaided cessation. One RCT has found that an app acting as a decision aid for smokers interested in stopping smoking resulted in higher 6-month abstinence rates than an information-only app8. Another RCT examined the effectiveness of a set of app components as an aid to cessation in pregnant smokers; engagement with the app was low and no specific components were found to increase short-term self-reported abstinence rates9. To date no RCTs have been published comparing apps designed to provide ongoing support for quit attempts with unaided quitting, or more intensive versus less intensive versions of an app.\n\nEvaluating the effectiveness of smoking cessation apps versus unaided cessation in RCTs is complicated by the fact that apps are widely available and participants who are randomised to the unaided quitting condition are likely to be motivated to drop out of the study or use one of the many freely available apps. An alternative is to compare full and reduced versions of an app in which the reduced version is sufficiently credible that participants who are randomised to receive it are not motivated to drop out of the study or seek out another app. That was the approach used in the present study.\n\nAnother challenge for RCTs of apps is how to address the problem of loss to follow up. With sufficient resources, high follow-up rates can be obtained in such trials6,10. However, the methods used can lead to problems of generalizability; study engagement processes that involve face-to-face visits, incentives and contracts on the part of participants that may exclude a substantial proportion of the target population. Moreover, the resources required are prohibitive for the kind of agile, iterative evaluation that is required during the development of these interventions, where evaluations need to be undertaken repeatedly11. Automated outcome assessment using the smartphone is low cost and does not require procedures that may undermine generalizability. Despite the fact that it may result in very low follow-up rates that is the approach used in the current study.\n\nIn smoking cessation trials it is common practice to use an intention-to-treat approach with participants lost to follow up considered to have resumed smoking12. This may bias effect sizes downwards if loss to follow up occurs for reasons other than relapse to smoking13. Conversely it may bias effect sizes upwards if the intervention condition leads to higher follow-up rates than the control condition. Only including participants who are successfully followed up may overestimate absolute success rates if participants refuse to engage with follow up because they have resumed smoking, but this would not affect the odds ratio comparing two conditions since this bias would affect both intervention and comparison groups equally. This approach is also immune to bias caused by differential follow-up in intervention and control groups. In practice ‘missing equals smoking’ (MES) and ‘follow-up only’ (FUO) approaches tend to produce very similar odds ratios in smoking cessation RCTs14, though the percentage point difference between conditions varies considerably. Multiple imputation methods are increasingly being used to estimate values for missing data arising from loss to follow up (e.g. Westmaas et al.15). However, these are only viable when the proportion of values that are missing is low. To address biases arising from loss to follow up, both the MES and FUO approaches were used in the present study. It may be expected that the true percentage point difference and odds ratios lie somewhere between the estimates provided by these two methods.\n\nBiochemical verification of abstinence is recommended in smoking cessation trials because of psychological pressure to claim abstinence12. However, this is highly resource intensive and may undermine generalizability to smokers who would use an app but not a more intensive interaction. Also, in a trial where there is no greater psychological pressure to claim abstinence in one condition than another, use of self-report should not bias the estimated effect size. Therefore, this study used self-report for outcome assessment.\n\nDuration of follow up is an important consideration in smoking cessation trials. Conventionally, follow-up at least 6 months after the start of an intervention is considered appropriate for definitive trials while shorter durations are acceptable for proof of concept trials12. A recent systematic review of continuous abstinence rates in smoking cessation trials has recently found, however, that rates at 6-month and 12-month follow up can be accurately predicted from findings after 12 weeks16. Loss to follow up may be greater with longer follow up so in the present study participants were followed up 12 weeks after the target quit date.\n\nThus, this study addressed the question of whether the full version of the Smoke Free app would result in higher 12-week self-reported continuous abstinence rates than a reduced version of the app in smokers downloading the app and using it to set a quit date.\n\n\nMethods\n\nParticipants were individually randomly allocated by the app on a 1:1 ratio using a random number generator to the full or reduced version and followed up automatically by the app 12 weeks after the target quit date to assess the outcome. This study was not registered as a randomised controlled trial because the lead author was not aware of this requirement at the time the data were collected, and so the study must be considered exploratory. We used the CONSORT-SPI checklist in preparing this report17 (Supplementary File 2). The study was approved by the University of East London Ethics Committee.\n\nParticipants were not actively recruited and received no financial incentive for taking part. Smokers who downloaded the Smoke Free app between February 2013 and January 2015 were informed by the app that it was being used in an evaluation and asked for permission to use their data for research purposes. If they agreed they completed baseline measures and were randomly assigned by a computer-generated random number sequence to be offered a full or reduced version of the app. Consent was given by users by means of the touchscreen on their device. They were then included in the analysis if they met the following criteria: aged 18 years or over, smoked cigarettes at the time of registration, set only one quit date, and used the app at least once on or after their target quit date. Those users who had started their quit attempt before the date of registration were excluded, and if users registered more than once on the same device only data from the first registration was used.\n\nParticipants were aware that they were taking part in an experiment but were not aware of the details of the condition to which they had not been assigned.\n\nSample size was determined pragmatically by recruiting from the point where the app was in a form that was stable to the deadline for delivery of the lead author’s project report. A total of 28,112 participants were included in the sample, of whom 14,228 received the full version and 13,884 received the reduced version.\n\nThe full version of the Smoke Free app took smokers through the first month of their quit attempt by helping them maintain their resolve by setting a clear goal, monitor their progress towards that goal and become aware of benefits achieved to date. There were several components: 1) a calculator that tracked the total amount of money not spent on buying cigarettes and the number of cigarettes not smoked; 2) a calendar that tracked the amount of time elapsed since cessation; 3) a scoreboard that awarded virtual ‘badges’ to users for not smoking; 4) progress indicators that informed users of health improvements expected since the start of their quit attempt; and 5) daily missions that were assigned from the start of a user’s quit date for one calendar month.\n\nThe daily missions included behaviour change techniques that research has suggested are likely to improve the chances of avoiding and resisting cravings and thereby promote abstinence18–20. A list of the daily missions can be found in the Supplementary File 1.\n\nThe full version of Smoke Free received daily push notifications for one calendar month from the start of their quit date. Users were prompted to open the app to read each day’s mission. The time of the push notification was preset to 8am local time but this could be changed to a time of the user’s preference. For screenshots of the app see Supplementary File 1.\n\nThe reduced version of the app was the same as the full version but without the daily missions.\n\nAfter consenting to take part in the experiment, users were asked to provide information on their: age, sex, educational level (high school or secondary school, undergraduate degree, or post-graduate degree), daily cigarette consumption, and time to first cigarette of the day (<5 minutes, 5–30 minutes, 31–60 minutes, >60 minutes)21.\n\nAfter filling out the baseline questionnaire, users were then requested to record their target quit date which could be any date in the past or future (with those having already quit being excluded from the analysis).\n\nThe primary outcome measure was self-reported continuous abstinence up to 12-week follow-up. The app sent users a push notification 12 weeks after the target quit date asking them to open the app and respond to a questionnaire. The questionnaire asked: 1) “Have you smoked at all in the last three months?” to which they could respond: “No, not a puff”, “1–5 cigarettes”, or “More than 5 cigarettes”. Those who responded “not a puff” were considered to be abstinent.\n\nBaseline characteristics of the two groups were compared using chi-squared tests or analyses of variance as appropriate. Outcomes were compared using logistic regression analyses with and without adjusting for all baseline variables. Two analytic approaches were used: 1) MES in which smokers who were lost to follow-up were counted as having smoked, and 2) FUO in which only smokers who responded to the 3-month follow-up were included in the analysis. Odds ratios and 95% confidence intervals were computed, along with p-values.\n\nData used in the analyses are available as Supplementary File 3 as an SPSS file and the SPSS syntax used to run the analyses is provided in Supplementary File 4. The full data set, including variables not included in the analysis are provided in Dataset 1.\n\n\nResults\n\nTable 1 shows participants’ baseline characteristics and Figure 1 shows the numbers allocated to each group and followed up. Participants who received the reduced version of the app were older, smoked more cigarettes per day, started smoking earlier in the day and were more likely to designate a quit date that was after the date of registration, but the differences were small. Complete data are shown in Dataset 122.\n\n*p<0.05 for comparison between groups, not adjusted for number of comparisons.\n\nOf the participants, 2,114 (7.5%) were followed up (full version 1,213, 8.5%, reduced version 901, 6.5%). In the MES analysis 1.6% (n=234) of the participants in the intervention group and 0.9% (n=124) of the participants in the control group reported as being abstinent from smoking (unadjusted Odds ratio=1.86; 95% CI=1.49-2.31; (p<0.001). In the FUO analysis, 19.3% in the intervention group and 13.8% in the comparison group reported being abstinent (unadjusted Odds Ratio=1.50; 95% CI=1.18-1.90; p<0.001).\n\nTable 2 shows the results from logistic regression analyses with app version and all baseline variables entered together. In both the MES and FUO analyses, participants randomized to the full version of the app had higher odds of reporting successful abstinence with odds ratios almost identical to the unadjusted regression analyses. A number of baseline variables also predicted reported abstinence. Older participants, and in the MES analysis, those with longer time to their first cigarette of the day were more likely to report abstinence, while those using Android (versus iOS devices) and those whose quit date was after the date of registration were less likely to report abstinence.\n\n*p<0.05 for linear trend or comparison with reference.\n\n\nDiscussion\n\nIn both the MES and FUO analyses the full version of the Smoke Free app produced higher self-reported abstinence rates than the reduced version 12 weeks after the target quit date. The odds ratios ranged between 1.50 and 1.90 in the two analyses and the percentage point differences ranged between 0.7% and 5.5%.\n\nEven with very low follow-up rates the study found a small but clear advantage to the full version of the app which may be attributed to the inclusion of the daily missions. This is on top of whatever effect the reduced version of the app may have had. Even in the follow-up only analysis the abstinence rates were relatively low, and lower than is found in studies involving face to face support or pharmacotherapy23. Therefore, this app should not be regarded as a substitute for those forms of support but rather as an alternative in smokers who do not wish to engage with them or do not have access to them. It is possible that this app could increase abstinence rates in smokers using such forms of support but this remains to be tested.\n\nThe fact that an intervention effect was found in both the MES and FUO analyses strengthens confidence that it was not due to bias arising from differential loss to follow up. The fact that adjusting for baseline variables did not influence the odds ratios similarly adds confidence that the results were not due to smokers who found it easier to stop being more likely to be followed up in the intervention condition.\n\nStrengths of the current study are the large sample size, the fact that it assessed an app that is very popular and therefore needs to be evaluated, and high generalizability to the population of interest, i.e. smokers finding the app on apps stores. Limitations are the very low follow-up rate, use of self-reported abstinence, the relatively short follow-up duration and the absence of process measures to assess what mediated the intervention effect.\n\nResearch continues into apps to support smoking cessation24–29, with improvements in technology providing new opportunities for intervention content, such as virtual reality and use of wearable devices. The popularity of the Smoke Free app should make it a useful vehicle for testing innovations in smoking cessation support, building on the version of the app tested in this study. Research is needed into improving follow-up rates without compromising generalizability and within the resource constraints operating on companies and research groups seeking to build incrementally on app performance.\n\nIn conclusion, the full version of the Smoke Free smartphone app appears to have small effect in improving 12-week abstinence rates in smokers trying to quit. This provides a basis for building a programme of incremental improvement in effectiveness.\n\n\nData availability\n\nDataset 1. Full de-identified data from each study participant, including download dates, quitting dates and all other data input into the app. Data are available in SAV and CSV formats. DOI: https://doi.org/10.5256/f1000research.16148.d21854122.",
"appendix": "Grant information\n\nDuring conduct of the study Harveen Kaur Ubhi’s post was funded by the UK’s National Centre for Smoking Cessation and Training and Robert West and Jamie Brown’s salary were funded by Cancer Research UK (C1417/A22962).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nSupplementary material\n\nSupplementary File 1. Details on the Smoke Free app, including each day’s mission.\n\nClick here to access the data.\n\nSupplementary File 2. Completed CONSORT-SPI checklist.\n\nClick here to access the data.\n\nSupplementary File 3. Data used in the analyses.\n\nClick here to access the data.\n\nSupplementary File 4. SPSS command syntax for the analyses.\n\nClick here to access the data.\n\n\nReferences\n\nUbhi HK, Michie S, Kotz D, et al.: Characterising smoking cessation smartphone applications in terms of behaviour change techniques, engagement and ease-of-use features. Transl Behav Med. 2016; 6(3): 410–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTaylor GMJ, Dalili MN, Semwal M, et al.: Internet-based interventions for smoking cessation. Cochrane Database Syst Rev. 2017; 9: Cd007078. PubMed Abstract | Publisher Full Text\n\nUbhi HK, Michie S, Kotz D, et al.: A mobile app to aid smoking cessation: preliminary evaluation of SmokeFree28. J Med Internet Res. 2015; 17(1): e17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBricker JB, Copeland W, Mull KE, et al.: Single-arm trial of the second version of an acceptance & commitment therapy smartphone application for smoking cessation. Drug Alcohol Depend. 2017; 170: 37–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBuller DB, Borland R, Bettinghaus EP, et al.: Randomized trial of a smartphone mobile application compared to text messaging to support smoking cessation. Telemed J E Health. 2014; 20(3): 206–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBricker JB, Mull KE, Kientz JA, et al.: Randomized, controlled pilot trial of a smartphone app for smoking cessation using acceptance and commitment therapy. Drug Alcohol Depend. 2014; 143: 87–94. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarrison KA, Pal P, O'Malley SS, et al.: Craving to Quit: A Randomized Controlled Trial of Smartphone app-based Mindfulness Training for Smoking Cessation. Nicotine Tob Res. 2018. PubMed Abstract | Publisher Full Text\n\nBinDhim NF, McGeechan K, Trevena L: Smartphone Smoking Cessation Application (SSC App) trial: a multicountry double-blind automated randomised controlled trial of a smoking cessation decision-aid 'app'. BMJ Open. 2018; 8(1): e017105. PubMed Abstract | Free Full Text\n\nTombor I, Beard E, Brown J, et al.: Randomized factorial experiment of components of the SmokeFree Baby smartphone application to aid smoking cessation in pregnancy. Transl Behav Med. 2018. PubMed Abstract | Publisher Full Text\n\nBrown J, Michie S, Geraghty AW, et al.: Internet-based intervention for smoking cessation (StopAdvisor) in people with low and high socioeconomic status: a randomised controlled trial. Lancet Respir Med. 2014; 2(12): 997–1006. PubMed Abstract | Publisher Full Text\n\nMichie S, Yardley L, West R, et al.: Developing and Evaluating Digital Interventions to Promote Behavior Change in Health and Health Care: Recommendations Resulting From an International Workshop. J Med Internet Res. 2017; 19(6): e232. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWest R, Hajek P, Stead L, et al.: Outcome criteria in smoking cessation trials: proposal for a common standard. Addiction. 2005; 100(3): 299–303. PubMed Abstract | Publisher Full Text\n\nBarnes SA, Larsen MD, Schroeder D, et al.: Missing data assumptions and methods in a smoking cessation study. Addiction. 2010; 105(3): 431–7. PubMed Abstract | Publisher Full Text\n\nBlankers M, Smit ES, van der Pol P, et al.: The Missing=Smoking Assumption: A Fallacy in Internet-Based Smoking Cessation Trials? Nicotine Tob Res. 2016; 18(1): 25–33. PubMed Abstract | Publisher Full Text\n\nWestmaas JL, Bontemps-Jones J, Hendricks PS, et al.: Randomised controlled trial of stand-alone tailored emails for smoking cessation. Tob Control. 2018; 27(2): 136–46. PubMed Abstract | Publisher Full Text\n\nJackson S, McGowan J, Ubhi HK, et al.: Modelling continuous abstinence rates over time from clinical trials of pharmacological interventions for smoking cessation. Addiction (Abingdon, England). 2018.\n\nGrant S, CONSORT-SPI Group: The CONSORT-SPI 2018 extension: a new guideline for reporting social and psychological intervention trials. Addiction. 2018. PubMed Abstract | Publisher Full Text\n\nMichie S, Churchill S, West R: Identifying evidence-based competences required to deliver behavioural support for smoking cessation. Ann Behav Med. 2011; 41(1): 59–70. PubMed Abstract | Publisher Full Text\n\nMichie S, Hyder N, Walia A, et al.: Development of a taxonomy of behaviour change techniques used in individual behavioural support for smoking cessation. Addict Behav. 2011; 36(4): 315–9. PubMed Abstract | Publisher Full Text\n\nWest R, Walia A, Hyder N, et al.: Behavior change techniques used by the English Stop Smoking Services and their associations with short-term quit outcomes. Nicotine Tob Res. 2010; 12(7): 742–7. PubMed Abstract | Publisher Full Text\n\nKozlowski LT, Porter CQ, Orleans CT, et al.: Predicting smoking cessation with self-reported measures of nicotine dependence: FTQ, FTND, and HSI. Drug Alcohol Depend. 1994; 34(3): 211–6. PubMed Abstract | Publisher Full Text\n\nCrane D, Ubhi HK, Brown J, et al.: Dataset 1 in: Relative effectiveness of a full versus reduced version of the ‘Smoke Free’ mobile application for smoking cessation: a randomised controlled trial. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16148.d218541\n\nWest R, Raw M, McNeill A, et al.: Health-care interventions to promote and assist tobacco cessation: a review of efficacy, effectiveness and affordability for use in national guideline development. Addiction. 2015; 110(9): 1388–403. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Agteren JEM, Lawn S, Bonevski B, et al.: Kick.it: The development of an evidence-based smoking cessation smartphone app. Transl Behav Med. 2018; 8(2): 243–67. PubMed Abstract | Publisher Full Text\n\nSchick RS, Kelsey TW, Marston J, et al.: MapMySmoke: feasibility of a new quit cigarette smoking mobile phone application using integrated geo-positioning technology, and motivational messaging within a primary care setting. Pilot Feasibility Stud. 2017; 4: 19. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMinami H, Brinkman HR, Nahvi S, et al.: Rationale, design and pilot feasibility results of a smartphone-assisted, mindfulness-based intervention for smokers with mood disorders: Project mSMART MIND. Contemp Clin Trials. 2018; 66: 36–44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoyd M, Wilson N: Just ask Siri? A pilot study comparing smartphone digital assistants and laptop Google searches for smoking cessation advice. PLoS One. 2018; 13(3): e0194811. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBaskerville NB, Struik LL, Dash D: Crush the Crave: Development and Formative Evaluation of a Smartphone App for Smoking Cessation. JMIR Mhealth Uhealth. 2018; 6(3): e52. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang H, Jiang Y, Nguyen HD, et al.: The effect of a smartphone-based coronary heart disease prevention (SBCHDP) programme on awareness and knowledge of CHD, stress, and cardiac-related lifestyle behaviours among the working population in Singapore: a pilot randomised controlled trial. Health Qual Life Outcomes. 2017; 15(1): 49. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "38627",
"date": "08 Oct 2018",
"name": "Erica Cruvinel",
"expertise": [
"Reviewer Expertise Smoking cessation",
"substance abuse"
],
"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 is a study of the “Relative effectiveness of a full versus reduced version of the ‘Smoke Free’ mobile application for smoking cessation: a randomised controlled trial”. The use of such applications as described therein can extend the reach of smoking cessation interventions to a large number of smokers through low-cost approaches. Thus, the results of this study will be of great interest to the tobacco treatment community. I have several questions about the details of the study and some suggested edits.\n\nIntroduction: The introduction discusses most methodology decisions, such as those about the comparator group, follow up loss, analysis, and the measure of abstinence. But, it lacks some conceptual information. For instance, it does not address why the full version of the app was expected to be more effective than the reduced version. I would like to see a more conceptual description used as a background to support the development of an extended version beyond the app’s basic version. Also, I think methodology decisions would fit better under methods or study limitations. In addition, it is not clear what the authors mean by, “Also, in a trial where there is no greater psychological pressure to claim abstinence in one condition than another, use of self-report should not bias the estimated effect size“. I’m not sure if the authors refer to a non-superiority trial since the last sentence in the introduction session shows that they expected a better result in the intervention group. Furthermore, I would like to have more information about the app itself. For instance, are people from all over the world able to download the app? In which languages is it available?\nMethods: Study design: Provide more details about the method used to generate the random allocation sequence (such as blocking and block size). Participants: Were included daily and occasionally smokers? In which countries do the participants reside? Is the app only in English? Intervention and comparator Show the theory that guided the development of these interventions in this section. Measures: Did you collect data about how often the app was used and for how long? Analysis: How did you control against the duplication of data, such as if someone installed the app twice? What strategy did you use to increase follow up response rates? Did you set up a reminder function to send notifications to the user to complete the follow-up questionnaire? How did you handle missing data? (details of any imputation method)\nResults: How did the participants use the app? If you included descriptive information from both groups, then it would help us to better understand participant interactions with the app. Did you have information about whether participants used other methods to quit smoking? For instance, nicotine replacement therapy? The sentence “Older participants, and in the MES analysis, those with longer time to their first cigarette of the day were more likely to report abstinence”. This interpretation is confusing because the results from participants who smoke cigarettes longer than 60 minutes were not significant. Also, those whose quit date was after the date of registration were less likely to report abstinence. What was the comparator? Table 2: It is not clear. What is the reference in table 2?\n\nDiscussion: The fifth sentence of the introduction states: “The odds ratios ranged between 1.50 and 1.90 in the two analyses and the percentage point differences ranged between 0.7% and 5.5%”. Is it not 0.85 and 0.7? The last sentence in the third paragraph is not clear when looking at table 2. Which variable measured the difficulty of quitting? The quit rate in this study was lower compared to other studies (ex: BinDhim, McGeechan, Trevena, 20181), as well as the follow-up rates. I would like to see more discussion about how the results from this study compare to those of other apps. The results showed that those using Android (versus iOS devices) were less likely to report abstinence. What’s the implication of this data?\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": "4350",
"date": "09 Jan 2019",
"name": "Robert West",
"role": "Author Response",
"response": "The introduction does not address why the full version of the app was expected to be more effective than the reduced version. I would like to see a more conceptual description used as a background to support the development of an extended version beyond the app’s basic version. Response: We have now added this to the introduction and made reference to a paper that gives more detail. Methodology decisions would fit better under methods or study limitations. Response: We thought of doing this but in the end decided to keep it where it is because it made the methods section harder to follow and is part of the rationale for the study rather than what was actually done. It is not clear what the authors mean by, “Also, in a trial where there is no greater psychological pressure to claim abstinence in one condition than another, use of self-report should not bias the estimated effect size“. I’m not sure if the authors refer to a non-superiority trial since the last sentence in the introduction session shows that they expected a better result in the intervention group. Response: We obviously didn’t make this point clear. We were referring to the pressure that respondents might feel to falsely claim that they were abstinent. We have reworded this to try to make it clearer. I would like to have more information about the app itself. For instance, are people from all over the world able to download the app? In which languages is it available? Response: We have now added this information. It is available globally but only in the English language. Provide more details about the method used to generate the random allocation sequence (such as blocking and block size). Response: We now included this information. The app generated a random number (1 or 2) with equal probability during the registration process for each user. Were included daily and occasionally smokers? Response: As indicated in the paper all smokers were included. We now make it explicit that this included non-daily smokers. Show the theory that guided the development of these interventions in this section. Response: We now include this in the explanation as to why we thought the full intervention would be more effective than the reduced version. Did you collect data about how often the app was used and for how long? Response: We did not collect information about number of times the app was used, unfortunately. How did you control against the duplication of data, such as if someone installed the app twice? Response: We already explain that if the app was downloaded more than once on the same device, we used the first occurrence only. Please see page no 5, paragraph 4 (continued on page 6). What strategy did you use to increase follow up response rates? Did you set up a reminder function to send notifications to the user to complete the follow-up questionnaire? Response: We already explain that push notifications appeared on the home screen of the device to solicit responses to the follow up questionnaire. We now make clear that reminders were not sent. Please see page no 7, paragraph 4. How did you handle missing data (details of any imputation method)? Response: We attempted to make this clear in the paper. We used both the traditional ‘missing equals smoking’ method and follow-up only. The aim was to address different types of bias that may arise from these two methods. As noted by reviewer 1, more sophisticated imputation methods would have required much greater follow-up rates. How did the participants use the app? If you included descriptive information from both groups, then it would help us to better understand participant interactions with the app. Response: Unfortunately we do not have this information. Did you have information about whether participants used other methods to quit smoking? For instance, nicotine replacement therapy? Response: Unfortunately not. It possible that participants in the intervention group used other methods to help them quit and indeed part of the effect of the app may have been to get them to do so since advice on medication use was one of the behaviour change techniques used. The sentence “Older participants, and in the MES analysis, those with longer time to their first cigarette of the day were more likely to report abstinence”. This interpretation is confusing because the results from participants who smoke cigarettes longer than 60 minutes were not significant. Response: We have rephrased this to make clear that the association was only statistically significantly for the middle two categories of time to first cigarette. Also, those whose quit date was after the date of registration were less likely to report abstinence. What was the comparator? Response: We now make clearer that the comparator was those whose quit date was the same as the date of registration. Table 2: It is not clear. What is the reference in table 2? Response: This is the standard way to report results from logistic regression analyses with categorical independent variables. The ‘reference’ is the group against which the other groups are compared. The fifth sentence of the introduction states: “The odds ratios ranged between 1.50 and 1.90 in the two analyses and the percentage point differences ranged between 0.7% and 5.5%”. Is it not 0.85 and 0.7? Response: We think the reviewer means the second sentence in the discussion but were still not sure what to make of this comment. This sentence just summarises the results from the comparison between the intervention and comparator conditions as stated in the results. However, to make this clearer we have rephrased this sentence. The last sentence in the third paragraph is not clear when looking at table 2. Which variable measured the difficulty of quitting? Response: We now make clearer that this was based on using baseline variables that predicted abstinence. The quit rate in this study was lower compared to other studies (ex: BinDhim, McGeechan, Trevena, 2018), as well as the follow-up rates. I would like to see more discussion about how the results from this study compare to those of other apps. Response: As noted by Reviewer 1, the follow-up rate was too low to be able to gain an accurate estimate of absolute quit rates. However, we now included discussion of how the results compare with those of previous RCTs of smoking cessation apps. The results showed that those using Android (versus iOS devices) were less likely to report abstinence. What’s the implication of this data? Response: We now include a brief discussion of this finding."
}
]
},
{
"id": "39534",
"date": "23 Oct 2018",
"name": "Jonathan B Bricker",
"expertise": [
"Reviewer Expertise mhealth for smoking cessation"
],
"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 reports on the results of a fully automated RCT that compared the Smoke Free app with and without 30 days of behavior change technique skills for quitting and preventing relapse (\"missions\").\nMajor strengths:\nThe scientific premise of the study is strong. Determining the added value of behavior change techniques for smoking cessation is useful in any context (not just apps) so the study affords a real-world opportunity to test the techniques' potential effectiveness. Testing the effectiveness of a widely utilized app in real-world conditions of consumers actually downloading to help them quit smoking is valuable from the point of view of implementation research. And perhaps that is the best context from which to frame this study--a real world implementation study. Novelty. There is a dearth of studies comparing apps for smoking cessation so the study adds to a small and important literature.\nMajor weaknesses:\nTrial was not registered. This is a very serious weaknesses that impacts ethical obligations to participants, research community, and publication bias. Reputable journals would not even consider reviewing this paper knowing that is was not registered. The authors are aware of this weakness and I have no reason to believe they acted intentionally unethically. Nonetheless, it would be appropriate not to title the study a randomized trial or use that term in the abstract. Instead, I recommend the authors use the same term employed in methods section of the study design: \"Exploratory Study.\" The authors themselves call it that. Now they just need to make it clear up front. Very low retention rate (7.5%) that appears to imbalanced by a relative rate of 24% between arms. This is a very unfortunate weakness that could have been avoided with some small and immediate incentives given that only one question is asked in the follow-up survey. Nonetheless, it is what it is. The level of bias from this extent of missing data (92.5%) cannot be overcome with imputation methods. And the argument that the relative difference (not absolute quit rate) is what matters for this study is undermined by the imbalanced retention rates. With 24% more data in one arm than the other, it very possible that the difference in quit rates is simply driven by the difference in retention rates.\nModerate weaknesses:\nThe main outcome of continuous abstinence after baseline is biased by differences in quit dates and simply does not give people enough time to even reach their quit date before they would already be counted a smoker. Thus, the actual quit rates could be higher. Its hard to say with this outcome. Generalizablity is overstated for two reasons. The low retention rate makes the sample highly biased toward the most motivated people who are most likely to be reporting that they quit. Comparing the baseline characteristics of all those enrolled vs only those completed the outcome survey item would be very important and instructive about bias. The second reason it is overstated is that the sample is limited to those who have already chosen an app and this particular app to help them quit smoking. The sample age (29) is young for a mhealth smoking study, which typically is about age 40.\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": "4349",
"date": "09 Jan 2019",
"name": "Robert West",
"role": "Author Response",
"response": "Major weaknesses Trial was not registered. This is a very serious weaknesses that impacts ethical obligations to participants, research community, and publication bias. Reputable journals would not even consider reviewing this paper knowing that is was not registered. The authors are aware of this weakness and I have no reason to believe they acted intentionally unethically. Nonetheless, it would be appropriate not to title the study a randomized trial or use that term in the abstract. Instead, I recommend the authors use the same term employed in methods section of the study design: \"Exploratory Study.\" The authors themselves call it that. Now they just need to make it clear up front. Response: We agree that the fact that the trial was not registered is a weakness. It does not affect the weakness of this trial but we note that even now, large numbers of behavioural trials published in high quality journals are not pre-registered. We gave a great deal of consideration before submitting the first version as to whether to call this an RCT and in the end decided to do so because that is what it was. We have checked authoritative definitions (e.g. the UK’s National Institute for Health and Care Excellence) and they do not specify pre-registration in the definition. As the reviewer acknowledges, we make it very clear that the trial was not pre-registered. We believe that we must use the term RCT in the title etc but have added the word ‘exploratory’ to highlight this issue. Very low retention rate (7.5%) that appears to imbalanced by a relative rate of 24% between arms. This is a very unfortunate weakness that could have been avoided with some small and immediate incentives given that only one question is asked in the follow-up survey. Nonetheless, it is what it is. The level of bias from this extent of missing data (92.5%) cannot be overcome with imputation methods. And the argument that the relative difference (not absolute quit rate) is what matters for this study is undermined by the imbalanced retention rates. With 24% more data in one arm than the other, it very possible that the difference in quit rates is simply driven by the difference in retention rates. Response: We agree that the low follow up is a major weakness and prevents estimation of absolute effect size. However, we disagree that the imbalance in follow-up rates could have contributed to the relative effect. As we explain in the introduction, we conducted an additional analysis only on those followed up and still got a highly significant effect with an odds ratio of 1.5. We have explored the issue of whether results from the follow-up only sample can possibly be affected by differential follow up rate and with both analytical reasoning and extensive modelling of extreme difference we find that it cannot. We were initially surprised but once we saw it, it was obvious. To help see this, imagine a study of 1000 smokers with 500 allocated to each group. Then imagine 200 (40%) are followed-up in the intervention group and 100 (20%) in the control group. Imagine that in fact there is no difference in effect between the conditions so 50 (10%) are actually abstinent in each group. In an intent to treat analysis with missing equals smoking we would expect 20 (40% of 50) successes in the intervention group and 10 (20% of 50) in the control group which would translate to 20/500=4% in the intervention group versus 10/500=2% in the control group – apparently double the success rate in the intervention than control group which is clearly wrong. But in the follow-up only analysis the figures are 20/200=10% in the intervention group and 10/100=10% in the control group – correctly showing no effect. You can try this with any permutation of outcomes and it always comes out the same. Moderate weaknesses The main outcome of continuous abstinence after baseline is biased by differences in quit dates and simply does not give people enough time to even reach their quit date before they would already be counted a smoker. Thus, the actual quit rates could be higher. Its hard to say with this outcome. Response: We agree that the actual quit rates may well be higher. In fact, given what we observe in prospective studies of unaided quitting it would be surprising if the quit rates were not higher. We were not sure what the reviewer was referring to when saying that some participants may not have reached their quit dates by the follow up; the follow up was always 12 weeks after their designated quit date. Generalizablity is overstated for two reasons. The low retention rate makes the sample highly biased toward the most motivated people who are most likely to be reporting that they quit. Comparing the baseline characteristics of all those enrolled vs only those completed the outcome survey item would be very important and instructive about bias. The second reason it is overstated is that the sample is limited to those who have already chosen an app and this particular app to help them quit smoking. The sample age (29) is young for a mhealth smoking study, which typically is about age 40. Response: We now note in the discussion that the generalizability of the findings is limited by the low follow-up rate and to smokers. However, it should be noted that in the intention to treat analysis the sample was not limited to those followed up and the odds ratio was slightly higher than for the follow-up only sample and adjusting for baseline variables that are predictive of successful cessation also did not influence the finding. The data are available for anyone who would like to compare the baseline characteristics of those followed-up compared with who were not."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1524
|
https://f1000research.com/articles/8-35/v1
|
09 Jan 19
|
{
"type": "Research Article",
"title": "Is the public-private mix approach increasing tuberculosis case notification in Azad Jammu and Kashmir, Pakistan? A cross-sectional study",
"authors": [
"Mumtaz Hussain",
"Razia Fatima",
"Zahida Qasim",
"Aashifa Yaqoob",
"Ahmed Wali",
"Sabir Abbasi",
"Sabira Tahseen",
"Razia Fatima",
"Zahida Qasim",
"Aashifa Yaqoob",
"Ahmed Wali",
"Sabir Abbasi",
"Sabira Tahseen"
],
"abstract": "Background: Pakistan ranks among high tuberculosis (TB) burden countries, with about 200,000 missing TB cases. Despite significant increase in case detection and treatment outcomes through public-private mix (PPM) globally, the contribution of the private sector may vary in different parts of Pakistan Methods: This was a cross sectional study conducted in state of Azad Jammu and Kashmir (AJK), Pakistan. The study was a retrospective record review of routine TB notification and treatment outcomes for 2015 to 2016 in districts with and without a PPM approach. The study was conducted in all districts of AJK, including all public and private health facilities. Intra-district comparison in PPM supported districts was also carried out.\n\nResults: Total notified TB cases during 2015-16 were 11479. Districts with PPM support has notified 5882 (51.24%) of TB cases while districts without PPM support contributed 5597 (48.87%). Pulmonary clinical diagnosis was higher in PPM supported districts (45.43%) as compared to districts without PPM support (31.11%). Notification of extra pulmonary TB cases was lower in PPM supported districts (1256; 21.35%) as compared to districts without PPM support (1877; 33.54%). Unfavourable treatment outcomes (treatment failure, died, lost to follow-up) was higher in PPM facilities (5.84%). Conclusion: The study shows minimal increase in TB case finding through the PPM approach. While this is an important aspect in END TB strategy, this needs more careful evaluation.",
"keywords": [
"Public Private Mix",
"Tuberculosis",
"case notification",
"treatment outcomes",
"AJK-Pakistan"
],
"content": "Introduction\n\nTuberculosis (TB) remains a major public health problem, as it is the leading cause of death from a single infectious agent. Globally, about 10 million people developed TB disease with 1.3 million TB deaths annually1. TB case notification is challenging as there was a 3.4 million gap globally between notified and incident cases2. To ensure universal access to quality TB services is a major challenge due to lack of systematic engagement of all health care providers, especially in the private sector1.\n\nThe public sector plays a major role in control of TB in Pakistan. More than 70% of total case notification was from public sector in 20171. Similarly, a study from India showed that the public sector has contributed 84% in TB case notification3. In addition, the role of the public sector in management of TB was assessed in Thailand and it was concluded that 90% of TB cases were diagnosed and treated in public health facilities, while the private sector contributed only 10% in case notification and treatment4.\n\nThe public-private mix (PPM) approach has been suggested by the World Health Organization to engage all health care providers5. PPM approach was tested in 23 countries from 1999 to 2000 and there was an increase in case detection rate and treatment outcome remained above 85%6. Evidence suggests that PPM has the capacity to achieve increased case notification, increased treatment outcome and more importantly it also improves patient satisfaction7.\n\nSeveral studies have been conducted in the region for estimation of cost effectiveness and the impact of PPM on case detection and TB management. These studies suggested that PPM is a promising model with increased case notification and improved treatment outcome8–10. However, PPM intervention can vary in terms of accessibility, cost effectiveness, acceptability and quality11.\n\nAbout 80% of TB patients in Pakistan attend private sector facilities for their initial diagnosis and management and most of them are not reported to the National TB Control Program7. Pakistan lists among countries with the highest number of un-notified TB cases12. In 2006, Pakistan adopted PPM model which has been gradually scaled-up13. The contribution of PPM in all case notification in Pakistan has been reported to be up to 18%14. Studies from Pakistan have revealed that PPM can play a substantial role in achieving the targets of treatment success rate and case detection7,15,16. Majority of TB cases which are managed by private sector, are not notified to National TB Control Program. Although the referral of presumptive TB cases is higher (70.9%), but only 29.1% are treated17.\n\nAll studies conducted in Pakistan focused the role of private sector but there is limited data available for inter-district and intra-district comparison. There is a need to compare the PPM districts with public sector districts, which are in close vicinity and with minimal geographical differences. Therefore, this study provided an inter-district as well as intra-district comparison in intervention districts in terms of case notification and treatment outcomes. This study also provided the overall contribution of public and private sector in terms of TB case notification and treatment outcomes.\n\n\nMethods\n\nThis was a cross-sectional study based on the retrospective review of routine data of State TB control program from 2015 to 2016.\n\nAzad Jammu and Kashmir (AJK) is a self-governing state in Pakistan with a population of 4.045million. Administratively it is divided into three divisions and ten districts. The total area of AJK is 5134 square miles (13,297 square km). The topography of the area is mainly hilly with high mountains, valleys and stretches of plains. The rural urban ratio is 88:1218. State TB Control Program is actively participating in the control of TB. PPM approach was adopted in 2011 in three districts and then scaled up to four districts in 2015. PPM approach in AJK is carried out by Sub-Recipient with Global Fund’s Principle Recipient.\n\nPPM approach in four districts (Bagh, Kotli, Mirpur and Bhimber) comprises almost half of the population of AJK (Figure 1). In these four districts, the General Practitioner (GP) model has been adopted and currently 73 GPs and 14 diagnostic laboratories work actively in these districts.\n\nTB case notification including pulmonary and extra pulmonary cases was assessed. TB treatment outcomes (cured, treatment completed, lost to follow-up, treatment failure, died and not evaluated) was compared between districts.\n\nData was collected from district reports. TB case notification report and TB treatment outcomes report was used for retrospective data analysis. Data collected at district level was verified and validated with TB registers available at each TB facility in all districts.\n\nData was collected from district reports and then entered into Epi-Data software (version 3.1 EpiData Association, Odense, Denmark). Data was analysed and percentages were obtained. Case notification of districts with and without PPM support was calculated. The proportion of both groups was assessed in terms of case notification and treatment outcomes.\n\nWritten approval (849/DHS-CDC/18) was obtained from the TB program manager of AJK-Pakistan. As secondary data was used, there was no direct involvement of patients/human subjects in this study.\n\n\nResults\n\nTotal notified TB cases during 2015-16 were 11479. Districts with PPM support has notified 5882 (51.24%) of TB cases while districts without PPM support contributed 5597 (48.87%). The percent change due to PPM intervention was 2.48%. Bacteriologically positive TB cases notified by PPM districts were 1954 (33.22%). Districts without PPM support notified 1979 (35.36%) bacteriologically positive TB cases. PPM supported districts notified 2672 (45.43%) pulmonary clinically diagnosed TB cases. Clinically diagnosed TB cases in districts without PPM support was 1741 (31.11%). The percent change in pulmonary clinically diagnosed TB cases was 14.32%. Total notification of extra pulmonary TB cases was 3133. PPM supported districts contributed 1256 (21.35%) as compared to districts without PPM support 1877 (33.54%). The percent change in extra pulmonary TB cases was -12.18% (Table 1).\n\nThe analysis of TB treatment outcomes are shown in Table 2. This shows that patients lost to follow-up was less in PPM supported districts with a percent change of -2.91%. Total number of TB cases under category of “not evaluated” was lower in PPM supported districts with percent change of -1.75%.\n\nIntra-district comparison of PPM supported districts is shown in Table 3. The PPM TB health facilities contributed 15% of TB case notification with a higher proportion of clinical diagnosis and lower proportion of extra pulmonary TB cases. Public health TB facilities contributed 24% of extra pulmonary cases, while PPM TB facilities contributed only 6.41%. Total TB case notification by PPM supported districts was higher in Mirpur, followed by Bhimber, Bagh and Kotli.\n\nDistricts with PPM support: n=4.\n\nTB treatment outcomes among PPM supported districts were analysed. Unfavourable treatment outcomes (treatment failure, died, lost to follow-up) was higher in PPM facilities (5.84%).\n\n\nDiscussion\n\nIt is evident from the study that the public sector remained the major contributor in TB case notification; the public sector notified 85% TB cases while the private sector contributed 15% only. In our study, inter-district comparison showed that case notification in districts with and without PPM support was almost the same with minimal difference. Pakistan has prioritized the PPM approach and there was a 30% increase in case detection rate in 20171. Data from Pakistan suggested that TB case notification has been increased due to involvement of private health care providers14. Similar studies from Pakistan also suggested that TB case notification has been improved due to involvement of private health care providers6,7,19. All these studies were conducted in PPM supported districts, but there is limited data available to compare PPM supported districts with non-PPM districts (public sector only).\n\nIn this study, pulmonary TB clinical diagnosis was higher in PPM support districts with a percent change of 14.32%. This showed that the private sector is relying more on X-ray based diagnosis or clinical diagnosis without referring the presumptive TB cases for laboratory investigation. A prevalence survey in Pakistan also showed that the majority of pulmonary TB cases are diagnosed by radiography in the private sector20. A study conducted in Karachi-Pakistan also concluded that the numbers of TB cases from the private sector are clinically diagnosed and there is a need for strengthening reliance on TB laboratories for screening of presumptive TB cases21. Therefore, the private sector is relying more on X-ray based diagnosis rather lab based diagnosis22. Laboratory based diagnosis in the private sector is limited and more than half of all presumptive TB cases in the private sector is referred to the public sector for diagnosis23.\n\nIn this study extra-pulmonary TB case notification in districts with PPM support is lower with a percent change -12.82%. Extra pulmonary TB case notification is usually difficult and needs extra investigation with high quality tests. GP model needs more attention for selection criteria in Pakistan. Different studies from Pakistan showed that there is a significant knowledge gap between public and private sector doctors, and private doctors have lesser knowledge for diagnosis and management of TB19.\n\nTB treatment outcomes are one of the major indicators for assessing a successful TB control program. A study showed that AJK had reported a 95% treatment success rate. The proportion of unfavourable TB treatment outcomes, died, TB treatment failure and lost to follow-up, were 3.8%, 0.1% and 0.2% respectively14. Our study also showed similar results with minor difference in public and private TB health facilities. Overall TB treatment success rate was above 95% in public TB health facilities. The proportion of unsuccessful TB treatment outcomes was higher in PPM facilities (5.84%).\n\nOur study showed that TB case notification through PPM was only 15% in PPM supported districts, but there was no such significant difference when PPM supported districts were compared with non-PPM districts (public sector only). A comparative performance of public and private health sector in low and middle income countries also support our study that the public sector is more efficient, accountable and medically effective than the private sector24. In Pakistan there is need of more stringent selection criteria for GP selection to improve their involvement in TB control7. PPM is an important approach to achieve global TB targets; however it could be affected by contextual characteristics in different areas25.\n\nThe strength of the study is that it compares PPM supported districts with other districts (non-PPM supported). Previous studies in Pakistan focus on few districts while this study covers a whole region of AJK, and all districts and health facilities were included in this study.\n\nThe limitation of our study is the reliance on program data. Accuracy and completeness of the data cannot be assured. There could be many reasons for deaths, and it is important to know that deaths occur due to TB rather that to another reason.\n\n\nConclusions\n\nThe study showed minimal increase in TB case finding using the PPM approach. While this is an important aspect in END TB strategy, this needs more careful evaluation. The public sector is contributing effectively to TB case notification and has better TB treatment outcomes. Therefore, there is a need to strengthen the public sector.\n\nAlthough the PPM approach is promising and is included in the END TB strategy, our study showed that the public sector is more efficient than PPM in terms of case notification and treatment outcomes. Public sector of AJK is comparable with PPM supported districts, with more good indicators. It could be further improved by monitoring and evaluation. PPM approach needs stringent selection criteria so that it could perform more efficiently. Contextual characteristics need to be addressed while implementing PPM.\n\n\nData availability\n\nFigshare: Dataset 1: Is the public-private mix approach increasing tuberculosis case notification in Azad Jammu and Kashmir, Pakistan? A cross-sectional study. This data set contains the TB case notification and TB treatment outcomes from both public and private health facilities of AJK-Pakistan, https://doi.org/10.6084/m9.figshare.7388036.v126.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Grant information\n\nThis research was conducted through the Structured Operational Research and Training Initiative (SORT IT), a global partnership led by the Special Programme for Research and Training in Tropical Diseases at the World Health Organization (WHO/TDR).\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\nThis research was conducted through the Structured Operational Research and Training Initiative (SORT IT), a global partnership led by the Special Programme for Research and Training in Tropical Diseases at the World Health Organization (WHO/TDR). The training model is based on a course developed jointly by the International Union Against Tuberculosis and Lung Disease (The Union, Paris, France) and Médecins Sans Frontières (MSF, Geneva, Switzerland). The specific SORT IT programme that resulted in this publication was implemented by the National Tuberculosis Control Programme of Pakistan, through the support of the Global Fund to Fight AIDS, Tuberculosis and Malaria (The Global Fund, Geneva, Switzerland). The publication fee was covered by the Special Programme for Research and Training in Tropical Diseases at the World Health Organization (WHO/TDR).\n\n\nReferences\n\nWHO Global TB Report 2018. 2018; 81–4. Reference Source\n\nGlobal Tuberculosis Report 2016. [cited 2018 Mar 30]. Reference Source\n\nLal SS, Sahu S, Wares F, et al.: Intensified scale-up of public-private mix: a systems approach to tuberculosis care and control in India. Int J Tuberc Lung Dis. 2011; 15(1): 97–104. PubMed Abstract\n\nChengsorn N, Bloss E, Anekvorapong R, et al.: Tuberculosis services and treatment outcomes in private and public health care facilities in Thailand, 2004-2006. Int J Tuberc Lung Dis. 2009; 13(7): 888–94. PubMed Abstract\n\nWHO/W document, CDS/TB/2001.285. Involving Private practitioners in TB control: Issues, interventions and policy intervention. 2001. Reference Source\n\nNaqvi SA, Naseer M, Kazi A, et al.: Implementing a public-private mix model for tuberculosis treatment in urban Pakistan: Lessons and experiences. Int J Tuberc Lung Dis. 2012; 16(6): 817–21. PubMed Abstract | Publisher Full Text\n\nPethani A, Zafar M, Khan AA, et al.: Engaging general practitioners in public-private mix tuberculosis DOTS program in an urban area in Pakistan: need for context-specific approach. Asia-Pacific J Public Heal. 2015; 27(2): NP984–NP992. PubMed Abstract | Publisher Full Text\n\nSehgal S, Dewan PK, Chauhan LS, et al.: Public-private mix TB activities in Meerut, Uttar Pradesh, North India: delivering dots via collaboration with private providers and non-governmental organizations. Indian J Tuberc. 2007; 54(2): 79–83. PubMed Abstract\n\nPradhan A, Datye V, Kielmann K, et al.: Sustaining PPM-DOTS: the case of Pimpri Chinchwad, Maharashtra, India. Indian J Tuberc. 2011; 58(1): 18–28. PubMed Abstract\n\nZhang T, Guo L, Zhang S, et al.: Improving detection and notification of tuberculosis cases in students in Shaanxi province, China: an intervention study. BMC Public Health. 2011; 11(1): 147. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMalmborg R, Mann G, Thomson R, et al.: Can public-private collaboration promote tuberculosis case detection among the poor and vulnerable? Bull World Health Organ. 2006; 84(9): 752–8. PubMed Abstract | Free Full Text\n\nFatima R, Qadeer E, Hinderaker SG, et al.: Can the number of patients with presumptive tuberculosis lost in the general health services in Pakistan be reduced? Int J Tuberc Lung Dis. 2015; 19(6): 654–6. PubMed Abstract | Publisher Full Text\n\nNyirandagijimana B, Edwards JK, Venables E, et al.: Closing the gap: decentralising mental health care to primary care centres in one rural district of Rwanda. Public Health Action. 2017; 7(3): 231–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChughtai AA, Qadeer E, Khan W, et al.: Estimation of the contribution of private providers in tuberculosis case notification and treatment outcome in Pakistan. East Mediterr Health J. 2013; 19(3): 213–8. PubMed Abstract | Publisher Full Text\n\nTb N, Program C: Public Private Mix Experience for TB control. TB Situation in Pakistan. Reference Source\n\nAhmed J, Ahmed M, Laghari A, et al.: Public private mix model in enhancing tuberculosis case detection in district thatta, sindh, Pakistan. J Pak Med Assoc. 2009; 59(2): 82–6. PubMed Abstract\n\nDisease L: Public Health Action. 2013; I: 193–8.\n\nAzad Jammu & Kashmir At A Glance 2017. 2017. Reference Source\n\nNaseer M, Khawaja A, Pethani AS, et al.: How well can physicians manage tuberculosis? A public-private sector comparison from Karachi, Pakistan. BMC Health Serv Res. 2013; 13(1): 439. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrevalence_Report.pdf.\n\nCreswell J, Khowaja S, Codlin A, et al.: An Evaluation of Systematic Tuberculosis Screening at Private Facilities in Karachi, Pakistan. PLoS One. 2014; 9(4): e93858. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCentre CLT, Medicine C, Date A, et al.: Tuberculosis case management and treatment outcome: Assessment of the effectiveness of public-private mix of tuberculosis programme in Kaduna state, Nigeria Abstract in French. 2009; 7–11.\n\nFatima R, Haq MU, Yaqoob A, et al.: Delivering Patient-Centered Care in a Fragile State: Using Patient-Pathway Analysis to Understand Tuberculosis-Related Care Seeking in Pakistan. J Infect Dis. 2017; 216(suppl_7): S733–S739. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBasu S, Andrews J, Kishore S, et al.: Comparative performance of private and public healthcare systems in low- and middle-income countries: a systematic review. PLoS Med. 2012; 9(6): e1001244. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLei X, Liu Q, Escobar E, et al.: Public-private mix for tuberculosis care and control: a systematic review. Int J Infect Dis. 2015; 34: 20–32. PubMed Abstract | Publisher Full Text\n\nHussain M: Is the public-private mix approach increasing tuberculosis case notification in Azad Jammu and Kashmir, Pakistan? A cross-sectional study. figshare. Dataset. 2018. https://doi.org/10.6084/m9.figshare.7388036.v1"
}
|
[
{
"id": "42824",
"date": "11 Feb 2019",
"name": "Saw Saw",
"expertise": [
"Reviewer Expertise Tuberculosis",
"Health Systems Research",
"Public Private partnership",
"operational 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\nThis article tried to answer whether the public-private mix approach increases TB case notification in different settings.\nMethods:\nStudy setting: The authors should describe more on the PPM approach and the public sector approach especially for case finding and notification. Is there any difference in case finding strategy between PPM and public sector? If yes, it needs to be addressed in the discussion. For example, if the public sector is using active case findings and PPM is only passive case finding, case notification would be different. A table or diagram showing PPM model in Pakistan would be helpful to understand the context since different countries use different models of PPM.\n\nData sources and collection: It is suggested to elaborate more about district reports; whether data from private sectors were included and completeness and proportion of missing data in reports and register.\n\nResults:\nIt is necessary to clarify some findings especially for comparison between PPM and non-PPM and intra-district comparison. E.g. in the 2nd paragraph of the results, Table 2 shows loss to follow up was less in PPM districts. But in the last sentence of the results, it stated that unfavorable treatment outcome was higher in PPM facilities. Clear description and comparison should be made.\n\nIt is surprising to see 15% in PPM facilities and 85% in public facilities for case notification (Table 3) while 80% of TB patients in Pakistan attend private sector for initial diagnosis. Under-reporting from the private sector and incompleteness of district reports might be the reasons. These interpretations should come up clearly in the discussion session.\n\nDiscussion and Conclusions:\nOther possible factors contributing to case notification and treatment outcomes need to be addressed and discussed.\n\nCaution needs to be taken for the conclusion saying that the public sector is contributing effectively to TB case notification and has better treatment outcome just by using programme data. Weakness in reporting and recording at PPM clinics and also the public sector may also need to be considered.\n\nRecommendations: The authors stated that the PPM approach needs stringent selection criteria. However, there is no evidence or findings related to selection criteria in the results section to support this recommendation.\n\nI have read this submission. I believe that I have an appropriate level of expertise to confirm that it is acceptable. However, I would suggest this as a short report rather than a research article. I have significant reservations, as outlined 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? 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? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "58993",
"date": "02 Mar 2020",
"name": "Olusola Adedeji Adejumo",
"expertise": [
"Reviewer Expertise TB epidemiology"
],
"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\nThank you for asking me to review this manuscript. While this is an interesting topic the authors assumed that the prevalence of TB will be the same in PPM supported and Non-PPM supported district. That is why they compared them. This assumption may not hold. Factors like the presence of slums and socio-economic status of peoples may not be evenly distributed.\nIn conducting this study, authors should have compared the Case notification rates in the PPM supported districts before and after the commencement of the PPM scheme. In that way, all potential bias will be corrected for. This is a major flaw for me in this manuscript\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? No source data required\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-35
|
https://f1000research.com/articles/8-32/v1
|
09 Jan 19
|
{
"type": "Research Article",
"title": "Effects of sub-lethal concentrations of copper ammonium acetate, pyrethrins and atrazine on the response of Escherichia coli to antibiotics",
"authors": [
"Hyunwoo Jun",
"Brigitta Kurenbach",
"Jack Aitken",
"Alibe Wasa",
"Mitja N.P. Remus-Emsermann",
"William Godsoe",
"Jack A. Heinemann",
"Hyunwoo Jun",
"Brigitta Kurenbach",
"Jack Aitken",
"Alibe Wasa",
"Mitja N.P. Remus-Emsermann",
"William Godsoe"
],
"abstract": "Background: Antibiotic resistance in human and animal pathogens is mainly the outcome of human use of antibiotics. However, bacteria are also exposed to thousands of other antimicrobial agents. Increasingly those exposures are being investigated as co-selective agents behind the rapid rise and spread of resistance in bacterial pathogens of people and our domesticated animals. Methods: We measured the sub-lethal effects on antibiotic tolerance of the human pathogen/commensal Escherichia coli caused by exposure to three common biocide formulations based on either copper, pyrethrins, or atrazine as active ingredients. The influence of the efflux pump AcrAB-TolC was investigated using deletion strains, and the persistence of observed effects was determined. Results: Some effects were seen for all biocides, but the largest effects were observed with copper in combination with the antibiotic tetracycline. The effect was caused by both the induction of the adaptive efflux system and by chelation of the antibiotic by copper. Finally, persistence of the adaptive response was measured and found to persist for about two generations. Conclusions: Through a combination of microbe-chemical and chemical-chemical interactions, humanity may be creating micro-environments in which resistance evolution is accelerated.",
"keywords": [
"biocides",
"antibiotic resistant bacteria",
"antibiotics",
"copper",
"pyrethrins",
"atrazine"
],
"content": "Introduction\n\nBesides antibiotics, a growing number of anthropogenic products are being found to affect antibiotic resistance in microorganisms (Heinemann & Kurenbach, 2017; Knöppel et al., 2017; Molina-González et al., 2014). These include non-antibiotic therapeutics (Kristiansen, 1992; Maier et al., 2018), food sweeteners (Wang et al., 2018), food perservatives (Capita & Alonso-Calleja, 2013; Capita et al., 2014), emulsifiers used in food and medicine (Kurenbach et al., 2017), paints, and cleaning products (Buffet-Bataillon et al., 2016; Molina-González et al., 2014).\n\nThe world’s industrial capacity to produce, distribute and consume manufactured chemical products is at an all time high and growing (American Chemistry Council, 2016). In the United States alone, 13,000 kg of industrial chemicals are produced per capita per year, and over 11,000 kg of 8,000 chemicals are produced or imported per capita per year (Wang et al., 2018).\n\nManufactured chemicals contribute to pollution, which is the leading cause of disease and premature death worldwide (Landrigan et al., 2018). The Lancet Commission on Pollution and Health said that less than half of “high-production volume chemicals have undergone any testing for safety or toxicity, and rigorous pre-market evaluation of new chemicals has become mandatory in only the past decade and in only a few high income countries. The result is that chemicals and biocides whose effects on human health and the environment were never examined have repeatedly been responsible for episodes of disease, death, and environmental degradation” (Landrigan et al., 2018).\n\nTo our knowledge, pre-market assessments of biocides that include tests of sub-lethal effects on microorganisms have not been performed yet (Kurenbach et al., 2015), although this may be changing, at least in Europe (Buffet-Bataillon et al., 2016). For every human exposure to a biocide, there may be 10s of trillions of exposures in our personal microbiota, not to mention microbiota exposures in soil, water and air, and on plants, livestock, companion animals and insects (Claus et al., 2016; Imfeld & Vuilleumier, 2012; Motta et al., 2018).\n\nWe have previously shown that active ingredients and commercial formulations based on dicamba, glyphosate, and 2,4-D induced changes in the response of Escherichia coli and Salmonella enterica to five different antibiotics from different classes. Increases in tolerance to antibiotics could be attributed in part to increased production of efflux pumps from the resistance-nodulation-division (RND) family (Kurenbach et al., 2015; Kurenbach et al., 2017). Unfortunately, the diversity of active and adjuvant ingredients of the tested herbicides provide little basis to produce general predictions of effects on different bacteria because of a common chemistry. Thus, at present, products must be tested on a case-by-case basis to determine whether or not there are sub-lethal responses in bacteria of interest.\n\nThe aim of the work described here was to determine whether other biocides used in agriculture and urban environments could induce a similar response in E. coli. The biocides used in the experiments were commercial formulations of a fungicide (copper ammonium acetate), an insecticide (pyrethrins) and an herbicide (atrazine).\n\nWe measured the initial response of bacteria to chemical exposures by the adaptive changes in the expression of TolC, an efflux pump component that controls transport across membranes (Corona & Martinez, 2013). This response is reversible in time, but may be heritable through epigenetic transmission (Bootsma et al., 2012; Motta et al., 2015). We used one biocide-antibiotic combination to attempt to empirically measure the transgenerational longevity of the adaptive response.\n\n\nMethods\n\nStrains used in this study are detailed in Table 1. Liquid cultures were grown in LB Lennox (Invitrogen, Auckland, NZ) at 37°C in a rotary incubator. Antibiotics used were tetracycline (Tet, Sigma, Auckland, NZ), streptomycin (Str, Sigma, Auckland, NZ), kanamycin (Kan, Gibco, Auckland, NZ), and ciprofloxacin (Cip, Pentex, Auckland, NZ). Biocides were commercial formulations Yates Liquid Copper Fungicide (Yates, Auckland, NZ) containing 92.8 g/L of Copper (Cu2+) in the form of copper ammonium acetate, Pyrethrum Natural Insect Spray (Yates, Auckland, NZ) containing pyrethrins (14 g/L) and 56.5 g/L of piperonyl butoxide, and Atranex WG (Adama, Nelson, NZ), containing 900g/kg atrazine. Relevant concentrations are given in the main text or Figure legends.\n\nAntibiotic responses were determined as described previously (Kurenbach et al., 2015). In brief, E. coli was grown to saturation (ca. 2 x 109 cfu/mL) in LB, and serial dilutions were plated on LB in the presence of antibiotics and/or biocides. When added, biocide concentrations were constant, while antibiotic concentrations varied. Plates were incubated at 37°C and examined daily for up to 10 days, at which point no new colonies emerged. To account for day to day variability, cfu counts were normalised to growth on nonselective medium. The efficiency of plating (EoP) is the ratio of a culture’s titre (cfu/mL) on treatment plates to the titre on LB [(cfu/mL)treatment/ (cfu/mL)LB] (Rosner, 1985). The detection range was an EoP of ca. 1 to 10-7.\n\nThe concentration of biocide that caused a significantly different response to an antibiotic (“dose response”) was determined as described previously (Kurenbach et al., 2015). In brief, E. coli were grown to saturation in LB and a serial dilution was plated on LB agar plates supplemented with varying concentrations of biocide and a constant concentration of antibiotic. The antibiotic concentration used was the one causing the greatest difference in EoP in the antibiotic response experiments. The inducing concentration of a biocide was defined as the lowest concentration for which a change occurred that was a) statistically significant and b) showed an at least 100-fold difference in EoP compared to the control containing only antibiotic. Plates were incubated at 37°C and examined daily for up to 10 days, at which point no new colonies emerged.\n\nTo construct plasmid pHJ01, E. coli BW25113 (GenBank accession number CP009273) was used as a template for the 204 bp upstream of the start codon of tolC. The tolC promoter was amplified by PCR and fused to mScarlet-I which was amplified from pTriEx-RhoA-wt_mScarlet-I_SGFP2 (Addgene plasmid #85071) and HindIII digested pFru97 (Tecon & Leveau, 2012) by isothermal assembly (Gibson et al., 2009; Schlechter et al., 2018). Touchdown PCRs were performed as described previously (Schlechter et al., 2018) using Phusion High-Fidelity DNA polymerase (Thermo Scientific, Auckland, NZ). Primers used were FWD_TolC (5' CAG GAC GCC CGC CAT AAA CTG CCA GGA ATT GGG GAT CGG ATG TTA ATG TCC TGG CAC TAA TAG TGA ATT AAA TGT 3’; Tm: 60°C), REV_TolC (5' TCG CCC TTG CTC ACC ATG GTT TGC ATT CCT TGT GGT GAA GCA G 3'; Tm: 60°C), TolC_mScarlet_FWD (5' CTT CAC CAC AAG GAA TGC AAA CCA TGG TGA GCA AGG GC 3’; Tm: 70°C), and mScarlet_REV (5' TTA CTG GAT CTA TCA ACA GGA GTC CAA GCT CAG CTA ATT ACT TGT ACA GCT CGT CCA TGC 3'; Tm: 71°C), where nucleotides shown in bold font are complementarity to the vector, and nucleotides shown in italics overlap with other primers. pHJ01 transformands of BW25113 were selected on kanamycin.\n\nPrior to microscopy, cells grown for 180 min either in LB or in LB + 450 µg/mL copper were fixed using 4% paraformaldehyde as described previously and stored at -20°C in 1:1 ethanol:phosphate buffered saline (Akkermans et al., 1996; Kowalchuk et al., 2004). Fixed cells were examined with an Axio Imager. M1 (Zeiss, Oberkochen, Germany) using an EC Plan-Neofluar 100x objective (NA 1.30) and Zeiss filter set 43HE (BP 550/25 (HE); FT 570 (HE); BP 605/70 (HE)). Multichannel images were acquired using an AxioCam 506 mono camera (Zeiss) in differential interference contrast (DIC) and Zeiss filter set 43HE. Single cell fluorescence was determined as described previously (Remus-Emsermann et al., 2016).\n\nSeven Erlenmeyer flasks (50 mL) containing LB (10 mL) were supplemented with copper (450 µg/mL) and tetracycline (35 μg/mL) (Flasks 1–4), tetracycline (35 μg/mL) without copper (Flasks 5 and 6), or copper (450 µg/mL) without tetracycline (Flask 7) at t0. All flasks were incubated continuously at 37°C with aeration. E. coli BW25113 was grown to saturation without selection and approximately 104 cfu were used to inoculate flasks 1, 6 and 7 at t0, and flasks 2 and 3 at t24 and t48, respectively. Flasks 4 and 5 were inoculated at t96. The culture in each flask was monitored for growth every 24 hours by plating appropriate dilutions onto LB agar plates.\n\nE. coli was grown to saturation with aeration at 37°C in liquid LB medium supplemented with both copper (450 µg/mL) and tetracycline (15 µg/mL) for 3 days. This culture was diluted 100-fold into 10 mL LB medium supplemented with only tetracycline (15 µg/mL) and incubated at 37°C for 12 hours with aeration. The concentration of E. coli at the start and end of the experiment was determined using a haemocytometer.\n\nR (version 3.2.0) was used for all statistical analyses (R Core Team, 2013). In experiments testing the responses to antibiotics during exposure to biocides we were interested in effects on EoP that were different in antibiotic+biocide combinations compared to either substance in isolation. We therefore tested the log-transformed EoP scores using a multifactor analysis of variance (ANOVA) by evaluating the significance of the antibiotic by biocide interaction term. Antibiotic concentrations were treated as separate categories in the ANOVA. Plots of residuals were used to test for violations of assumptions. We fit these models using the lm function.\n\nSince many data points used for the determination of the concentration of biocide that caused a significantly different response to an antibiotic were near or below the detection limit, residuals from a standard ANOVA were not normally distributed. We therefore used the equivalent non-parametric Kruskal-Wallis one-way ANOVA to test for differences in log-transformed EoP/EoP(0) scores among biocide concentrations. The P-value reported is derived for a null model where EoP/EoP(0) is the same across all biocide concentrations versus and alternative model where the ratio differs among some concentrations.\n\nWe tested whether cfu scores depended on “flask” using a single factor ANOVA at 24 and 48 hours post inoculation. In each case, we first used an analysis of covariance (ANCOVA) to test if cfu scores post inoculation were confounded with the cfu count and the time of inoculation. Cfu scores at inoculation did not influence final scores (data not shown). Cfu scores were transformed to log (cfu +0.0001) to ensure normality of the residuals. We used a sequential Bonferroni contrast to test for differences among treatments (Flasks 1–4) and between treatments and controls (flasks 1 and 6, 1 and 6, and 4 and 5). Residuals were used to check assumptions. With the exception of two low-influence outliers, the data matched our expectations under normality.\n\nWhere fluorescence was measured, differences between median fluorescence values of the reporter strain grown under two conditions (+/- copper) were determined using a non-parametric Mann-Whitney U T-test because residuals were not normally distributed. Violin plots were created using ggplot2 (Wickham, 2016).\n\n\nResults\n\nMIC was defined as the minimum concentration of agent in an agar plate at which no growth was observed after ca. 108 cfu were applied to the surface. It was not possible to determine the MIC for atrazine because E. coli BW25113 survived to the limit of solubility of atrazine in our standard culture medium, LB. The No Observable Effect Level (NOEL) was defined as the highest concentration of a substance that had no effect on the EoP (Table 2).\n\nBacteria were cultured on LB agar supplemented with one of the three commercial formulations of biocide (at respective NOEL concentrations) as well as different concentrations of selected antibiotics. Changes in response to particular concentrations of antibiotic because of exposure to the biocide are revealed as a differential EoP (Figure 1). As reported for other biocide*antibiotic combinations, the observed responses were specific for the combination of biocide and antibiotic used (Kurenbach et al., 2015; Kurenbach et al., 2017). We observed increases and decreases in tolerance to different antibiotics as well as no effect in some cases. As a conservative threshold, we used the antibiotic concentration for which we saw an at least 103-fold decrease in EoP as the cut-off point to determine the fold-change in survival (Table 3).\n\nThe x-axes scale is antibiotic concentrations in µg/mL. Biocide concentrations used were 450 µg/mL for copper ammonium acetate, 140 µg/mL for pyrethrin, and 1000 µg/mL for atrazine. Values are means of at least three independent experiments; error bars are standard errors (SEM, with SEM=standard deviation/√n). Asterisks indicate P-values for antibiotic*herbicide interaction terms (see Materials and Methods). * P<0.05; ** P<0.01; *** P<0.001; ns, not significant.\n\naWhile EoP dropped below our threshold at the same concentration in the presence and absence of biocides, the ANOVA showed a statistically significant interaction term for this combination.\n\nbThe ANOVA did not show a statistically significant interaction term, despite the drop in EoP below our threshold at different concentrations.\n\nCopper significantly increased the EoP over a 40-fold concentration range of tetracycline (from 2 to 80 µg/mL) and decreased it on a 5-fold concentration range of streptomycin (from 5 to 1 µg/mL). Copper caused non-statistically significant decreases in the EoP on either ciprofloxacin or kanamycin.\n\nPyrethrins increased EoP over a 5-fold streptomycin concentration range. They caused a statistically significant difference in EoP on ciprofloxacin, but no change in MIC. This has been sometimes observed for various herbicide-antibiotic combinations (Kurenbach et al., 2015; Kurenbach et al., 2018).\n\nAtrazine caused statistically significant but small increases in EoP on ciprofloxacin, kanamycin, and streptomycin.\n\nIn the experiments described above, the concentration of antibiotic was varied while the biocide concentration was constant. To determine the minimum biocide concentration necessary to cause the observed effects, we chose an antibiotic concentration for which there was a maximum resolution between treatments and decreased the biocide concentration for each biocide.\n\nAs a conservative measure, we report the biocide concentration that caused a statistically significant and at least a 100-fold change in the EoP compared to the EoP of the antibiotic-only plate (EoP(0)). To aid visualization, we calculated log EoP/EoP(0) (Figure 2). A value >0 indicates that the biocide increases EoP of bacteria on higher concentrations of the antibiotic. Our threshold of a 100-fold change was reached at 120 µg/mL copper with tetracycline, 250 µg/mL copper with streptomycin, and 100 µg/mL pyrethrins with streptomycin.\n\nAntibiotic concentrations used (in µg/mL) were as follows: Tet: 10 µg/mL for copper; Str: 2 µg/mL for copper and 10 µg/mL for pyrethrins. Values are means of at least 3 independent experiments; error bars are standard errors (SEM, with SEM=standard deviation/√n). Asterisks indicate the lowest biocide concentration for which a statistically significant change in EoP by at least 100-fold compared to the antibiotic only occurred. * P<0.05; ** P<0.01; *** P<0.001; ns, not significant.\n\nThe minimum biocide concentration was not determined for some other statistically significant combinations shown in Figure 1. This was because the affected antibiotic concentrations were of such a small range. As a consequence, we also concentrated on copper exposures in the remainder of the study.\n\nWe previously found that the herbicidal formulations based on 2,4-D, dicamba and glyphosate (Kurenbach et al., 2015), as well as the corresponding purified active ingredients (Kurenbach et al., 2017) caused changes in the expression pattern of genes that may alter antibiotic susceptibility. This response is phenotypic, resulting from a change in gene expression rather than genotype. It is distinguished from the outgrowth of rare spontaneous mutants by the uniform reversion on the population level when the environment changes (Motta et al., 2015).\n\nWe further characterized the copper-induced tetracycline response as an adaptive response by following the phenotype of induced clones. Randomly chosen colonies from cultures plated on LB, LB+Tet, or LB+Tet+Cu were transferred to plates containing 35 μg/mL tetracycline or LB. At this tetracycline concentration, E. coli survived only when simultaneously exposed to copper.\n\nRegardless of whether the colonies were transferred from LB or LB+Tet+Cu, they all again formed colonies on LB. However, none of the colonies transferred from either LB or LB+Tet+Cu grew on LB+Tet plates, indicating that the response to tetracycline was reversible and dependent upon ongoing stimulation by copper.\n\nThe efflux pump AcrAB-TolC was shown to contribute to the altered EoP of E. coli on different antibiotics when simultaneously exposed to various herbicides (Kurenbach et al., 2017). Here, the same set of strains from an isogenic series carrying single gene deletions, ΔacrA, ΔacrB and ΔtolC, were used to test whether copper induced an adaptive response via this pump. NOEL and MIC of copper were determined for all three strains (Table 2). Changes in EoP on tetracycline-supplemented media were measured as described above, using the NOEL copper concentration (Figure 3).\n\nThe x-axis indicates antibiotic concentration in µg/mL. Copper was added at 450 µg/mL. Error bars are standard errors (SEM, with SEM=standard deviation/√n). Asterisks indicate P values for interaction terms (see Materials and Methods). * P<0.05; ** P<0.01; *** P<0.001; ns, not significant.\n\nThe MIC but not the NOEL of tetracycline was lower in all three deletion strains compared to the wildtype BW25113. This is consistent with the observations of others (de Cristóbal et al., 2006) and suggests that the AcrAB-TolC efflux pump is responding to copper and contributing to tetracycline resistance. Concurrent copper exposure significantly increased tolerance to tetracycline in all strains. However, with increases of 4-fold for ΔacrA, 22.5-fold for ΔacrB, and 5-fold for ΔtolC these effects were smaller than those observed for the parental strain (40-fold). This suggests that the AcrAB-TolC efflux pump is responding to copper and contributing to tetracycline resistance, but it is not the only mechanism involved.\n\nAccumulation of copper directly effects the transcription factor MarR, derepressing the MarRAB operon (Hao et al., 2014). Increased production of the transcription factor MarA leads to increased transcription of among others the acrAB and tolC genes (Weston et al., 2018). We chose to investigate this further by using a tolC reporter strain.\n\nThe E. coli strain BW25113 (pHJ01) has the mScarlet fluorescent protein open reading frame transcriptionally fused to the tolC promoter region. This technique was used previously to demonstrate e.g. the accessibility of fructose to bacterial cells on leaves and the availability of phenol to bacteria on leaves (Leveau & Lindow, 2001; Sandhu et al., 2007).\n\nThe fluorescence of BW25113 (pHJ01) was statistically significantly lower (p < 0.001) when cultured in LB compared to LB + 450 µg/mL copper (Figure 4). The median relative fluorescence of the reporter strain increased 77% from 70 arbitrary fluorescence units (afu) when cultured in LB to 124 afu when cultured with additional copper. This indicates induction of tolC by the copper fungicide.\n\nThe violin plots show the distribution of the single-cell fluorescence within the cell population. The median is depicted by the bar in the box; the box represents the 25% and 75% quartiles.\n\nCopper had a large effect on the EoP of E. coli exposed to tetracycline, increasing the concentration necessary to decrease EoP by >103-fold from 2 to 80 µg/mL tetracycline. This was the largest effect of any biocide on any antibiotic that we have observed. When cultured in a combination of copper and tetracycline at copper-induced sub-lethal concentrations of tetracycline, we observed a significant delay in the growth of the culture. This could be due to copper chelation of tetracycline (Tong et al., 2015), or to the outgrowth of rare tetracycline resistant mutants. Since we have not detected the latter (see above), we tested the former hypothesis.\n\nA series of E. coli BW25113 cultures were used to estimate tetracycline bioavailability. The series was composed of four flasks with a medium supplemented with copper and tetracycline and incubated at 37°C with aeration. The medium in the flasks was inoculated with bacteria in successive 24 hour intervals (flask 1 at t0 – flask 4 at t72) and the titre of each culture was determined at the same intervals by plating dilutions of samples on LB. Control cultures with medium supplemented with only tetracycline were started in parallel with flasks 1 and 4, and a positive control culture using medium only supplemented with copper was inoculated in parallel to flask 1. These controls showed that tetracycline alone, even after 92 hours of pre-incubation, prevented growth of the culture, and that the copper concentration was sub-lethal. Cultures began to grow only after the medium with a mixture of copper and tetracycline was over 72 hours old (Figure 5).\n\nX-axis is time in hrs after adding copper and tetracycline (flasks 1–4, solid lines), tetracycline (flasks 5 & 6, dashed lines), or copper (flask 7, dotted line). The first data point in each series indicates time of inoculation with E. coli BW25113. Values are means of three independent experiments ± SD.\n\nUsing ANOVAs, we tested for significant differences between flasks in cfu counts 24 and 48 hours after inoculation with bacteria (see Underlying data: ‘Chelation experiment_ANOVA tables’; https://doi.org/10.17605/OSF.IO/RZKWU (Kurenbach, 2018)). At 24 hrs, the cfus of flask 4, inoculated at t72, was significantly different to either flasks 1, 2, or 3. At 24 hrs, these flasks had not passed the t72 point. At 48 hrs, flasks 3 and 4, now both past t72, were not significantly different from each other. Flask 4 was still different from flasks 1 and 2, while differences between 3 and 1 and 2 were not significant or marginally significant, respectively. This is in general agreement with the interpretation that bacteria start growing after t72 regardless of the point in time at which they were inoculated (see Underlying data: ‘Chelation experiment_ANOVA tables’; https://doi.org/10.17605/OSF.IO/RZKWU (Kurenbach, 2018)).\n\nThis observation is consistent with the notion that copper forms a complex with tetracycline (Tong et al., 2015), or facilitates degradation of tetracycline, over time. Because tetracycline is bacteriostatic, the bacteria are able to recover once the effective concentration of tetracycline falls to sub-inhibitory levels.\n\nE. coli’s response to copper exposure was consistent with an adaptive response through a change in efflux pump levels rather than a change in DNA sequence conferring antibiotic resistance. We therefore hypothesized that the tetracycline-resistant physiotypes created by the adaptive response should continue to reproduce in medium supplemented with tetracycline above the MIC until the number of efflux pumps, and possibly other contributing factors, per cell fell below an efficacious threshold (Bootsma et al., 2012; Motta et al., 2015).\n\nThis hypothesis could be tested by determining the number of generations E. coli was able to reproduce after removal of copper but not tetracycline from a previously induced (Cu+Tet) culture. A complication was encountered when we observed a reversible filamentation of the bacteria after their transfer from a medium with both copper and tetracycline. E. coli are known to form filaments when stressed (Justice et al., 2006). Filamentation made the determination of growth by measuring OD600 inaccurate. To address this, densities of bacteria were determined visually using a haemocytometer.\n\nImmediately after transfer to tetracycline-supplemented medium, the concentration of bacteria was determined. The cultures were then incubated at 37°C for 16 hours. Testing the limits of our method, we were consistently able to distinguish 4-fold differences in population growth, i.e. two generations. Our experimental data fell below that threshold, with populations growing by ca. 3-fold, or just over one generation. We therefore estimate that the adaptive phenotype in this experiment was heritable for less than two generations.\n\n\nDiscussion\n\nAbout 2 million metric tons of the 30 most commonly used commercial pesticides are released into the environment annually worldwide. Of these, 55.4% are herbicides, 28.6% are fungicides and 5.7% are insecticides (Casida & Bryant, 2017). Despite their long and widespread use, to our knowledge they have never been tested for sub-lethal effects on potential human or animal pathogenic bacteria.\n\nWe have tested three common pesticides for sub-lethal effects on the bacterium E. coli. Copper-based formulations are the third largest fungicide usage group. The triazine herbicide ingredient atrazine is by amount used the third most commonly used herbicide in the world. Pyrethroids are medium use insecticides, occupying positions of 11, 12, 16 and 26 of the top 30 insecticides used worldwide (Casida & Bryant, 2017).\n\nSimilar to our previous findings for the herbicides based on glyphosate, dicamba, and 2,4-d active ingredients, the three biocides tested here did alter the response of the human and animal commensal and potential pathogen E. coli to some clinical antibiotics. The concentrations of biocide that caused the change in response to antibiotics were at or below label-recommended application rates, which are 30 - 2320 µg/mL for copper, 70 µg/mL for pyrethrins, and 500 µg/mL for atrazine.\n\nStreptomycin resistance was most affected by pyrethroids, while effects on other antibiotics tested were small. Moreover, results were not statistically significant for the other tested aminoglycoside antibiotic, kanamycin. Likewise, atrazine caused only small effects for all antibiotics tested. We observed the largest changes using copper, which increased survival on 40 times higher concentrations of tetracycline.\n\nThe response seen to tetracycline from copper exposure was the largest we have observed from a biocide and antibiotic combination. Some of this is attributed to the chelation of copper by tetracycline, resulting in a decrease in effective concentration of both agents. Nevertheless, some of the response was confirmed to be adaptive because as shown by use of the strains with gene deletions, it depended in part on acrA, acrB and tolC. Moreover, exposure to copper was directly observed to increase the expression of the red fluorescent protein gene mScarlet under the control of the tolC promoter, and the fully susceptible phenotype was uniformly restored to the population when induced bacteria were transferred to LB+Tet medium. Because the gene deletion strains continued to respond to copper and tetracycline, the full effect of copper was not explained only through the expression of the AcrAB-TolC efflux pump.\n\nCopper is a common supplement for animal feeds which can also contain traces of copper from biocide residues. In a European Union survey of copper content in animal feed used in member countries, copper was found over a broad range of concentrations (in the mg/kg range) and mean concentrations of 8 to ~20 mg/kg in the feed of most surveyed animals, including pets such as dogs and cats. The highest mean was 119 mg/kg for piglets (EFSA, 2016). The lowest statistically significant tetracycline-resistance inducing concentration of copper in our study was 120 mg/L, just above routine piglet exposures. Other exposures to copper from use of biocides would be in addition to these.\n\nAnimal feed can also be unintentionally contaminated with antibiotics. Tetracycline-class antibiotics are approved for use in animal feed and are among the most frequently used. This alone resulted in concentrations of chlortetracycline and doxycycline at concentrations of 10 mg/kg and 4 mg/kg, respectively, in the feces of pigs. The level was high enough to select for resistance (Gavilán et al., 2015; Granados-Chinchilla & Rodríguez, 2017).\n\nA study in Vietnam that examined nearly 1500 chicken and pig feed formulations estimated that 77.4 mg and 286.7 mg, respectively, of antimicrobials were used to raise each 1 kg of animal. The level of antimicrobial agent in the feed ranged from 25.7–62.3 mg/kg. Chlortetracycline was among the most common additives in chicken and pig feed (Van Cuong et al., 2016). Thus it is not unusual to find both copper and tetracycline in the same environments.\n\n\nConclusion\n\nPreservation of antibiotics as useful medicines requires stewardship of populations of bacteria that remain susceptible to them. It is imprudent to base stewardship on frequency of resistance because even low numbers of resistant bacteria will dominate a population when antibiotics are used. Environments that maintain phenotypes caused by adaptive resistance or genotypes with a fitness advantage during antibiotic exposure thus could contribute to the rate at which populations of pathogens evolve resistance (Kurenbach et al., 2018).\n\nThe number of circulating high use commercial chemicals being associated with sub-lethal effects on bacteria is growing, as are the number of environments that are being contaminated with antibiotics themselves. Exposure to herbicides and antibiotics simultaneously accelerates the evolution of genotypically resistant bacteria (Kurenbach et al., 2018). The effects seen for atrazine, copper and pyrethrins were more limited than for some other herbicide active ingredients and commercial formulations, but may contribute to the overall burden of resistance.\n\n\nData availability\n\nAll underlying data is available on the Open Science Framework: Effects of sub-lethal concentrations of copper ammonium acetate, pyrethrins and atrazine on the response of Escherichia coli to antibiotics, https://doi.org/10.17605/OSF.IO/RZKWU (Kurenbach, 2018).\n\nThe following files are available:\n\nEffects of biocides on antibiotic response.Antibiotic resistance in the presence and absence of biocide. Data presented in Figure 1.\n\n○ Atrazine+Cip_killing curves.csv\n\n○ Atrazine+Kan_killing curves.csv\n\n○ Atrazine+Str_killing curve.csv\n\n○ Atrazine+Tet_killing curve.csv\n\n○ Cu+Cip_killing curves.csv\n\n○ Cu+Kan_killing curves.csv\n\n○ Cu+Str_killing curves.csv\n\n○ Cu+Tet_killing curves.csv\n\n○ Pyrethrins+Cip_killing curves.csv\n\n○ Pyrethrins+Kan_killing curves.csv\n\n○ Pyrethrins+Str_killing curves.csv\n\n○ Pyrethrins+Tet_killing curves.csv\n\nMinimum inducing concentration. Data presented in Figure 2.\n\n○ Cu+Str_Minimum inducing concentration.csv\n\n○ Cu+Tet_Minimum inducing concentration.csv\n\n○ Pyrethrins+Str_Minimum inducing concentration.csv\n\nDependence on the AcrAB-TolC efflux pump as evidence of an adaptive response. Antibiotic resistance response in the presence and absence of copper. Data presented in Figure 3.\n\n○ AcrA_Cu+Tet_killig curves.csv\n\n○ AcrB_Cu+Tet_killig curves.csv\n\n○ TolC_Cu+Tet_killig curves.csv\n\ntolC was induced by copper. Relative fluorescence data for BW21003(pHJ101) in the absence and presence of copper. Data presented in Figure 4.\n\nFluorescense_PtolC induction.csv\n\nCopper directly reduced available tetracycline. Data presented in Figure 5.\n\n○ Chelation_all timepoints.csv\n\n○ Chelation experiment_ANOVA tables.docx\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\nFunding was received from a variety of sources none of whom played any role in the study, preparation of the article, or decision to publish. This project received funding from the Brian Mason Trust (Grant # 2015/08 to JAH) and donations to the UC Foundation (JAH) including from, inter alia, donors Third World Network (Malaysia) and the Sustainable Food Trust (UK).\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\nWe are grateful to Stuart Levy for the gift of Keio strains and Dorus Gadella for the gift of pTriEx-RhoA-wt_mScarlet-I_SGFP2. We also acknowledge Lynn Clark for support with the violin plots.\n\n\nReferences\n\nAkkermans ADL, van Elsas JD, de Bruija FJ: Molecular Microbial Ecology Manual. (Kluwer Academic Publishers: Dortrecht, The Netherlands) 1996. Publisher Full Text\n\nAmerican Chemistry Council: US Chemical industry continues to outpace industrial output; Accounts for more than one-half of construction spending by manufacturing sector. 2016. Reference Source\n\nBaba T, Ara T, Hasegawa M, et al.: Construction of Escherichia coli K-12 in-frame, single-gene knockout mutants: the Keio collection. Mol Syst Biol. 2006; 2(1): 2006.0008. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBootsma MC, van der Horst MA, Guryeva T, et al.: Modeling non-inherited antibiotic resistance. Bull Math Biol. 2012; 74(8): 1691–1705. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBuffet-Bataillon S, Tattevin P, Maillard JY, et al.: Efflux pump induction by quaternary ammonium compounds and fluoroquinolone resistance in bacteria. Future Microbiol. 2016; 11(1): 81–92. PubMed Abstract | Publisher Full Text\n\nCapita R, Alonso-Calleja C: Antibiotic-resistant bacteria: a challenge for the food industry. Crit Rev Food Sci Nutr. 2013; 53(1): 11–48. PubMed Abstract | Publisher Full Text\n\nCapita R, Riesco-Peláez F, Alonso-Hernando A, et al.: Exposure of Escherichia coli ATCC 12806 to sublethal concentrations of food-grade biocides influences its ability to form biofilm, resistance to antimicrobials, and ultrastructure. Appl Environ Microbiol. 2014; 80(4): 1268–80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCasida JE, Bryant RJ: The ABCs of pesticide toxicology: amounts, biology, and chemistry. Toxicol Res (Camb). 2017; 6(6): 755–763. PubMed Abstract | Publisher Full Text | Free Full Text\n\nClaus SP, Guillou H, Ellero-Simatos S: The gut microbiota: a major player in the toxicity of environmental pollutants? NPJ Biofilms Microbiomes. 2016; 2: 16003. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCorona F, Martinez JL: Phenotypic Resistance to Antibiotics. Antibiotics (Basel). 2013; 2(2): 237–255. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Cristóbal RE, Vincent PA, Salomón RA: Multidrug resistance pump AcrAB-TolC is required for high-level, Tet(A)-mediated tetracycline resistance in Escherichia coli. J Antimicrob Chemother. 2006; 58(1): 31–6. PubMed Abstract | Publisher Full Text\n\nEFSA: Revision of the currently authorised maximum copper content in complete feed. EFSA J. 2016; 14: 4563. Publisher Full Text\n\nGavilán RE, Nebot C, Miranda JM, et al.: Analysis of Tetracyclines in Medicated Feed for Food Animal Production by HPLC-MS/MS. Antibiotics (Basel). 2015; 5(1): pii: E1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGibson DG, Young L, Chuang RY, et al.: Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods. 2009; 6(5): 343–5. PubMed Abstract | Publisher Full Text\n\nGranados-Chinchilla F, Rodríguez C: Tetracyclines in food and feedingstuffs: from regulation to analytical methods, bacterial resistance, and environmental and health implications. J Anal Methods Chem. 2017; 2017: 1315497. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHao Z, Lou H, Zhu R, et al.: The multiple antibiotic resistance regulator MarR is a copper sensor in Escherichia coli. Nat Chem Biol. 2014; 10(1): 21–8. PubMed Abstract | Publisher Full Text\n\nHeinemann JA, Kurenbach B: Post-Antibiotic Era is Business as Usual. Alliance for the prudent use of antibiotics (APUA) Newsletter. 2017; 35(1): 7–9.\n\nImfeld G, Vuilleumier S: Measuring the effects of pesticides on bacterial communities in soil: A critical review. Eur J Soil Biol. 2012; 49: 22–30. Publisher Full Text\n\nJustice SS, Hunstad DA, Seed PC, et al.: Filamentation by Escherichia coli subverts innate defenses during urinary tract infection. Proc Natl Acad Sci U S A. 2006; 103(52): 19884–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKnöppel A, Näsvall J, Andersson DI: Evolution of Antibiotic Resistance without Antibiotic Exposure. Antimicrob Agents Chemother. 2017; 61(11): pii: e01495-17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKowalchuk GA, de Bruijn FJ, Head IM, et al. eds.: Molecular Microbial Ecology Manual. Dordrecht: Springer Netherlands. 2004. Reference Source\n\nKristiansen JE: The antimicrobial activity of non-antibiotics. Report from a congress on the antimicrobial effect of drugs other than antibiotics on bacteria, viruses, protozoa, and other organisms. APMIS Suppl. 1992; 30: 7–14. PubMed Abstract\n\nKurenbach B: “Effects of Sub-Lethal Concentrations of Copper Ammonium Acetate, Pyrethrins and Atrazine on the Response of Escherichia Coli to Antibiotics.” OSF. 2018. http://www.doi.org/10.17605/OSF.IO/RZKWU\n\nKurenbach B, Gibson PS, Hill AM, et al.: Herbicide ingredients change Salmonella enterica sv. Typhimurium and Escherichia coli antibiotic responses. Microbiology. 2017; 163(12): 1791–1801. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKurenbach B, Hill AM, Godsoe W, et al.: Agrichemicals and antibiotics in combination increase antibiotic resistance evolution. PeerJ. 2018; 6: e5801. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKurenbach B, Marjoshi D, Amábile-Cuevas CF, et al.: Sublethal exposure to commercial formulations of the herbicides dicamba, 2,4-dichlorophenoxyacetic acid, and glyphosate cause changes in antibiotic susceptibility in Escherichia coli and Salmonella enterica serovar Typhimurium. mBio. 2015; 6(2): pii: e00009-15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLandrigan PJ, Fuller R, Acosta NJR, et al.: The Lancet Commission on pollution and health. Lancet. 2018; 391(10119): 462–512. PubMed Abstract | Publisher Full Text\n\nLeveau JH, Lindow SE: Appetite of an epiphyte: quantitative monitoring of bacterial sugar consumption in the phyllosphere. Proc Natl Acad Sci U S A. 2001; 98(6): 3446–3453. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMaier L, Pruteanu M, Kuhn M, et al.: Extensive impact of non-antibiotic drugs on human gut bacteria. Nature. 2018; 555(7698): 623–628. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMolina-González D, Alonso-Calleja C, Alonso-Hernando A, et al.: Effect of sub-lethal concentrations of biocides on the susceptibility to antibiotics of multi-drug resistant Salmonella enterica strains. Food Control. 2014; 40: 329–334. Publisher Full Text\n\nMotta EVS, Raymann K, Moran NA: Glyphosate perturbs the gut microbiota of honey bees. Proc Natl Acad Sci U S A. 2018; 115(41): 10305–10310. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMotta SS, Cluzel P, Aldana M: Adaptive resistance in bacteria requires epigenetic inheritance, genetic noise, and cost of efflux pumps. PLoS One. 2015; 10(3): e0118464. PubMed Abstract | Publisher Full Text | Free Full Text\n\nR Core Team: R: a language and environment for statistical computing. 2013. Reference Source\n\nRemus-Emsermann MN, Gisler P, Drissner D: MiniTn7-transposon delivery vectors for inducible or constitutive fluorescent protein expression in Enterobacteriaceae. FEMS Microbiol Lett. 2016; 363(16): pii: fnw178. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRosner JL: Nonheritable resistance to chloramphenicol and other antibiotics induced by salicylates and other chemotactic repellents in Escherichia coli K-12. Proc Natl Acad Sci U S A. 1985; 82(24): 8771–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRuiz C, Levy SB: Regulation of acrAB expression by cellular metabolites in Escherichia coli. J Antimicrob Chemother. 2014; 69(2): 390–399. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSandhu A, Halverson LJ, Beattie GA: Bacterial degradation of airborne phenol in the phyllosphere. Environ Microbiol. 2007; 9(2): 383–392. PubMed Abstract | Publisher Full Text\n\nSchlechter RO, Jun H, Bernach M, et al.: Chromatic Bacteria - A Broad Host-Range Plasmid and Chromosomal Insertion Toolbox for Fluorescent Protein Expression in Bacteria. Front Microbiol. 2018; 9: 3052. Publisher Full Text\n\nTecon R, Leveau JH: The mechanics of bacterial cluster formation on plant leaf surfaces as revealed by bioreporter technology. Environ Microbiol. 2012; 14(5): 1325–1332. PubMed Abstract | Publisher Full Text\n\nTong F, Zhao Y, Gu X, et al.: Joint toxicity of tetracycline with copper(II) and cadmium(II) to Vibrio fischeri: effect of complexation reaction. Ecotoxicology. 2015; 24(2): 346–355. PubMed Abstract | Publisher Full Text\n\nVan Cuong N, Nhung NT, Nghia NH, et al.: Antimicrobial consumption in medicated feeds in vietnamese pig and poultry production. Ecohealth. 2016; 13(3): 490–498. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang QP, Browman D, Herzog H, et al.: Non-nutritive sweeteners possess a bacteriostatic effect and alter gut microbiota in mice. PLoS One. 2018; 13(7): e0199080. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang A, Gerona RR, Schwartz JM, et al.: A Suspect Screening Method for Characterizing Multiple Chemical Exposures among a Demographically Diverse Population of Pregnant Women in San Francisco. Environ Health Perspect. 2018; 126(7): 077009. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWeston N, Sharma P, Ricci V, et al.: Regulation of the AcrAB-TolC efflux pump in Enterobacteriaceae. Res Microbiol. 2018; 169(7–8): 425–431. PubMed Abstract | Publisher Full Text\n\nWickham H: \"ggplot2: Elegant graphics for data Analysis.\" In. New York, Springer Verlag. 2016. Publisher Full Text"
}
|
[
{
"id": "42816",
"date": "28 Jan 2019",
"name": "Xue-Xian Zhang",
"expertise": [
"Reviewer Expertise Environmental Microbiology",
"Bacterial Genetics"
],
"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 the effects of three commonly used agrochemicals (copper ammonium acetate, pyrethrins and atrazine) on bacterial resistance to antibiotics when they are present at sub-lethal concentrations. The data revealed significant antagonistic interactions between copper and the tetracycline antibiotic. It is particularly interesting that copper is capable of restoring the growth of E. coli cells when tetracycline was added at high concentrations that would normally cause full growth inhibition. Furthermore, the authors performed a series of hypothesis-driven experiments, and the results propose two underlying mechanisms: copper-induced expression of the AcrAB-TolC multidrug efflux system, and the chelation of tetracycline by copper. These results are highly significant, and I have no major reservations towards the experimental design and data interpretation. The few minor comments listed below are mostly on writings.\nFigure 1: I am wondering if bacteria were inoculated immediately after the addition of copper (and antibiotics), and whether there was a specific order for the addition of copper and antibiotics. I think this information is important and it should be provided in the Methods. This is because the toxic effects of copper are exerted by free Cu ions, not directly by copper ammonium acetate. There will be a process of copper ion releasing, and subsequent binding with other organic compounds present in the LB medium. We would thus expect different results when bacteria were inoculated immediately or left for some time after copper addition. Figure 2: ppm is used here, but the results are discussed in the main text using ug/ml. It would be helpful to use the same concentration unit. Also, it is difficult to distinguish the two types of grey bars, so the authors should consider changing one of them into an empty bar. Table 2: There was a two-fold difference of copper MICs between the acrA/acrB and tolC mutants, which has been overlooked in both the Results and Discussion. The AcrAB-TolC system is supposed to be induced by copper, and confers resistance to tetracycline (but not copper). Thus, it needs an explanation why inactivation of acrAB-tolC caused a reduction of MIC to copper, and furthermore, why there was a difference among the three mutants. For data presented in Figure 3 on copper plus tetracycline, the difference among three mutants has been noted but not discussed. A triple mutant of acrA, acrB and tolC may help further understand the mechanisms. Figure 5: I like this experiment and would be interested in what the results will be if the authors have set up an additional control with the addition of copper only (tetracycline added at the time of inoculation). As mentioned above, copper ions released from copper ammonium acetate can potentially react with organic compounds in the medium. Thus, copper-tetracycline chelation is just one of the plausible explanations for the obtained data. It will certainly help improve our understanding if the authors had determined the dynamic change of copper ion concentrations in this experiment. The current “discussion” is a little bit weak, given that many interesting points need to be specifically addressed in this work, e.g., the roles of the AcrAB-TolC system and the genotypic versus phenotypic adaptation.\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": []
},
{
"id": "42817",
"date": "01 Feb 2019",
"name": "Pal J. Johnsen",
"expertise": [
"Reviewer Expertise Microbiology",
"antimicrobial resistance 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 manuscript addresses important questions on potential non-antibiotic drivers of antimicrobial resistance evolution. Jun et al test the co-selective abilities of widely used pesticides and provide data suggesting that in particular copper ammonium acetate may co-select for pre-existing tetracycline resistance in E. coli.\n\nMajor points: 1. Relevance: The manuscript would benefit from a stronger contextualisation with respect to how relevant the co-selective effects reported are. The minimum copper biocide concentration causing a change in tetracycline response is reported to be 120microgramml- this is according to published data referred to in the Discussion in the upper range of what is measured in piglets- questioning the relevance of the findings. There is a huge literature on these issues - for example a study from Europe suggest that copper concentrations can be relatively high in soils, see for example (Heijerick DG1, Van Sprang PA, Van Hyfte AD.,20061).\n\nTo this end, the experiments on both the effects on acrAB-TolC knock-outs and mScarlet tolC promoter- fusions are performed at a much higher concentration (450microgramml) - this needs justification beyond the stated NOEL.\n\nThe dose response data presented in Figs 1 and 2 for the Copper/tetracycline combinations strongly suggest that there are selective effects at much lower concentrations than the authors conservatively report. I fully support this approach considering the chosen methodology. However, I suspect that clear fitness effects may be seen at much lower doses of copper ammonium acetate. This could be shown by mixed competitions between WT and tolC knock-out mutants over a range of copper concentrations, similar to what was shown for sub-MIC selection of antibiotic resistance (Gullberg et al 20112). I do however suspect that a simpler assay using relative growth rates as a proxy for relative fitness at lower concentrations than 120microgramml copper ammonium acetate would increase the sensitivity of identifying a “minimum co-selective” coper concentration.\n\n2. Non- inheritable resistance/tolerance as underlying mechanism for reduced tetracycline susceptibility: The reported patching experiments where it demonstrated that E. coli survival rely on simultaneous presence of copper were done at 35microgramml tetracycline. In the dose responses presented in Fig. 2 10microgramml is used. Given the large mutational target for efflux alteration- how can you so categorically rule out existing mutations with MICs lower than 35microgramml without presenting sequence data and or doing proper MIC assays of several colonies recovered on LB plates?\n\nOther points:\nTable 3: the fold - change description does separate synergy/antagonism- the strep/copper combination is reported as 5 fold shift- but it has an opposite sign as compared to the 40 fold difference reported for copper/tet- this is a little confusing. Fig. 2: Why are the biocide concentrations here reported in ppm?- microgramml is used throughout the manuscript. Tetracycline MICs for acrAB-tolC mutants are not presented- this should be included. Figure 4: a less than 2- fold increase in expression is reported as significant. The use of single cells here takes stochastic variation in fluorescence into account and I do find the data solid but it would be good if more experimental details were provided. How many individual cells were measured, number of biological replicates ect- it is hard to extract from the deposited data (at least for this reviewer).\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": []
},
{
"id": "42814",
"date": "18 Feb 2019",
"name": "Piklu Roy Chowdhury",
"expertise": [
"Reviewer Expertise Evolution of drug resistant bacteria",
"Mobile genetic elements and Genomic Epidemiology of drug resistant bacteria."
],
"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, one of a series of reports by the group, is on the impact of sub-lethal concentrations of biocides on the ability of Escherichia coli to resist three different antibiotics, commonly used in the treatment of human diseases. Presently, the primary focus of research on antibiotic resistance is either on surveillance of drug resistance genes or on characterizing mechanisms of drug resistance. Research on anthropogenic agents which co-select for drug resistance or equip bacteria to resist drugs by alternative mechanisms is underexplored. Data presented in the manuscript addresses this knowledge gap.\n\nPreviously published experiments protocols and analytical pipelines were used to generate data in this study. Significant findings are logically discussed, citing examples from available literature.\n\nAlthough I am not an expert in statistical analysis, the authors have used the standard methodology. The overall results on the statistical analyses were therefore easy to follow. Interpretation of results is convincing and supports conclusions presented.\n\nThere is one typographic error in the manuscript. Page 5, right hand column first line: “pHJ01 transformands of BW25113…..” Please change “transformands” to “transformants”.\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/8-32
|
https://f1000research.com/articles/8-31/v1
|
09 Jan 19
|
{
"type": "Research Article",
"title": "Teenage pregnancy in Nigeria: professional nurses and educators’ perspectives",
"authors": [
"Oluwaseyi Abiodun Akpor",
"Gloria Thupayagale-Tshweneagae",
"Gloria Thupayagale-Tshweneagae"
],
"abstract": "Background: Teenage pregnancy has been regarded as a negative occurrence in recent times due to its various negative consequences on the overall wellbeing of the teenage mother and her child, the whole family system and the entire community at large. Teenage pregnancy rate is a powerful indicator of the total well-being of a population. Purpose: The aim of the study was to explore professional nurses and secondary school teachers’ own perspectives on teenage pregnancy and to ascertain the current teenage pregnancy prevention programmes within the two selected communities in Kwara and Edo states in Nigeria. Methods: The study was qualitative and contextual with an exploratory strategy. A total of 80 participants, who were professional nurses and secondary school teachers, responded to the semi-structured interview and completed a questionnaire on demographic data. Template analysis style was combined with content analysis for data analysis. Results: Our findings revealed limited teenage pregnancy prevention initiatives in the communities. The majority of the participants expressed that teenage pregnancy is a common occurrence in their communities, but it is not acceptable. More than half of the participants did not accept the usage of contraceptives by teenagers. Conclusions: The study identified a number of factors that may influence the rate of teenage pregnancy in the communities. Therefore, strategies to reduce teenage pregnancy should focus on building social capital for teenagers in communities, making information on contraception more accessible and offering programmes that empower girls in the area of sexuality.",
"keywords": [
"Teenagers",
"teenage pregnancy",
"professional nurses",
"educators",
"contraceptives",
"perspectives",
"intervention programmes",
"Nigeria"
],
"content": "Introduction\n\nAccording to Kanku and Mash1, teenage pregnancy can be defined as a teenaged or under-aged girl within the ages of 13–19 years becoming pregnant. Globally, the world’s population comprises 1.2 billion young people between the age of 15–24 years; the majority of these young people live in Sub-Saharan Africa and are susceptible to teenage pregnancies and sexually transmitted infections (STIs)2. Teenage pregnancy has been regarded as a negative occurrence in recent times due to its various negative consequences on the overall wellbeing of teenagers2–4. The consequences of the sudden role change that occurred to a teenager due to an unplanned childbirth are felt not only by the teenage mother and her child, but the whole family system as well as the entire community5,6.\n\nRegardless of the various teenage pregnancy prevention strategies and programmes that have been established, teenage pregnancies and birth-rates are still outrageously high7–9. Teenage pregnancy is regarded as an outcome of inconsistent or non-usage of contraceptives and is a risk factor for STIs, including HIV1. Teenage pregnancy has been linked to teenagers indulging in regular unprotected sexual intercourse without a reliable contraceptive, inadequate sexuality education, sexual coercion, peer-pressure, proof of one’s fertility, poor socio-economic status and promiscuity, among others1,10,11.\n\nHigh incidence of teenage pregnancy is a global phenomenon4. Teenage pregnancy is prevalent in the USA, with nearly a million teenagers becoming pregnant annually3,7,8. Likewise, in the United Kingdom, teenage pregnancy rate is still12,13. Teenage pregnancy rate is an important index that portrays the health status of a population but in Nigeria, the rate of teenage pregnancy and child bearing is a social health burden2,12,14. According to a study conducted by Amoran2 on the predictors of teenage pregnancy and its prevention in a rural town in Western Nigeria, the prevalence of teenage pregnancy among the study population was 22.9%. This is similar to other studies conducted in Nigeria12,14.\n\nOne of the most important obligations that a country can make in order to ensure a stable future economic, social and political progress is to address the health and developmental needs of its younger citizens12. Thus, there is need to respond effectively to the health and developmental challenges of teenagers in Nigeria.\n\nThe two communities used in the study were selected based on the previous findings by Edukugbo14, indicating the regions with highest rate of teenage pregnancy in Nigeria. In the study, Edukugbo indicated that the Northern states had the highest rate, while states in the South-South (SS) region had least rate of teenage pregnancy. The two professional groups were chosen based on the fact that nurses and teachers usually have constant contact with teenagers (both in the hospitals and at schools) and they have important roles to play in the prevention of unplanned teenage pregnancy. The study therefore aimed to understand and explore the professional nurses and secondary school teachers’ own perspectives on teenage pregnancy, to identify factors that in their view may influence the risk of teenage pregnancy. Besides, the study also attempted to identify current teenage pregnancy prevention programmes and initiatives that are in place within the two selected communities, thereby contributing to a deeper understanding of teenage pregnancy and suggest positive strategies that may be put in place to reduce the occurrence of teenage pregnancy.\n\n\nTheoretical application\n\nThe theoretical framework for this study is social cognitive theory15,16, which was initially referred to as social learning theory. Personal determinants of behaviour, as well as the sociocultural determinants, were addressed by the theory. As explained by Bandura15, the best way to accomplish health promotion within a setting encompasses modification of activities within social frameworks that have adverse effects on the health as opposed to changing the tendencies of individuals.\n\nAccording to social cognitive theory16, the important aspects in development are behaviour, environment and cognition. Furthermore, social cognitive theory may be used to explain why people acquire and maintain certain behavioural patterns. For example, if a teenager is to avoid certain behaviours such as risky sexual practices and teenage pregnancy, then she needs an exposure to positive and healthy environment because positive and a healthy environment will definitely produce more positive and healthy behaviours.\n\nBandura15, stated that most human conduct and behaviours are acquired observationally; by watching others, one gains an understanding of how new practices are performed, and later, these thoughts are used as guidance for actions. So, teenagers within the community observe the behaviours of others, especially the role models (such as nurses and teachers), before engaging in sexual practices. Thus, positive influences from these role models can indirectly influence the teenagers and in turn reduce risky behaviours that may lead to teenage pregnancy and childbirths.\n\nFor example, in building behavioural capability, nurses and teachers as role models can teach teenagers positive behavioural skills such as good communication skills and positive sexual practices. Adequate information must be provided and teenagers must be given the opportunities to practice the skills and receive feedback. Likewise, in observational learning, nurses and teachers are expected to model the skills taught to teenagers and modelling is most effective when the person being observed is influential and respected. Expectations mean that nurses and teachers must motivate teenagers so as to convince them appropriately, while self-efficacy involves direct observation to know teenagers’ level of confidence to practice the learned skills. Finally, there must be reinforcement in order to build their confidence in trying out the new skills and this will ensure positive sexual behaviours, thus contributing to the prevention of teenage pregnancy and its associated risks.\n\n\nMethods\n\nA qualitative approach using an exploratory research design was primarily used to explore professional nurses and educators’ perceptions and experiences of teenage pregnancy in the two communities in Nigeria.\n\nThe study was conducted in two selected secondary schools and primary health care centres in Kwara and Edo states in Nigeria. Kwara is a state located in the North-Central geopolitical zone while Edo is an inland state in the SS geopolitical zone.\n\nA purposive sampling technique was used to guide the recruitment of 80 participants who were professional nurses and secondary school teachers. In total, 10 of the participants dropped out due to time constraint and work schedule. Criteria for inclusion in the study were as follows:\n\n(1) Participants must either be a professional nurse working with teenagers in the selected primary health clinics or a teacher from the selected secondary schools during the time of data collection.\n\n(2) Participants must reside in the study area.\n\n(3) Participants must be willing to participate in the study.\n\nAppointment were booked with participants both at the schools and clinics through physical visits and follow-up phone calls to confirm availability. All participants agreed to be interviewed at their workplace. Interviews were conducted at prearranged times at the clinic and school. Recruitment of participants was performed between August and September, 2016.\n\nThe sample size was determined by saturation of data, which was achieved when 70 participants had been interviewed. Data collection and analysis were done concomitantly. Transcription of recorded information was completed within 24 hours. Data saturation was attained when no new information was forthcoming from the participants. A total of professional nurses and 40 teachers were interviewed and each sample group saturated independently. A total of 70 interviews were analyzed with new categories and 10 interviews analyzed without new categories evolving. Referential adequacy was attained, partially fulfilling the requirement of trustworthiness.\n\nData was collected from September to November 2016. The first author (O.A.A.) conducted and audiotaped the semi-structured interviews with professional nurses and secondary school teachers in each of the selected schools and clinics. The researcher explained the goals and reasons for conducting the research individually to all participants during the recruitment period. To guide the interviewer, an interview schedule was developed. The interview was guided by questions such as: knowledge of teenage pregnancy prevention programmes, occurrence, contributing factors and prevention of teenage pregnancy. Probing questions were asked so as to implore certain information from the participants. Follow-up questions were also asked in order to clarify participants ideas and thoughts. The interviews took an average of 45–50 minutes per participants. Furthermore, pretest interviews were conducted during the recruitment phase before the actual data collection with four interviewees (from each community) using participants who have similar characteristics to the study population (one teacher from a nearby school and a nurse from a different PHC centre), but who were not included in the final data. Categorical variables such as socio-cultural group, age and gender. The interviews were written and audio-recorded with participants’ permission and field notes were taken to complement the recorded information. No repeat interview was conducted.\n\nThe interviews were transcribed verbatim by OAA and samples of the transcripts were given to GTT for verification and coding. Data was analysed using a combination of three qualitative data analysis methods, namely, the template analysis style, Tesch’s content analysis approach using open coding and quasi-statistics17,18. Quasi-statistic is used as a validating mechanism so as to confirm that the inferred themes and categories precisely echo the viewpoints of the participants involved in the study. Quasi-statistics are a tabulation of the frequency certain themes or categories are supported by the data17,18. Themes were generated from the data, they were first identified by the first author and later confirmed by the second author. Software was not used in qualitative data analysis. Descriptive statistics were also used for the presentation of data.\n\nTo ensure trustworthiness, strategies such as interpersonal relationship and trust building, triangulation of data gathering methods, peer examination, member checking, authority of the researcher, dense description, consensus with independent coder and dependability audit were employed.\n\nBefore the commencement of the study, the research proposal was submitted to the Local Government Health Authority and permission to conduct the research was given (IRLG/CA/TC/6/T.1/96). An official letter was also written to the selected primary health care clinics as well as the secondary schools and permission letters were obtained from the Head of Nursing Services and the School Principals. Prior to the interview, each participant’s rights were explained and written informed consent was obtained, and also the permission to use audio recorder. To guarantee privacy, the interviews were conducted in a private room with only the participant and the researcher present.\n\n\nResults\n\nAll eighty participants (N = 80) verbatim interviews were subjected to template analysis after initial data saturation was attained after 70 interviews and the template constructed according to the categories and sub-categories that emerged from the data. As shown in Table 1, the participants were grouped into four age groups; the majority (63.7) were aged 20–40 years. In terms of socio-cultural groups, the Yoruba socio-cultural group had the most participants with 52.5% (42 of 80) followed by the Benin with 15% (12 of 80). The majority (66.3%) of the participants (53 of 80) were females.\n\nThe qualitative findings of the study are presented according to the themes and various categories generated from the data (Table 2). Each theme is described with a summary of its categories it represents. This served as a template according to which accounts (interview data) from participants were considered after the initial data analysis and the point of data saturation. Complete de-identified interview transcripts are available on figshare19.\n\nKnowledge of community teenage pregnancy prevention programmes. More than half (44 of 80) of the participants interviewed were not aware of any programmes, events or interventions that were currently in place at the community level focusing on the prevention of unplanned teenage pregnancies. About one-third (15/40) of the professional nurses mentioned different programmes such as School Sexuality Education (5/15), Society of Family Health Programme (6/15) and Church-Based Teen Programmes (5/15) but only 1 (1/15) was from the NC and mentioned Sexuality Education in the Family Planning Clinic. While more than half (21/40) of teachers from both regions mentioned various intervention programmes such as School Sexuality Education (14/21), HIV/AIDS programme on the Radio (2/21), State Family Support Programme (2/21), Church Based Teenage Programme (2/21) while one teacher from the NC mentioned School Health visit from the Ministry of Health. Notable in these findings is that the majority (34/40) of the participants from the NC region (19/20 of the professional nurses) were not aware of any teenage pregnancy prevention programme.\n\nMeaning, acceptance and cultural beliefs about dating. Dating was viewed by all 80 participants as an intimate relationship between two parties, primarily of the opposite sex, who are supposed to have known each other before they get married or relationship that will result in marriage or a love affair. However, the majority (74 of 80) expressed that it must not involve sexual intercourse.\n\nMajority (66 of 80) of the participants expressed during the interviews that dating is accepted. Of these, 44/66 believed it must only commence after the age of 20 years, while the remaining 22/66 assumed it can begin at 18 years. Notable in the findings was that all teachers from the NC opined that dating is acceptable.\n\nParticipants opinions were taken as regards their cultural belief about dating, the majority (54/80) of the participants expressed that dating is culturally acceptable in their communities. Almost half (8/20) of the professional nurses from both regions respectively, all (n=20) teachers from the SS and 18/20 of teachers from the NC. About one third (26 of 80) of the respondents emphatically stated that their culture does not support dating. The following is a citation from a professional nurse from the NC relating to cultural belief about dating:\n\n“The Yoruba culture beliefs that dating is for knowing each other to the background, if you date, you take your spouse to your parents, they will either agree or disagree due to the background of the man (especially). The background of the man or woman is checked out during dating.” [Professional nurse from NC]\n\nAcceptance and expected age of teenage sexual experimentation. The interviewed participants were asked about acceptance and the age at which teenagers are expected to get involve in sexual relationship. Almost half (37/80) of the participants stated that sexual relationship is not accepted, nor expected from teenagers at whatsoever age they are, they must wait till after marriage. However, the majority (43 of 80) expressed that teenagers are only allowed to engage in a sexual relationship when they are well above 20 years of age. The followings are citations from the participants:\n\n“I will say when the teenager is married. So, if you cannot get married as a teenager, then wait till you are married.” [Professional nurse from SS]\n\n“No age if you are a teenager. There should not be any such thing among teenagers.” [Teacher from NC]\n\nTeenage pregnancy: occurrence, acceptance and family support. During the interviews, participants were asked about the frequency of teenage pregnancies in their communities. A large number (54 of 80) of the respondents stated that they have witnessed cases of teenage pregnancy very often as it is frequent and rampant in their communities, more than two third (28 of 40) of the participants from the NC while more than half (26 of 40) from the SS. Notable from the findings was that more teachers (29 of 40) opined that teenage pregnancy was more rampart in their communities.\n\nWith regards to teenage pregnancy acceptability, the general perceptions (75 of 80) was that pregnancy among teenagers is absolutely not acceptable for whatever reason. Conversely, 5 of 40 of the professional nurses (3 from NC and 2 from SS) indicated that it is conditionally acceptable as the parents or guardians has nothing else to do rather than accept to avoid more complications.\n\nDuring the interviews, participants were if there is family support for the pregnant teenager, the majority of the sampled participants (58 of 80) responded that no family in the community will accept and support a pregnant girl, they rather refute and treat her harshly; they resulted to the fact that it is sometimes depends on the moral and religious background of that family because in it real sense, it is uncultured and biblically it is wrong, and morally it is bad. While 22 of 80 of the participants opined that even if the girl’s family do not accept the case from the beginning, but later resolutions on the matter usually see the family supporting her. Some of the observed responses are as follows:\n\n“Teenage pregnancy is rampant because some of our girls are not serious, even primary six pupils get pregnant, 12 years old are getting pregnant.” [Professional nurse from NC]\n\n“Acceptance here is not an issue. There is no community that will accept a child who is not married with pregnancy. It is supposed to be unheard of.” [Teacher from SS]\n\nDefinition, knowledge, type and acceptance of contraceptives. Participants’ opinions were taken with regards contraceptives. All the professional nurses (n = 40) had adequate understanding of contraceptives. Contraceptives was defined by all 40 participants as material things, instruments or devices, or chemical substances used in prevention of pregnancy and sexually transmitted diseases (STDs). Likewise, all the teachers (n = 40) also had previous knowledge of contraceptives. Around two-thirds (27/40) explained that contraceptives are methods of preventing unwanted pregnancy.\n\nParticipants opinions were taken as regards the types of contraceptives that they know; the professional nurses mentioned different types of contraceptives, and the majority of the teachers (38/40) also mentioned various contraceptive methods, such as condoms, 32/40 stated contraceptives pills and injections; more than half (25/40) indicated IUDs and calendar methods; while only 18/40 mentioned implant, sterilization and emergency contraceptives.\n\nAcceptance of contraceptive usage among teenagers. Participants’ also responded to the enquiry that seek to find out if they accept or can encourage teenagers to use contraceptives as a means to prevent unwanted pregnancies and STDs. Responses observed were categorized as follows; 9/20 of professional nurses from the NC and more than half (12/20) from the SS expressed that they accept and also encourage the use of contraceptives among teenagers. However, only 13/40 of teachers (6 from NC and 7 from SS) stated that they accept teenagers to use contraceptives. Significant in this was that despite the adequate knowledge of the professional nurses, only half (21/40) accept the use of contraceptives by teenagers. The following is a citation from a professional nurse from the SS relating to contraceptive usage:\n\n“It must not even exist. Teenagers should not even know where contraceptives are sold. They should wait for the right age, as there are side effects like infertility that can occur in future, it is not a thing for teenagers, and they are for grownups.” [Professional nurse from SS]\n\nProblems and benefits associated with teenage pregnancy. Perceived consequences of teenage pregnancy and child bearing as expressed by the participants were classified as health, economic and social consequences as indicated in Table 3. A notable finding was vesico-vaginal fistula that was mentioned by only the professional nurses from both regions as a possible consequence of teenage pregnancy and child bearing.\n\nTo compare the responses of the samples, the following grading scale was used: a total number of 10 responses and above (n = 20) are indicated a yes √, indicating that the majority of the sample had the same perception; 9-5 responses are indicated with, Δ indicating that there was not a consensus perception in the sample; and 0-4 responses from the sample were indicated with an x.\n\nA: Professional nurses (North-Central); B: Professional nurses (South-South); C: Teachers (North-Central); D: Teachers (South-South).\n\nThe interviewed participants were asked about the benefit of teenage pregnancy; all the participants unanimously opined that there is no benefit associated with teenage pregnancy, it is always found to pose one problem or the other. The following are examples of citations from a professional nurse from the SS and a teacher from the NC:\n\n“I don’t think any benefit can come out from such a shameful act.” [Professional nurse from SS]\n\n“There is no benefit because the unborn suffer and the mother also suffers.” [Teacher from NC]\n\nPerceived risk factors, importance and existence of parent-child communication and preventive strategies for teenage pregnancy. As revealed in Table 4, perceived risk factors that might contribute towards teenage pregnancies were classified into six categories: personal, psychological, cultural, societal and media pressures, family and economic reasons. A notable difference from the findings was that only teachers from the NC mentioned religious belief of early marriage as a perceived cause of teenage pregnancy and polygamy was also mentioned by participants from the NC.\n\nTo compare the responses of the samples, the following grading scale was used: a total number of 10 responses and above (n = 20) are indicated a yes √, indicating that the majority of the sample had the same perception; 9-5 responses are indicated with, Δ indicating that there was not a consensus perception in the sample; and 0-4 responses from the sample were indicated with an x.\n\nA: Professional nurses (North-Central); B: Professional nurses (South-South); C: Teachers (North-Central); D: Teachers (South-South).\n\nImportance and existence of parent-child relationship. On the issue of parent-child communication, participants were asked about its meaning and importance, all (n = 80) indicated that parent-child communication as an open relationship that exist between the parent and the child such that both parties express their thoughts, feelings and experience in a friendly manner, or an act of interaction between the parent and the child such that both parties discuss matters that are bothering them and as such, the parents can have a knowledge of what the child is up to.\n\nAll the participants stressed the importance of the parent-child communication as something every home must practice so as give their children the freedom to express themselves. Similarly, all the participants unanimously expressed the existence of parent-child communications in their homes. The followings are statements made by participants:\n\n“Is the process whereby the parents communicate or interacts with the child about their education, sexual health and spiritual life in a cordial way and the child should feel free enough to tell the parents about those they like and how they feel about them and those that tell them they like them and the parents will then be able to advise them on the right path, my children talk to me about everything.” [Teacher from SS]\n\n“Some children do the wrong thing in ignorance but when the parents and children communicate, they will correct an attitude in that child, the parents can impact into the life of their children through communication. It brings closer relationship between the child and the parent and it will make the children to express their feelings (what they like and don’t like) to their parents. Parents can become their children’s confidants.” [Professional nurse from NC]\n\nPreventive strategies for teenage pregnancy. With regards to teenage pregnancy prevention strategies, the participants’ responses were grouped into sex education, preventive health care, youth programmes and community engagement, personal and family life strategies (Table 5). A notable finding was that the use of contraceptives as a preventive strategy for teenage pregnancy was only mentioned by the professional nurses.\n\nTo compare the responses of the samples, the following grading scale was used: a total number of 10 responses and above (n = 20) are indicated a yes √, indicating that the majority of the sample had the same perception; 9-5 responses are indicated with, Δ indicating that there was not a consensus perception in the sample; and 0-4 responses from the sample were indicated with an x.\n\nA: Professional nurses (North-Central); B: Professional nurses (South-South); C: Teachers (North-Central); D: Teachers (South-South).\n\n\nDiscussion\n\nLearning sexual behaviour is a common developmental task of teenagers, and the teenage years are seen as a period when teenagers develop their gender identity as well as sexual exploration20,21. The majority of the participants opined that dating is acceptable and many teenagers engaged in sexual relationships.\n\nThe social stigma that was once a major consequence of out of wedlock pregnancy has been declining, although the associated health risks for both the teenage mother and her child remain. In some cultures, teenage pregnancy is perceived as a normal occurrence, a God-given gift as well as evidence of fertility5,6,22. Conversely, as shown in the findings of this study, almost all the participants expressed that teenage pregnancy, although rampant within their communities, is not acceptable.\n\nEasy access of teenagers and women to contraceptive services has been revealed as an important factor in delaying, spacing and limiting childbearing within a population9,23. As revealed in this study, although all participants have good knowledge of contraceptives, the majority (46/80) were totally against the idea of teenagers using contraceptives. Therefore, educational programmes are needed to provide adequate contraceptive information to young men and women so as to enable them to make sound informed decisions.\n\nUniversal access to sexual information and skills is necessary for all young people, as they will have to deal with their sexuality at some point in time so as to be able to make informed choices about sexuality; this will go a long way in curbing negative sexual practices among teenagers6,20,24. Likewise, as opined by all the participants in the study, parent-child communication is very essential in the prevention of risky sexual behavior among teenagers. Sexuality education should be part of the school curriculum, as all young children must be reached at an early age before they become sexually active. Also, it is paramount to linked school sexuality education to the primary health care services so as to ensure easier accessibility24.\n\nSex before marriage can be problematic because it fosters increased likelihood of an early, unwanted pregnancy and all the adverse effects likely to follow. Effects of early childbirth can be more devastating on a teenager, as she is more likely to face serious social issues such as poverty, poor education, poor health, remain a single parent, reduced chances of getting married as well as inadequate care for the baby3,7,25. Likewise, all participants in the study expressed various consequences of teenage pregnancy and childbirth.\n\nAccording to previous studies1,6,24, factors such as poor family relationships (divorce, lack of love and parental guidance), poverty, social grants, alcohol, substance abuse, rape and incest, and failure to use contraceptives has been linked to the high incidence of teenage pregnancy. Equally as discovered in this study, professional nurses and teachers mentioned all the above as perceived causes of teenage pregnancy, with the addition of factors such as covetousness, living with grandparents, illiteracy, street hawking and polygamy. This is also similar to the findings of Demographic and Health Survey26 and Ekefre27 also revealed in their findings that poverty, illiteracy and poor sexuality education seem to be the major indicators that determine the rate of unplanned teenage pregnancies in Nigeria as well as in other developing countries.\n\nFindings from this study revealed several preventive strategies of teenage pregnancy as expressed by the participants, includes adequate parent-child communication, sexuality education in schools and homes, control over social media and good role models. Equally, it has been shown that when it comes to the issue of teenage pregnancy prevention, a single intervention strategy by only a sector of the society will not solve the problem but a comprehensive approach which connects and foster linkages with one another8,9,28.\n\n\nLimitations\n\nLimitations of the study include the purposive sampling of professional nurses and secondary school teachers living in the two selected communities in Nigeria; hence, the results are not generalizable to a larger context.\n\n\nConclusions and recommendations\n\nAs reported in the study, lack of sexuality education and poor acceptance of contraceptives play a significant role in the occurrence of teenage pregnancy. Hence, it is vital for government, communities and all policy makers to target programmes and initiatives that will be directed at enlightening teenagers, parents and all stakeholders in the community on behavioural change that will encourage dissemination of appropriate and effective sexuality education.\n\nLikewise, as reported in the study, economic deprivation is an important risk factors in the occurrence of teenage pregnancy thus, social differences are an important factor to be considered in the design and implementation of programmes and strategies targeting teenage pregnancy prevention. Also, teenage pregnancy intervention programmes are still not visible within the communities; hence, there is need for the establishment and sustenance of competent and accessible teenage pregnancy intervention programmes and initiatives.\n\n\nData availability\n\nDe-identified interview transcripts are available on figshare. DOI: https://doi.org/10.6084/m9.figshare.753166419.\n\nData are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).",
"appendix": "Grant information\n\nThis study was funded by the University of South Africa Post-Doctoral Research Fund.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nKanku T, Mash R: Attitudes, perceptions and understanding amongst teenagers regarding teenage pregnancy, sexuality and contraception in Taung. South Africa Family Prac. 2010; 52(6): 563–572. Publisher Full Text\n\nAmoran OE: A comparative analysis of predictors of teenage pregnancy and its prevention in a rural town in Western Nigeria. Int J Equity Health. 2012; 11(1): 37. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAkella D, Jordan M: Impact of social and cultural factors on teen pregnancy. J Health Dispar Res Pract. 2015; 8(1): 41–62. Reference Source\n\nWorld Health Organisation: Maternal, newborn, child and adolescent health. 2017. Reference Source\n\nWhitehead E: Understanding the association between teenage pregnancy and inter-generational factors: a comparative and analytical study. Midwifery. School of Health and Social Care, University of Chester, 2009; 25(2): 147–154. PubMed Abstract | Publisher Full Text\n\nOyedele OA, Wright SCD, Maja TMM: Community participation in teenage pregnancy prevention using the community-as-partner model. Int J Nurs Midwifery. 2014; 6(6): 80–89. Reference Source\n\nNational Campaign: Linking teen pregnancy prevention to other critical social issues. 2010.\n\nSolomom-Fears C: Teenage pregnancy prevention: statistics and programs. Congressional Research Service. 2011; 5–6. Reference Source\n\nOyedele OA, Wright SCD, Maja TMM: Community participation in teenage pregnancy prevention programmes: a systematic review. Journal of Research in Nursing and Midwifery. 2015; 4(2): 24–36. Reference Source\n\nMacleod C, Tracey T: Review of South Africa research and interventions in the development of a policy strategy on teen-aged pregnancy. South Africa: Depart of Health. 2009. Reference Source\n\nSolomon-Fears C: Teenage pregnancy prevention: statistics and programs. Congressional Research Service. 2012; 4–5.\n\nAjala AO: Factors associated with teenage pregnancy and fertility in Nigeria. Journal of Economics and Sustainable Development. 2014; 5(2): 63–70. Reference Source\n\nDennison C: Teenage pregnancy: an overview of the research evidence. 2004. Reference Source\n\nEdukugbo E: Teenage-pregnancy-anatomy-number-one-killer-girls. 2015. Reference Source\n\nBandura A: Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice-Hall, 1986. Reference Source\n\nBandura A: Observational learning, encyclopedia of learning and memory. New York: Macmillan, 1992.\n\nPolit DF, Beck CT: Nursing research: generating and assessing evidence for nursing practice. Philadelphia: Lippincott, 2008. Reference Source\n\nCreswell JW: Research design: qualitative and quantitative approaches. London: Sage, 2009.\n\nAkpor O, Thupayagale-Tshweneagae G: Teenage pregnancy in Nigeria: professional nurses and educators’ perspectives. figshare. Paper. 2018.\n\nJewkes R, Christofides N: Teenage pregnancy: rethinking prevention. Keynote Address at the 5th Youth Policy Initiative Roundtable on Teenage Pregnancy, 2008; 1–14.\n\nOyedele OA, Wright SCD, Maja TMM: Prevention of teenage pregnancies in Soshanguve, South Africa: Using the Johnson Behavioural System Model. Afr J Nurs Midwifery. 2013; 15(1): 95–108. Reference Source\n\nBabafemi AA, Adeleke JJ: Health and social problems of teenage pregnancy and future childbearing in Amassoma community, Bayelsa State, Nigeria. Res J Med Sci. 2012; 6(5): 251–260. Publisher Full Text\n\nIrja NK: Contraceptive use among women of Namibia: determinants and policy implications. PhD thesis in the Department of Sociology, University of Pretoria, South Africa, 2007. Reference Source\n\nRitcher MS, Mlambo GT: Perceptions of rural teenagers on teenage pregnancy. Health SA Gesondheid. 2005; 10(2): a195. Publisher Full Text\n\nRosenthal MS, Ross JS, Bilodeau R, et al.: Economic evaluation of a comprehensive teenage pregnancy prevention program: pilot program. Am J Prev Med. 2009; 37(6 Suppl 1): S280–287. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDemographic and Health Survey: Teenage pregnancy in Nigeria: facts and truth. 2014. Reference Source\n\nEkefre EN, Ekanem SA, Ekpenyong EA: Teenage pregnancy and education in Nigeria: a philo-sociological management strategy. Journal of Educational and Social Research. 2014; 4(3): 41–47. Publisher Full Text\n\nPanday S, Makiwane M, Ranchod C, et al.: Teenage pregnancy in South Africa - with a specific focus on school-going learners. Pretoria: Department of Basic Education. 2009. Reference Source"
}
|
[
{
"id": "45180",
"date": "19 Aug 2019",
"name": "Electra V. González Araya",
"expertise": [
"Reviewer Expertise My area of expertise are teenage pregnancy",
"adolescent fatherhood",
"sexual education",
"sexual behaviour in adolescents",
"contraception in adolescents",
"sexual abuse"
],
"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 the work is clearly and accurately presented and cites the current literature.\n\nThe style is appropriate, and follows a logical sequence.\n\nThe study design is appropriate to achieve the aim of the study.\n\nDetails of methods and analysis provided are sufficient to allow replication by others.\n\nThe source data underlying the results are sufficient, appropriate and available to ensure full reproducibility.\n\nThe conclusions drawn are adequately supported by the results.\nAccording the results of this study the lack of appropriate and effective sexuality education and, the social differences, are important factors to be considered in the design and implementation of programmes and strategies targeting teenage pregnancy prevention in Nigeria.\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": "49097",
"date": "07 Oct 2019",
"name": "Rasak Bamidele",
"expertise": [
"Reviewer Expertise sociology"
],
"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 an excellent and important addition to the literature on the subject of teenage pregnancy in Nigeria. The abstract is informative and self-explanatory. The introduction provides a clear statement of the problem, relevant literature on the subject, and an adequate approach to the subject matter. The study is understandable, while the methods used are appropriate, nevertheless, the sample size is small and therefore the question of generalization is brought to bear. However, the procedures were adequately described; the more recently published study should be cited. The theory used for the study was well situated in the study and was able to explain why people acquire and maintain certain behavioural patterns, thus, as shown in the study, there are a positive influence role models has on teenagers and this, in turn, reduce risky behaviours that may lead to teenage pregnancy and childbirth. Results were presented with clarity and precision. The discussion was able to interpret the findings given the results obtained.\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/8-31
|
https://f1000research.com/articles/7-1604/v1
|
05 Oct 18
|
{
"type": "Software Tool Article",
"title": "Improve your Galaxy text life: The Query Tabular Tool",
"authors": [
"James E. Johnson",
"Praveen Kumar",
"Caleb Easterly",
"Mark Esler",
"Subina Mehta",
"Arthur C. Eschenlauer",
"Adrian D. Hegeman",
"Pratik D. Jagtap",
"Timothy J. Griffin",
"Praveen Kumar",
"Caleb Easterly",
"Mark Esler",
"Subina Mehta",
"Arthur C. Eschenlauer",
"Adrian D. Hegeman",
"Pratik D. Jagtap"
],
"abstract": "Galaxy provides an accessible platform where multi-step data analysis workflows integrating disparate software can be run, even by researchers with limited programming expertise. Applications of such sophisticated workflows are many, including those which integrate software from different ‘omic domains (e.g. genomics, proteomics, metabolomics). In these complex workflows, intermediate outputs are often generated as tabular text files, which must be transformed into customized formats which are compatible with the next software tools in the pipeline. Consequently, many text manipulation steps are added to an already complex workflow, overly complicating the process and decreasing usability, especially for non-expert bench researchers focused on obtaining results. In some cases, limitations to existing text manipulation are such that desired analyses can only be carried out using highly sophisticated processing steps beyond the reach of most users. As a solution, we have developed the Query Tabular Galaxy tool, which leverages a SQLite database generated from tabular input data. This database can be queried and manipulated to produce transformed and customized tabular outputs compatible with downstream processing steps. Regular expressions can also be utilized for even more sophisticated manipulations, such as find and replace and other filtering actions. Using several Galaxy-based multi-omic workflows as an example, we demonstrate how the Query Tabular tool dramatically streamlines and simplifies the creation of multi-step analyses, efficiently enabling complicated textual manipulations and processing. This tool should find broad utility for users of the Galaxy platform seeking to develop and use sophisticated workflows involving text manipulation on tabular outputs.",
"keywords": [
"Galaxy",
"Workflows",
"SQLite",
"Multi-omics",
"Genomics",
"Proteomics",
"Metaproteomics",
"Proteogenomics",
"Metabolomics"
],
"content": "Introduction\n\nThe Galaxy platform1 offers a highly flexible bioinformatics workbench in which disparate software tools can be deployed and integrated into sophisticated workflows. Frequently, these workflows contain many steps and different software tools, with many different types of outputs. Each output can then act as the input for a subsequent software tool. Often, the results outputted from a software tool are in the form of a tabular file, which serve as input to a subsequent tool in the workflow. To make these workflows functional, usually the tabular output(s) must be manipulated, extracting and re-formatting the original file and creating a new tabular file with a data structure which can be read by a downstream software tool. In some cases, the final tabular results file from the workflow must be further processed and manipulated to obtain desired information for interpretation by the user.\n\nThere are many examples of multi-step workflows requiring manipulations of tabular text files employed across the diverse analysis applications facilitated by Galaxy. One example is emerging “multi-omic” analyses, which integrate software from different ‘omic domains and are well suited to the strengths of Galaxy2. For example, proteogenomics integrates tools for RNA-Seq assembly and analysis, software for matching tandem mass spectrometry (MS/MS) data to peptide and protein sequences, and other customized tools to characterize novel, variant protein sequences expressed within a sample3,4. To enable compatibility between the software tools composing a proteogenomics workflow, tabular files often must be manipulated into appropriate formats recognized by specific tools. Another example is Galaxy workflows for metaproteomics5,6, a multi-omics analysis which requires text manipulations in workflows integrating metagenomic, MS-based proteomics and other functional and taxonomic software tools. Finally, Galaxy-based metabolomics data analysis solutions are also emerging7–9, which utilize tabular inputs and outputs within multiple step workflows.\n\nUnder the category of “Text Manipulation”, the Galaxy Tool Shed has long offered many tools for extracting and transforming information within tabular files produced in workflows. However, sophisticated workflows (e.g. multi-omics, metabolomics), can require numerous manipulations to tabular files in order to build fully integrated and automated pipelines. Consequently, workflows can grow to hundreds of steps, dominated by sequential text manipulation steps. This situation makes the building and optimizing of such workflows highly time-consuming and prone to errors, requiring much effort even by experienced Galaxy users. It also hampers efforts to further customize or modify workflows by other users, if these change formats of the tabular files, necessitating another round of optimization of many text manipulations.\n\nTo improve the available options for text manipulation in Galaxy, we have developed a new suite of tools, which we generally refer to as Query Tabular. Query Tabular leverages the power of SQLite, automatically creating a database directly from desired tabular outputs within a workflow using the Query Tabular tool. The SQLite database can be saved to the Galaxy history, and acted upon by the companion SQLite_to_Tabular tool, generating additional tabular outputs containing desired information and formatting. As such, Query Tabular streamlines complicated text manipulations, greatly simplifying the creation and customization of Galaxy workflows, and in some cases enabling new analyses. Here, we show the use of Query Tabular in several example Galaxy-based workflows, demonstrating its value. Query Tabular is available through the Galaxy Tool Shed and should prove highly useful to a broad community of Galaxy users.\n\n\nMethods\n\nThe Query Tabular tools use Python applications to read and filter tabular files, and the Python package sqlite to create and query a SQLite database. There are 3 main functions performed:\n\n1. Line filtering. For a tabular file, a sequence of line filters can be used to transform each line as it is read. A line filter takes one TAB-separated line and produces 0 or more TAB-separated lines. For example, a line filter that filters out comment lines only produces an output line when an input line does not begin with a comment character. The normalization line filter splits a line that has a comma-separated value in one (or more) specified fields into one output row per list item.\n\n2. Loading a SQLite table. The filtered tabular file is inspected for number of TAB-separated fields and the SQLite type of the values in each field - Real, Integer, or Text - followed by generation of a database table for that file. Each line from the filtered tabular file is then loaded as a row in that table.\n\n3. Querying the database. A SQL query is executed on the database. The results are written out as a new tabular-formatted text file.\n\nThe query_tabular.py application can perform all three of the steps above. However, the query can be omitted when the SQLite database is the only desired output. The sqlite_to_tabular.py application only performs the query function given an existing SQLite database as input. This can be useful when one needs to perform several queries on the same database. The filter_tabular.py application performs the line filtering function to directly produce a tabular file. This can be sufficient for simple selection of rows and columns from a single file.\n\nGalaxy tools have been developed for each of the actions described above, and are called “Query Tabular”, “SQLite to Tabular”, and “Filter Tabular”, respectively. These Galaxy tools provide a web form for a user to specify input files and settings for line filters, table and column names, and a SQL query. The Galaxy framework makes it easy to link these tools with other software and processing steps, creating multi-step workflows.\n\nFigure 1 shows a screenshot of the Galaxy-based Query Tabular tool. The Query Tabular Galaxy tool loads any number of tabular datasets into a new or existing SQLite database allowing the full power of a SQL query to produce a new tabular output. Long, complicated workflows of Galaxy text manipulation tools can be replaced by Query Tabular in a single step.\n\nThe user can select the tabular data which acts as input and is converted to a SQLite database. The input data tables can be filtered if desired. The interface also provides a field to define the queries for the SQLite database that will be carried out, along with options for displaying results from the query in the tabular output.\n\nThe Query Tabular tool provides default names for tables - t1, t2, etc. - and columns - c1, c2, etc. - but a user can specify more specific and meaningful names for tables and columns. When column names are specified in the first row of the tabular file, the user has the option to use those names when selecting the columns to be loaded into the SQLite database.\n\nRegex functions, which apply regular expressions, are added to sqlite connections so that re.search, re.match, and re.replace functions are available for use in the SQL query. Line filters can apply regular expressions while reading tabular input files to include, exclude, or modify lines before entering the values as rows in the database table. A column replace line filter can use a regex function to change, for example, a date value to the SQLite recognized format. A normalize filter can convert list fields in the input to first normal form with an individual list item per row; when several fields are specified in a normalization filter, an input line having lists of length n in the specified columns results in n output row, each with one respective pair of values from the specified fields.\n\n\nUse cases\n\nBelow we provide examples of use cases for Query Tabular, focusing on Galaxy-based workflows for proteogenomics, metaproteomics and metabolomics.\n\nA common task in a proteogenomics data analysis is to match MS/MS fragmentation spectra to variant peptide sequences, which derive from genomic mutations, expression from genomic regions thought to be non-coding or silenced, or unexpected RNA splicing events10. The veracity of putative variant sequences matched to MS/MS spectra must be confirmed, which can be accomplished by querying the variant peptide sequences against NCBI’s non-redundant (nr) protein database using the BLASTP tool, which is implemented in Galaxy4. Those peptides which do not have a 100% alignment and sequence match to known sequences within the database qualify as verified variant sequences, which are then passed on for further analysis3,4.\n\nThe workflow for carrying out this analysis of putative variant peptide sequences is shown in Figure 2. This workflow takes as input the peptide spectrum matches (PSMs) containing matches to putative variant amino acid sequences, and analyzes these using BLASTP, producing a list of verified PSMs to true variant sequences. Figure 2A outlines the initial workflow, which contained 9 total steps and required multiple text manipulations with Galaxy tools. The text manipulations format the input tabular file for BLASTP analysis, extracting and re-formatting information from the PSM input. A number of manipulations are also required on the BLASTP alignments: querying the tabular files for peptides with alignment identities less than 100%, those with any gaps in the sequence alignment or those which lacked full-length matching of the known peptides to the putative variant sequence.\n\nThis workflow takes as input peptide spectrum matches (PSMs) of putative variant peptide sequences and further analyzes them using BLASTP to verify sequences which are truly variants compared to the reference proteome. A) The initial workflow comprised of nine total steps, including multiple text manipulation steps in Galaxy; B) The simplified workflow when using Query Tabular, which reduces the number of steps to 4 to obtain the same results.\n\nWhen Query Tabular is used, the individual text manipulation steps are not needed, and the number of steps is reduced from 9 to 4 (Figure 2B). We have made this workflow available for demonstration purposes at z.umn.edu/proteogenomicsgateway. Supplementary File 1 provides instructions on accessing and using this workflow.\n\nMetaproteomic workflows seek to identify peptide sequences expressed by a community of microorganisms, usually bacteria. These sequences are further analyzed to characterize the taxonomic distributions of the bacteria present in the community; the peptides are also mapped to protein groups which have known biochemical functions, such that the peptides can be indicators of specific functional responses of the community to external perturbations11,12.\n\nIn one established metaproteomics Galaxy workflow6, the microbial peptides must be verified by matching to the NCBI nr database, using the BLASTP tool (Figure 3). A number of text manipulation steps are required to make the file of identified peptide sequences compatible with BLASTP. The BLASTP-aligned sequences are outputted in a tabular file, and this file must be further manipulated via several steps in order to create a tabular file in correct format for downstream functional and taxonomic analysis. In all, this workflow ends up requiring many text manipulations in order to achieve desired results. Figure 3A highlights these numerous manipulation steps.\n\nThis workflow verifies the presence of detected microbial peptides by matching peptides against the NCBI nr protein sequence database using the BLASTP tool. (A) Using conventional Galaxy text manipulation tools, the workflow requires 17 steps to achieve desired outputs. (B) When utilizing Query Tabular, desired results are obtained in seven steps.\n\nQuery Tabular greatly simplifies this metaproteomics workflow. As shown in Figure 3B, use of Query Tabular eliminates many of the initial steps required to generate a tabular input compatible with BLASTP. It also greatly simplifies the second part of the workflow where the BLASTP outputs are further manipulated to generate a tabular file which is required for further taxonomic and functional analysis. In all, using Query Tabular reduced the length of the workflow from 17 steps to 7. We have made this workflow available for demonstration purposes at z.umn.edu/metaproteomicsgateway. Supplementary File 1 provides instructions on accessing and using this workflow.\n\nA Galaxy-based metabolomics workflow provides an example where Query Tabular was used to enable efficient data correction and analysis that was not possible with other existing Galaxy tools. This workflow utilizes VKMZ, a metabolomics tool under development which predicts and plots metabolites from liquid chromatography (LC)-MS data. Metabolite predictions are made by comparing the neutral mass of observed signals to a dictionary of known mass-formulas. When a signal’s neutral mass is within a given mass error range of a known mass, a prediction is made.\n\nFor the use-case presented here, targeted metabolomics data were collected on a low resolution LC-MS instrument. Low mass standards in the data, used to provide more accurate mass assignments to observed signals, had a systematic mass shift caused by using an instrument calibration method for high mass molecules. Figure 4 shows the two-part SQL query inputted in the Query Tabular tool and used to correct this shift, operating on the tabular data generated from MS data by VKMZ, which assumes charge (z) is 1. The inner-query determines the average relative mass error for molecules with low mass-to-charge (mz) values (molecular mass <250 Daltons) in the data. The outer-query adjusts all detected molecules within this same mz range by the average mass error. Before making mass adjustment with Query Tabular, VKMZ was able to predict 85.7% of the features for the standards. After the mass adjustment, VKMZ was able to correctly predict all features for the standards. This two-step manipulation, with dependency of the outer-query on the result from the inner-query, is concise and would require generation of a nested, multiple step workflow within the larger workflow if using existing text manipulation tools in Galaxy. We have made this workflow available for demonstration purposes at z.umn.edu/metaproteomicsgateway. Supplementary File 1 provides instructions on accessing and using this workflow.\n\nThe two part SQL query corrects mass errors in low resolution MS-based metabolomics data, using an inner- and outer-query. The inner-query (lines 9-11) determine the average mass error for mz values of detected molecules below 250 Daltons. The outer-query (all other lines) adjusts all mz values in this range based on the determined mass error. Chromatographic retention time (rt) and signal intensities are also assigned values for the molecules detected by LC-MS.\n\n\nConclusions\n\nWe have described a new Galaxy tool, Query Tabular, which significantly improves the development and application of multi-step workflows in Galaxy. Leveraging a SQLite database, and utilizing regular expressions, the tool can minimize the need for lengthy workflows using conventional Galaxy-based text manipulation tools. This eases the process of workflow development, producing more efficient workflows which can be utilized and understood more easily by non-expert bench researchers. We have provided use-case examples in the area of multi-omics (proteogenomics and metaproteomics) demonstrating the value of Query Tabular in this way. Via an example in metabolomics, we also demonstrate how Query Tabular can enable new manipulations and analyses of textual data within a single, simplified workflow, that would otherwise require separate workflow development if attempted using existing Galaxy tools. The Query Tabular tool has also proven useful and versatile for developing workflows used for multi-omic informatic training workshops (http://galaxyp.org/workshops/) and online training via the Galaxy Training Network (http://galaxyproject.github.io/training-material13). The free and open tool is available to any Galaxy user, and should provide a valuable addition to the Galaxy tool box for developing analysis workflows.\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\nThe Query Tabular suite of tools can be added to a Galaxy server from the Galaxy Tool Shed: https://toolshed.g2.bx.psu.edu/view/iuc/query_tabular/1ea4e668bf73.\n\nSource code available from: https://github.com/galaxyproject/tools-iuc/tree/master/tools/query_tabular.\n\nArchived source code at time of publication: https://doi.org/10.5281/zenodo.143929614.\n\nLicense: MIT license.\n\nAdding tools from the Tool Shed is an administrative function of a Galaxy server, and as a security precaution is restricted to users designated as admins for the server. From the Galaxy server, an admin simply searches for the tool in the toolshed and clicks the install button.\n\nAs we described above, we have also made available example workflows for demonstration purposes using Query Tabular on outputs from proteogenomics data (z.umn.edu/proteogenomicsgateway) and metaproteomics & metabolomics data (z.umn.edu/metaproteomicsgateway). Supplementary File 1 contains instructions on how to access these example workflows.",
"appendix": "Grant information\n\nThis work was supported in part by NSF award 1458524 and NIH award U24CA199347 to T.J. Griffin and the Galaxy for proteomics (Galaxy-P) research team.\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\nWe thank the Supercomputing Institute at the University of Minnesota for support and maintenance of software and hardware infrastructure used in the development of this tool. We also thank the Jetstream team at the University of Indiana for support and maintenance of software and hardware infrastructure used for hosting publicly accessible Galaxy instances described in this manuscript. Additionally, we would like to thank Kevin Murray from the Department of Veterinary Population Medicine at the University of Minnesota for contributing the metabolomics data set.\n\n\nSupplementary material\n\nSupplementary File 1. Detailed instructions on accessing and operating the demonstration workflows which utilize Query Tabular.\n\nClick here to access the data.\n\n\nReferences\n\nAfgan E, Baker D, Batut B, et al.: The Galaxy platform for accessible, reproducible and collaborative biomedical analyses: 2018 update. Nucleic Acids Res. 2018; 46(W1): W537–W44. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoekel J, Chilton JM, Cooke IR, et al.: Multi-omic data analysis using Galaxy. Nat Biotechnol. 2015; 33(2): 137–9. PubMed Abstract | Publisher Full Text\n\nChambers MC, Jagtap PD, Johnson JE, et al.: An Accessible Proteogenomics Informatics Resource for Cancer Researchers. Cancer Res. 2017; 77(21): e43–e46. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJagtap PD, Johnson JE, Onsongo G, et al.: Flexible and accessible workflows for improved proteogenomic analysis using the Galaxy framework. J Proteome Res. 2014; 13(12): 5898–908. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBlank C, Easterly C, Gruening B, et al.: Disseminating Metaproteomic Informatics Capabilities and Knowledge Using the Galaxy-P Framework. Proteomes. 2018; 6(1): pii: E7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJagtap PD, Blakely A, Murray K, et al.: Metaproteomic analysis using the Galaxy framework. Proteomics. 2015; 15(20): 3553–65. PubMed Abstract | Publisher Full Text\n\nDavidson RL, Weber RJ, Liu H, et al.: Galaxy-M: a Galaxy workflow for processing and analyzing direct infusion and liquid chromatography mass spectrometry-based metabolomics data. GigaScience. 2016; 5: 10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGuitton Y, Tremblay-Franco M, Le Corguillé G, et al.: Create, run, share, publish, and reference your LC-MS, FIA-MS, GC-MS, and NMR data analysis workflows with the Workflow4Metabolomics 3.0 Galaxy online infrastructure for metabolomics. Int J Biochem Cell Biol. 2017; 93: 89–101. PubMed Abstract | Publisher Full Text\n\nWeber RJM, Lawson TN, Salek RM, et al.: Computational tools and workflows in metabolomics: An international survey highlights the opportunity for harmonisation through Galaxy. Metabolomics. 2017; 13(2): 12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNesvizhskii AI: Proteogenomics: concepts, applications and computational strategies. Nat Methods. 2014; 11(11): 1114–25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHettich RL, Pan C, Chourey K, et al.: Metaproteomics: harnessing the power of high performance mass spectrometry to identify the suite of proteins that control metabolic activities in microbial communities. Anal Chem. 2013; 85(9): 4203–14. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRudney JD, Jagtap PD, Reilly CS, et al.: Protein relative abundance patterns associated with sucrose-induced dysbiosis are conserved across taxonomically diverse oral microcosm biofilm models of dental caries. Microbiome. 2015; 3: 69. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBatut B, Hiltemann S, Bagnacani A, et al.: Community-Driven Data Analysis Training for Biology. Cell Syst. 2018; 6(6): 752–758.e1. PubMed Abstract | Publisher Full Text\n\nJohnson JE: query_tabular (Version 3.0.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1439296"
}
|
[
{
"id": "39106",
"date": "05 Nov 2018",
"name": "Daniel Blankenberg",
"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\nSummary:\n\nThe authors describe a set of Galaxy tools collectively referred to as “Query Tabular” (but composed of 3 individual tools “Query Tabular”, “SQLite to Tabular”, and “Filter Tabular”). This tool allows user-defined database operations to be performed on tabular files within Galaxy through the use of generated sqlite intermediate files. Tabular files are very common outputs of bioinformatics tools, and a significant number of Galaxy tools exist for manipulating these types of files. By enabling the the use of SQL statements to transform tabular files, a great deal of effort that currently requires several tools from the ‘standard’ Galaxy text manipulation toolbox can be performed in far fewer steps.\nGeneral comments:\nQuery Tabular does enable powerful manipulations to be performed, and it can simplify a workflow which may otherwise have many simple text manipulation tools connected together to achieve a similar result. A significant caveat is that the most powerful functions require the user to have working knowledge of SQL (‘simple’ things like filtering do not). For pre-canned workflows, this is not a problem, but for a typical ‘bench scientist’ attempting to use Query Tabular this may prove to be a formidable barrier to usage when developing their own analysis pipelines. This isn’t necessarily a problem with the tool, just a fact of the intended tool design, but it does place it into more of the power-user category. This does enable someone with SQL knowledge to easily do a bunch of neat things inside of Galaxy, and it might be beneficial to include a link to a resource with general help on writing SQL and perhaps provide an additional resource that provides examples of some typical operations relevant to common Galaxy tools.\nEssential changes:\nFigure 1: Have “Table Options” section expanded to show that table name is being set and that header line is being used for column names. It would also be helpful to provide a small snippet (~5 lines or so) of the input tabular file that is selected in an additional panel.\n\nProvide direct downloads for each example input file and exported workflow (perhaps at Zenodo or a Github repository, etc). Currently a reader needs to visit and register an account at two separate Galaxy servers to gain access to these examples.\n\nMinor suggestions:\nIt might be useful to provide a link to, or list by name, the specific Galaxy Training material tutorial that currently makes use of Query Tabular (https://galaxyproject.github.io/training-material/topics/proteomics/tutorials/metaproteomics/tutorial.html) when mentioning the online training available.\n\nMore examples of ‘real-world’ usage could be helpful, especially to users that are less experienced with SQL. Perhaps as a linked external Github page, or similar.\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? Partly\n\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article? Yes",
"responses": [
{
"c_id": "4322",
"date": "31 Dec 2018",
"name": "Tim Griffin",
"role": "Author Response",
"response": "We thank the reviewer for the comments. Below in bold text we provide responses to these comments and revisions we have made in the updated version.General comments:Query Tabular does enable powerful manipulations to be performed, and it can simplify a workflow which may otherwise have many simple text manipulation tools connected together to achieve a similar result. A significant caveat is that the most powerful functions require the user to have working knowledge of SQL (‘simple’ things like filtering do not). For pre-canned workflows, this is not a problem, but for a typical ‘bench scientist’ attempting to use Query Tabular this may prove to be a formidable barrier to usage when developing their own analysis pipelines. This isn’t necessarily a problem with the tool, just a fact of the intended tool design, but it does place it into more of the power-user category. This does enable someone with SQL knowledge to easily do a bunch of neat things inside of Galaxy, and it might be beneficial to include a link to a resource with general help on writing SQL and perhaps provide an additional resource that provides examples of some typical operations relevant to common Galaxy tools.>The suggestion about clarifying the target user audience for this tool is well-taken. It is true that Query Tabular does require working knowledge of SQL, and as such higher level users and developers of Galaxy benefit most directly from this tool. We have acknowledged this in the Conclusions section, and also we point readers to training material for those unfamiliar with SQL. Essential changes:1) Figure 1: Have “Table Options” section expanded to show that table name is being set and that header line is being used for column names. It would also be helpful to provide a small snippet (~5 lines or so) of the input tabular file that is selected in an additional panel.>We have expanded Figure 1 to show the “collapsed” view of the Query Tabular tool, as well as expanded views of the Table options and Filtering menus which open when selected in the tool. We have also provided a view of snippets of the tabular data that comprise input and output for the use case workflows shown in Figure 2 (proteogenomics) and Figure 3 (metaproteomics). The figure legends have also been updated reflecting these changes.2) Provide direct downloads for each example input file and exported workflow (perhaps at Zenodo or a Github repository, etc). Currently a reader needs to visit and register an account at two separate Galaxy servers to gain access to these examples.>We have deposited workflow files (.ga format) for the 3 use cases and also the example input data for these into an accessible Github repository at: https://github.com/galaxyproteomics/query_tabular_supplementary_materialWe describe access to the workflow files and data in this Github repository in the Software Availability section. This repository also contains a README file describing the use case workflows and input data.Minor suggestions:1. It might be useful to provide a link to, or list by name, the specific Galaxy Training material tutorial that currently makes use of Query Tabular (https://galaxyproject.github.io/training-material/topics/proteomics/tutorials/metaproteomics/tutorial.html) when mentioning the online training available.>We have added a link to a newly created training tutorial on proteogenomics that is part of the GTN (https://galaxyproject.github.io/training-material/topics/proteomics/tutorials/proteogenomics-novel-peptide-analysis/tutorial.html). We have added this link to the Conclusions section and we also mention this link as another example of application of Query Tabular for complex data manipulations in the introductory text of the Use Case section.2. More examples of ‘real-world’ usage could be helpful, especially to users that are less experienced with SQL. Perhaps as a linked external Github page, or similar.>We have emphasized in the text the Galaxy wrapper which provides some simplified explanation of how the Query Tabular tool operates (see second line of text in Use Cases section). For those wanting more exposure to more complex data workflows which mimic those encountered by many researchers, we point them to the three use case workflows which were designed as representative, more complex examples where this tool has value and are now available for direct download and install from a Github repository (https://github.com/galaxyproteomics/query_tabular_supplementary_material). Additionally, at the start of the Use Case section, we now point the readers to the proteogenomics workflow on the Galaxy Training Network which provides another “real-world” application of Query Tabular (https://galaxyproject.github.io/training-material/topics/proteomics/tutorials/proteogenomics-novel-peptide-analysis/tutorial.html). We also point readers to Galaxy Training Network material focused on metaproteomics (https://galaxyproject.github.io/training-material/topics/proteomics/tutorials/metaproteomics/tutorial.html) which also utilizes Query Tabular and provides a detailed explanation of the tool and its use within this complex, real-world workflow."
}
]
},
{
"id": "39107",
"date": "13 Nov 2018",
"name": "Maria A. Doyle",
"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\nJohnson et al’s paper describes Query Tabular, a tool for simplifying text manipulation in the Galaxy platform. The paper describes the tool and shows example use cases for proteogenomics, metaproteomics and metabolomics. The Query Tabular tool can create an SQLite database that can then be queried, saving on multiple text manipulation steps. Example workflows are provided for the use cases.\n\nI really like that the tool enables a user to create a database easily from text files. The instructions in the Supplementary Material for running the example workflows were easy to follow and all the workflows ran without issue. The tool also installed easily from the toolshed and ran without error using a example provided in the tool help section.\n\nMinor suggestions for future revisions:\n\nPart of the text is a bit confusing, Query Tabular is referred to as a single tool in most of the text, but then it is also described as “the Query Tabular tools” in the Implementation section. The user needs to know some SQL to use the tool which could be highlighted more. It could perhaps be noted that this tool might provide a way to introduce/teach SQL to Galaxy users, as it enables them to create a database easily without the need to install anything, so they could just focus on learning SQL queries, like the examples provided by Data Carpentry: https://datacarpentry.org/sql-ecology-lesson/ The SQL query is shown for workflow 3 (Figure 4) but not for workflows 1 and 2. While the queries are available within the example workflows, it could be helpful to show them all in the text, especially for users not familiar with SQL. In the workflows described, it’s not always obvious what some of the steps are for (e.g. the Compute step in Figure 2, why there are two Filter sequences by length steps in Figure 3.) Perhaps the generically-named steps in the workflow figure could be changed to have more descriptive names, similar to what some of the steps already have (e.g. “Deduplicate peptides”) And/or perhaps a table could be provided describing what the individual steps are for. For the workflows, while the inputs and outputs are available in the example workflows provided, it could be helpful to provide screenshots in the text showing what the input and outputs for the use cases look like. The metabolomics workflow utilizes a tool called VKMZ that’s described as a tool under development. Would be good to note if the workflow is very specific to this tool or also relevant for other metabolomics analyses. Are there known limits on how big tables can be e.g. are tables with tens/ hundreds of thousands/ millions of rows possible. The workflows are available in the authors’ Galaxy but it could be noted that the tool is available in other public Galaxies (e.g usegalaxy.eu) for people running analyses there. And perhaps the example workflows could be added to the workflows tested there: https://github.com/usegalaxy-eu/workflow-testing Would be good to provide a link to the training material that shows further examples and details on how Query Tabular can be used e.g. https://galaxyproject.github.io/training-material/topics/proteomics/tutorials/metaproteomics/tutorial.html\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": "4321",
"date": "31 Dec 2018",
"name": "Tim Griffin",
"role": "Author Response",
"response": "We thank the reviewer for the comments. Below in bold text we provide responses to these comments and revisions we have made in the updated version.Johnson et al’s paper describes Query Tabular, a tool for simplifying text manipulation in the Galaxy platform. The paper describes the tool and shows example use cases for proteogenomics, metaproteomics and metabolomics. The Query Tabular tool can create an SQLite database that can then be queried, saving on multiple text manipulation steps. Example workflows are provided for the use cases. I really like that the tool enables a user to create a database easily from text files. The instructions in the Supplementary Material for running the example workflows were easy to follow and all the workflows ran without issue. The tool also installed easily from the toolshed and ran without error using a example provided in the tool help section.>We thank the reviewer for the kind comments and positive impressions of the Query Tabular tool.Minor suggestions for future revisions:--Part of the text is a bit confusing, Query Tabular is referred to as a single tool in most of the text, but then it is also described as “the Query Tabular tools” in the Implementation section.>We have clarified this in the Methods section under “Implementation”. We now clarify that Query Tabular is a single tool, but contains three different modules, in the form of python scripts, which carry out different functions. We describe these three functions, and also clarify that Query Tabular is the main Galaxy tool that provides all of this functionality and is the focus of the description manuscript.-- The user needs to know some SQL to use the tool which could be highlighted more. It could perhaps be noted that this tool might provide a way to introduce/teach SQL to Galaxy users, as it enables them to create a database easily without the need to install anything, so they could just focus on learning SQL queries, like the examples provided by Data Carpentry: https://datacarpentry.org/sql-ecology-lesson/>We appreciate the suggestion to include this link to help in learning SQL queries, which we have now included in the text in the Conclusions section. We have also clarified in the text that Query Tabular does require some SQL knowledge, and its use is targeted towards more advanced Galaxy users with SQL knowledge. For those readers without SQL knowledge, this link will provide a useful resource for training.-- The SQL query is shown for workflow 3 (Figure 4) but not for workflows 1 and 2. While the queries are available within the example workflows, it could be helpful to show them all in the text, especially for users not familiar with SQL.>We have now expanded Figures 2 and 3 and show the queries utilized in these workflows within the inset boxes. The figure legends have been updated to reflect these additions to the figures.-- In the workflows described, it’s not always obvious what some of the steps are for (e.g. the Compute step in Figure 2, why there are two Filter sequences by length steps in Figure 3.) Perhaps the generically-named steps in the workflow figure could be changed to have more descriptive names, similar to what some of the steps already have (e.g. “Deduplicate peptides”) And/or perhaps a table could be provided describing what the individual steps are for.>We have clarified in the text description of the figures that Figures 2 and 3 are meant to show the steps involved in these workflows with or without using Query Tabular, offering a visual depiction of how Query Tabular simplifies these complex workflows. Given this purpose to the figure, we decided not to go into detail on each specific step shown in the workflows.-- For the workflows, while the inputs and outputs are available in the example workflows provided, it could be helpful to provide screenshots in the text showing what the input and outputs for the use cases look like.>We have revised Figures 2 and 3 and now show small snippets of the tabular input and output data formats for these workflows.-- The metabolomics workflow utilizes a tool called VKMZ that’s described as a tool under development. Would be good to note if the workflow is very specific to this tool or also relevant for other metabolomics analyses.>The workflow shown is specific to data manipulations necessary for the VKMZ tool. Query Tabular however is generally useful for any other data manipulations that may be required for a metabolomics workflow. We have added a statement about the general applicability of Query Tabular in the Conclusions section.-- Are there known limits on how big tables can be e.g. are tables with tens/ hundreds of thousands/ millions of rows possible.>We have yet to encounter limits in terms of table size – we have used on data with millions of rows successfully. We would note that it is important to create indices on tables when dealing with a large number of rows or columns. We state this in the Operation section of the methods.-- The workflows are available in the authors’ Galaxy but it could be noted that the tool is available in other public Galaxies (e.g usegalaxy.eu) for people running analyses there. And perhaps the example workflows could be added to the workflows tested there: https://github.com/usegalaxy-eu/workflow-testing>We have added the workflows and input data for each to a Github repository, where these can now be accessed and downloaded (https://github.com/galaxyproteomics/query_tabular_supplementary_material). We also mention in the Software Availability section that the Query Tabular tool is available in the Tool Shed, and can be used on local instances or instances such as usegalaxy.eu.We are also in process of adding the three demonstration workflows to the Github site established for testing workflows (https://github.com/usegalaxy-eu/workflow-testing). These are listed under names “F1000_Metaproteomics_QueryTabular”, “F1000_Proteogenomics_QueryTabular”, etc.-- Would be good to provide a link to the training material that shows further examples and details on how Query Tabular can be used e.g. https://galaxyproject.github.io/training-material/topics/proteomics/tutorials/metaproteomics/tutorial.html>In the future, we will pursue adding Query Tabular to the Galaxy Training Network in the future, under the category of Text Manipulation. We now do point out in the text that the Galaxy Training Network material does contain a metaproteomics tutorial that utilizes Query Tabular, and provides a detailed description of using this tool within this workflow. This provides readers another example of using Query Tabular on complex data, in addition to our Use Cases described in this manuscript."
}
]
},
{
"id": "39108",
"date": "19 Nov 2018",
"name": "Margaret E Staton",
"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\nJohnson et al. present a new Galaxy tool, Query Tabular, that enables a Galaxy user to load a tab-separated value (tsv) file and then make SQL-based manipulations of that data. The tool leverages a sqlite database and is publicly available.\n\nIs the rationale for developing the new software tool clearly explained?\nYes. The authors created this tool to make sophisticated text manipulations possible within Galaxy, enabling workflows with many fewer steps than previously possible. However, the rationale that this will be used by the “non-expert bench researcher” is weak - SQL is a computational language that is not a common skill among bench researchers. Galaxy itself was built to help researchers who do not have in depth computational skills to still be able to run sophisticated informatics analysis.\nHowever, I think the tool may be used by a slightly different group of Galaxy users - Galaxy server administrators or informaticians who build and maintain workflows for others, or the tool developers that are interested in embedding their own or others' tools into useful workflows. This is a slightly different user group than the average user, but very important for making Galaxy powerful. Pointing this out in the manuscript would provide a more compelling reason for the tool and is more in line with the use cases.\n\nIs the description of the software tool technically sound?\nThe implementation section is short but covers the three basic functions.\n\nIn the methods, the original list of functions is in the order line filtering, loading the table, then querying. The next paragraphs discuss the functions in the reverse order, making it confusing.\n\nDetails on filtering the input file are scant; it would be nice to have a figure that illustrates the user interface for that part of the tool. This function appears as an unexpanded box in Figure 1.\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?\nThe tool is publicly available through the Galaxy tool shed, which is the central place for Galaxy tools.\n\nThe instructions for the tools inside Galaxy are good and helpful for figuring out how to use them. I was able to use the provided Jetstream instance to test “Query Tabular”, “SQLite to Tabular”, and “Filter Tabular”, all worked with my very simple tests. I do not think it’s feasible for the jetstream instance to persist indefinitely - could the workflows be shared via the main public Galaxy instance instead to help future readers?\n\nA README for the github repo would be helpful for other developers.\n\nIs sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool?\nIs there any way for a user to see what database tables are in their history and how they are structured (i.e. column names and data types in each column?). It would be difficult to debug why an SQL query is not working if it’s impossible to see the table and its data somehow (I ran into this problem trying the tool out). For example, someone who works with an SQL database from the command line would use sqlcmd, or through a web server, something like PhpLiteAdmin. Something like that would be very helpful inside the Galaxy interface.\nAre the conclusions about the tool and its performance adequately supported by the findings presented in the article?\nYes.\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": "4320",
"date": "31 Dec 2018",
"name": "Tim Griffin",
"role": "Author Response",
"response": "We thank the reviewer for the comments. Below in bold text we provide responses to these comments and revisions we have made in the updated version.Johnson et al. present a new Galaxy tool, Query Tabular, that enables a Galaxy user to load a tab-separated value (tsv) file and then make SQL-based manipulations of that data. The tool leverages a sqlite database and is publicly available.Is the rationale for developing the new software tool clearly explained?Yes. The authors created this tool to make sophisticated text manipulations possible within Galaxy, enabling workflows with many fewer steps than previously possible. However, the rationale that this will be used by the “non-expert bench researcher” is weak - SQL is a computational language that is not a common skill among bench researchers. Galaxy itself was built to help researchers who do not have in depth computational skills to still be able to run sophisticated informatics analysis.However, I think the tool may be used by a slightly different group of Galaxy users - Galaxy server administrators or informaticians who build and maintain workflows for others, or the tool developers that are interested in embedding their own or others' tools into useful workflows. This is a slightly different user group than the average user, but very important for making Galaxy powerful. Pointing this out in the manuscript would provide a more compelling reason for the tool and is more in line with the use cases.>We agree with this comment, and we have modified our description of the target audience as those who are more advanced Galaxy users and developers, with knowledge of SQL (see comments for reviewers above). We have removed the mention of “non-expert bench researchers” as being the main beneficiaries of this tool from the Conclusions section.Is the description of the software tool technically sound?The implementation section is short but covers the three basic functions. In the methods, the original list of functions is in the order line filtering, loading the table, then querying. The next paragraphs discuss the functions in the reverse order, making it confusing. Details on filtering the input file are scant; it would be nice to have a figure that illustrates the user interface for that part of the tool. This function appears as an unexpanded box in Figure 1.>We have expanded Figure 1 to show the view of the interface when a user selects functions in the main tool (Table options and Filtering options).Are sufficient details of the code, methods and analysis (if applicable) provided to allow replication of the software development and its use by others?The tool is publicly available through the Galaxy tool shed, which is the central place for Galaxy tools. The instructions for the tools inside Galaxy are good and helpful for figuring out how to use them. I was able to use the provided Jetstream instance to test “Query Tabular”, “SQLite to Tabular”, and “Filter Tabular”, all worked with my very simple tests. I do not think it’s feasible for the jetstream instance to persist indefinitely - could the workflows be shared via the main public Galaxy instance instead to help future readers?A README for the github repo would be helpful for other developers.>As we have mentioned for the reviewer comments above, we have now deposited the workflows and input data in a Github repository: https://github.com/galaxyproteomics/query_tabular_supplementary_materialThis repository also contains a README file as suggested.Is sufficient information provided to allow interpretation of the expected output datasets and any results generated using the tool?Is there any way for a user to see what database tables are in their history and how they are structured (i.e. column names and data types in each column?). It would be difficult to debug why an SQL query is not working if it’s impossible to see the table and its data somehow (I ran into this problem trying the tool out). For example, someone who works with an SQL database from the command line would use sqlcmd, or through a web server, something like PhpLiteAdmin. Something like that would be very helpful inside the Galaxy interface.>This is an excellent suggestion, although beyond the scope of our Query Tabular tool which is the focus of this paper. This would most likely require the development of a new Galaxy tool to make this possible, which is something that we will consider pursuing in future developments."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1604
|
https://f1000research.com/articles/8-28/v1
|
08 Jan 19
|
{
"type": "Research Article",
"title": "Antitryptical, anticoagulant and hemagglutinating activities of Eucalyptus sp. seeds",
"authors": [
"Yago Queiroz dos Santos",
"Gabriella Silva Campos Carelli",
"Bruno Oliveira de Veras",
"Virgínia Cunha Batista",
"Anderson Felipe Jácome de França",
"Márcia Vanusa da Silva",
"Elizeu Antunes dos Santos",
"Gabriella Silva Campos Carelli",
"Bruno Oliveira de Veras",
"Virgínia Cunha Batista",
"Anderson Felipe Jácome de França",
"Márcia Vanusa da Silva",
"Elizeu Antunes dos Santos"
],
"abstract": "Background: Plant biodiversity has great value for science being an inexhaustible source for new bioactive molecules capable of offering environmentally friendly and innovative solutions for various areas of the industry. The scientific community has increased their interest in the study of plant species in the search of new molecules and to determine their mechanisms of action. Plant seeds are natural sources of bioactive compounds, such as carbohydrates, lipids and proteins with special focus on enzymatic inhibitors which protect them against digestive enzymes of phytopathogens and lectins that play an important role on carbohydrate signalization and metabolism during germination. The objective of the present study was to evaluate and describe the protein profile and to test the hemagglutinating, hemolytic and anticoagulant activities, as well as the antitryptic effect of extracts and fractions obtained from seeds of Eucalyptus species. Methods: The crude protein extract was obtained from the seed of Eucalyptus sp. with 0.02 M sodium phosphate buffer, at pH 6.6, and fractionated using ammonium sulfate in order to study its antitryptical properties as well as the capacity of hemagglutination and influence on hemostasis. Results: The crude extract showed a high effectiveness for trypsin inhibition. For hemagglutinating activity, the ammonium sulfate fraction 0-30% presented better activity, while no hemolytic activity was present in the obtained fractions. For anticoagulation assay, the fraction 0-30% showed better results. Conclusions: Taken together, the obtained results demonstrate the biotechnological potential of Eucalyptus sp. seeds, although further study is still necessary to better isolate as well as describe the bioactive compounds.",
"keywords": [
"Eucalyptus sp.",
"seed",
"tripsin inhibitor."
],
"content": "Introduction\n\nThe high biodiversity of plants has led to them becoming of increasing interest to research communities due to their potential for providing new bioactive molecules with new mechanisms of action (Viegas et al., 2006). Among the most important studied plant structures, seed extracts have demonstrated potential biological activities such as protease inhibition, hemagglutinating, antibacterial and anticoagulant activities (Otieno & Analo, 2012).\n\nThe seed is the structure of a plant responsible for the propagation, and dispersion of plants in the environment, as well as nourishing and protecting the embryo at the first critical stages of germination and establishment in soil (Mello et al., 2010). To fulfill such functions these organs require a true arsenal of molecules, such as carbohydrates, lipids, amino acids and proteins (Banik et al., 2018; Mello et al., 2010).\n\nEucalyptus belongs to Myrtaceae family (Otieno & Analo, 2012) and some species of this genus are used in the treatment of certain bacterial or fungal infections in humans. Eucalyptus monoculture provides distinct products, such as wood, charcoal, resins, plywood, cellulosic ethanol, cellulose and paper (Takahashi et al., 2004). The present study had as objective to evaluate the protein profile and to test the hemagglutinating, hemolytic and anticoagulant activities, as well as the antitryptic effect of the crude extracts and fractions obtained from the seeds of Eucalyptus sp.\n\n\nMethods\n\nThe seeds used in the present study were donated by the seed bank of the National Forest (Flona) of Nísia Floresta, located in the district of Nísia Floresta, Rio Grande do Norte, Brazil. They were powdered using refrigerated mill (TE® 631/2) until obtaining a fine flour. The ground seeds were then homogenized in 0.02 M sodium phosphate buffer, pH 6.0, under constant stirring using magnetic stirrer (Solab® SL-91/A) for 4 hours at 4°C. The homogenate was centrifuged (Hettich® MIKRO 200/200R) at 10.000 x g for 30 minutes at 4°C. The supernatant was termed crude extract (EB). EB was sequentially fractionated in two steps (0–30% named after F1 and 30–60% named as F2) with ammonium sulfate at 30% (w/v) then 60% (w/v) and further centrifuged at 10,000 x g for 30 minutes at 4°C. The pellet was resuspended in distilled water and dialyzed against its same solvent. Protein quantification was performed according to the method described by Bradford (Bradford, 1976) with adaptations for microplate assay. Plate reading was performed at 595 nm using EPOCH (Biotek®) microplate reader.\n\nThe electrophoretic protein pattern of Eucalyptus sp. fractions were observed by SDS-PAGE 12.5% (SDS-PAGE kit 1615100, Bio-Rad®) according to Laemmli (Laemmli, 1970). The protein bands also were visualized by silver staining and the approximate molecular mass were estimated by SDS-PAGE using as reference the molecular weight (Kaleidoscope™, Bio-Rad®) and migration pattern of Bovine Serum Albumin (BSA) (code A9418, Sigma-Aldrich®).\n\nHuman red blood cells (from blood bags generously donated by the Hemocentro Dalton Cunha, Rio Grande do Norte, Brazil) from different types (A, B and O) treated with papain or trypsin (both of them at 0.5 mg/mL) were incubated with serial dilutions of EB, F1 or F2 in saline solution (NaCl 0.15 M) in a 96-well plate, at a ratio of 1:1. The plate was incubated for 1 hour (at pH 7.4 and 22°C), and a negative control (saline solution and red blood cells) was performed for further comparison. The degree of agglutination was visually analyzed and the titre expressed in hemagglutination unit (U.H.), which is defined as the inverse of the highest dilution where Red Blood Cells (RBCs) agglutination was observed.\n\nRBCs were separated from the plasma by sedimentation and washed three times with saline solution. Then, 100 μL of the red blood cell suspension were incubated with 100 μL of the samples (EB, F1 or F2) for 60 minutes at 25°C. For positive control, 100 μL of RBCs suspension was incubated with 100 μL of 1% Triton X-100, while 100 μL of saline was incubated with same volume of RBCs suspension for negative control. After incubation, the reaction mixture was centrifuged (Hettich® MIKRO 200/200R) at 3.200 x g for 5 minutes at 25°C. Aliquots of 100 μL of supernatants were transferred to 96-well plates and analyzed by spectrophotometry with readings at 405 nm (Pharmacia Biotec® Ultrospec 2100 pro). The mean and standard deviation was determined by three replicate assays.\n\nIn order to evaluate the capacity of trypsin inhibition by Eucalyptus sp. seeds protein, the test was performed according to Xavier-Filho (Xavier-Filho et al., 1989), where aliquots of 10 μL of bovine trypsin (code T8802, Sigma-Aldrich®) solution (0.3 mg/ml in 50 mM tris-HCl buffer, pH 7.5) were preincubated with 120 μL of 2.5 mM HCl as well as 320 μL of 50 mM Tris-HCl buffer, pH 7.5 and 50 μL of samples from Eucalyptus sp. for 15 minutes at 37°C. After this period, the reaction was started by adding 200 μL of 1% (w/v) Azocasein (code A 2765, Sigma-Aldrich®) solution for another 30 min. The reaction was finally stopped by adding 300 μL of 20% TCA. The reaction mixture was centrifuged for 12 min at 10.000 x g and 500 μL aliquot of the supernatant were alkalinized with 500 μL 2N NaOH. The effect of the fractions on the proteolytic activity was monitored by spectrophotometer (Pharmacia Biotec® Ultrospec 2100 pro) in the 410 nm wavelength.\n\nHuman blood was added to tubes containing sodium citrate and centrifuged at 3.200 x g for 5 minutes at room temperature for separation of plasma and red blood cells. This assay was performed with serial dilutions of the samples (EB, F1 and F2) in 0.15 M PBS buffer, pH 7.4. Aliquots of 90 μL plasma were mixed with 10 μL of samples in different protein amounts (100, 25, 12.5, 6.25, 3.12, 1.56, 0.78, 0.39 μg) and incubated for 3 minutes at 37°C. Then, 100 μL of 25 mM calcium chloride was added, after 1 hour the presence or not of coagulation was observed. For the negative control, 0.15 M PBS buffer, pH 7.4 with plasma was used. The tests were adapted from the United States Pharmacopeia (1965).\n\n\nResults and discussion\n\nThe electrophoretic profile showed two major bands, one near the region corresponding to the molecular weight of bovine serum albumin (BSA), in this case, used as a marker with a known molecular mass of 66 kDa, and a second band presenting a lower molecular weight, analyzed by linear regression for the used molecular marker as performed by Dos Santos (Dos Santos et al., 2018)(Figure 1).\n\nM - Marker; EB: Crude extract, F1: Fraction precipitated with ammonium sulphate 0–30%; F2: Precipitated fraction with ammonium sulphate 30–60%.\n\nEB and fractions were tested for their capacity to inhibit the activity of trypsin, a serine protease. The inhibition of trypsin was calculated based on the mass of protein estimated by the Bradford method in each sample. The results showed higher specific inhibitory activity for F1 when compared to the other fractions (Table 1). Since just a small group of proteins with same shared hydrophobicity indices are equally precipitated on each ammonium sulfate fractionation step, it increases the specific activity of proteins (de Oliveira et al., 2018) leading to a higher activity as previously described (Kunitz & Northrop, 1936; Oliveira et al., 2009).\n\nThe EB and fractions (F1 and F2) obtained from the seeds were tested for anticoagulant activity and the results obtained are presented in Table 2. Anticoagulant activity was identified in all tested samples, with the highest activity exhibited by F1 with 12.5 μg of protein. Protease inhibitors, such as those studied in the present work, have the potential to inhibit blood coagulation cascade proteases arising as potential hemostasis modulators with clinical application on the treatment of blood clots (Harish & Uppuluri, 2018; Tagnon & Soulier, 1946).\n\nThe amount of protein present per dilution is expressed in μg. (-) Absence of coagulation; (+) Presence of coagulation.\n\nThe extract and fractions were tested with erythrocytes A, B and O, treated separately, with papain and trypsin, in triplicate (Figure 2 and Figure 3). EB showed maximum titer 512 U.H. of hemagglutination for all blood types treated with papain (code P4762, Sigma-Aldrich®), and a minimum of 64 U.H. for blood B untreated with enzymes. The F1 fraction had a maximum titer of 1024 U.H for A and O erythrocytes treated with papain and trypsin, presenting the same titer for papain-treated B erythrocytes. The lowest titer of hemagglutination obtained for F1 was 256 U.H for untreated B and O types erythrocytes as described for other studies on protein seeds fractionation (Braga et al., 2015; Vodkin & Raikhel, 1986).\n\ntiter expressed in hemagglutination units (UH).\n\ntiter expressed in hemagglutination units (UH).\n\nNone of the tested samples showed hemolytic activity for any blood type even at concentrations as high as 300 μg/mL (Table 3), wich makes possible an eventual intravenous administration requiring, nonetheless, a further cytotoxicity studies in order to evaluate a proper pharmacological concentration, since same property was described for others trypsin inhibitor molecules (Sintsova et al., 2018).\n\nSaline phosphate buffer (PBS) and 1% Triton X-100, were respectively used as negative (-) and positive (+) control for hemolytic activity.\n\n\nConclusions\n\nIn summary, this study described the extraction and fractionation of proteins from Eucalyptus sp. seeds with antitryptic and hemagglutinating activities, suggesting the possible occurrence of bioactive proteins like lectin and protease inhibitors. In addition, none sample showed hemolytic activity against human erythrocytes. Taken together, the obtained results demonstrate the biotechnological potential of Eucalyptus sp. seeds, being still necessary to perform further studies in order to better isolate as well as describe the bioactive compounds to be detected.\n\n\nData availability\n\nOpen Science Framework: The underlying data generated in the present study, https://doi.org/10.17605/OSF.IO/KFAEX (Santos, 2018).\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- SDS-PAGE profile of Eucalyptus sp. seeds proteins revealed with silver nitrate\n\n- Raw data for Figure 2 and Figure 3\n\n- Protein quantification\n\n- Hemolytic Assay\n\n- Trypsin activity",
"appendix": "Grant information\n\nThis work was supported by the Federal University of Rio Grande do Norte as well as by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nBanik S, Biswas S, Karmakar S: Extraction, purification, and activity of protease from the leaves of Moringa oleifera [version 1; referees: 2 approved, 1 approved with reservations]. F1000Res. 2018; 7: 1151. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBradford MM: A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 1976; 72(1–2): 248–254. PubMed Abstract | Publisher Full Text\n\nBraga AA, Rodrigues e Lacerda R, Medeiros GK, et al.: Antibacterial and hemolytic activity of a new lectin purified from the seeds of Sterculia foetida L. Appl Biochem Biotechnol. 2015; 175(3): 1689–1699. PubMed Abstract | Publisher Full Text\n\nde Oliveira CFR, Oliveira CT, Taveira GB, et al.: Characterization of a Kunitz trypsin inhibitor from Enterolobium timbouva with activity against Candida species. Int J Biol Macromol. 2018; 119: 645–653. PubMed Abstract | Publisher Full Text\n\nHarish BS, Uppuluri KB: Potential Anticoagulant Activity of Trypsin Inhibitor Purified from an Isolated Marine Bacterium Oceanimonas Sp. BPMS22 and its Kinetics. Mar Biotechnol (NY). 2018; 20(6): 780–791. PubMed Abstract | Publisher Full Text\n\nKunitz M, Northrop JH: ISOLATION FROM BEEF PANCREAS OF CRYSTALLINE TRYPSINOGEN, TRYPSIN, A TRYPSIN INHIBITOR, AND AN INHIBITOR-TRYPSIN COMPOUND. J Gen Physiol. 1936; 19(6): 991–1007. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970; 227(5259): 680–685. PubMed Abstract | Publisher Full Text\n\nMello JIDO, Barbedo CJ, Salatino A, et al.: Reserve carbohydrates and lipids from the seeds of four tropical tree species with different sensitivity to desiccation. Braz Arch Biol Technol. 2010; 53(4): 889–899. Publisher Full Text\n\nOliveira AS, Migliolo L, Aquino RO, et al.: Two Kunitz-type inhibitors with activity against trypsin and papain from Pithecellobium dumosum seeds: purification, characterization, and activity towards pest insect digestive enzyme. Protein Pept Lett. 2009; 16(12): 1526–1532. PubMed Abstract | Publisher Full Text\n\nOtieno NE, Analo C: Local indigenous knowledge about some medicinal plants in and around Kakamega forest in western Kenya [version 2; referees: 2 approved]. F1000Res. 2012; 1: 40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDos Santos YQ, de Veras BO, de França AFJ, et al.: A New Salt-Tolerant Thermostable Cellulase from a Marine Bacillus sp. Strain. J Microbiol Biotechnol. 2018; 28(7): 1078–1085. PubMed Abstract\n\nSantos YQd: Raw Data. OSF. 2018. http://www.doi.org/10.17605/OSF.IO/KFAEX\n\nSintsova O, Gladkikh I, Chausova V, et al.: Peptide fingerprinting of the sea anemone Heteractis magnifica mucus revealed neurotoxins, Kunitz-type proteinase inhibitors and a new β-defensin α-amylase inhibitor. J Proteomics. 2018; 173: 12–21. PubMed Abstract | Publisher Full Text\n\nTagnon HJ, Soulier JP: Anticoagulant activity of the trypsin inhibitor from soya bean flour. Proc Soc Exp Biol Med. 1946; 61: 440. PubMed Abstract | Publisher Full Text\n\nTakahashi T, Kokubo R, Sakaino M: Antimicrobial activities of eucalyptus leaf extracts and flavonoids from Eucalyptus maculata. Lett Appl Microbiol. 2004; 39(1): 60–64. PubMed Abstract | Publisher Full Text\n\nViegas C Jr, da Silva Bolzani V, Barreiro EJ: Os produtos naturais e a química medicinal moderna. Química Nova. 2006; 29(2): 326–337. Publisher Full Text\n\nVodkin LO, Raikhel NV: Soybean lectin and related proteins in seeds and roots of le and le soybean varieties. Plant Physiol. 1986; 81(2): 558–65. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXavier-Filho J, Campos FAP, Ary MB, et al.: Poor correlation between the levels of proteinase inhibitors found in seeds of different cultivars of cowpea (Vigna unguiculata) and the resistance/susceptibility to predation by Callosobruchus maculatus. J Agr Food Chem. 1989; 37(4): 1139–1143. Publisher Full Text"
}
|
[
{
"id": "47250",
"date": "09 Jun 2020",
"name": "Devaraja Sannaningaiah",
"expertise": [
"Reviewer Expertise Protein chemistry",
"platelet biology",
"thrombosis and hemostasis",
"nanotechnology."
],
"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\nQueiroz dos Santos et al. made an attempt here. I really appreciate their effort but there are some points to mention.\nObjective of the study is not clear.\n\nResults are not presented well.\n\nDiscussion part is also not up to the mark.\n\nLot of grammatical errors.\n\nSDS-PAGE is good, rest of the graphs are of poor quality.\nThus in its present form it is difficult to accept, but it can be reconsidered as a short communication rather than a full length paper with major revision.\nPresent the results and discussion separately.\n\nGive the microscopic image of hemagglutination and hemolysis.\n\nBetter try anticoagulant activity in-vivo.\n\nCarry out APTT and PT assay\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? No\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "64932",
"date": "29 Jun 2020",
"name": "Maria G Pereira",
"expertise": [
"Reviewer Expertise Biochemistry."
],
"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, submitted by Queiroz dos Santos et al., aims to extract (0.02 M sodium phosphate buffer, pH 6.0) and fractionate the proteins [with ammonium sulfate at 30% (w/v) – F1 and at 30–60% (w/v) - F2] of seeds from Eucalyptus sp, and to evaluate the protein pattern of the fractions by SDS-PAGE 12.5% electrophoresis, as well its hemagglutinant, hemolytic, trypsin inhibition, anticoagulant activities. This one needs to provide a major review and English review.\nIs the work clearly and accurately presented and does it cite the current literature?\nResponse: in part. See comments below.\nIs the study design appropriate and is the work technically sound?\nResponse: in part. See comments below.\nAre sufficient details of methods and analysis provided to allow replication by others?\nResponse: should be better detailed to replication.\nIf applicable, is the statistical analysis and its interpretation appropriate?\nResponse: it is important to include.\nAre all the source data underlying the results available to ensure full reproducibility?\nResponse: in part. See comments below.\nAre the conclusions drawn adequately supported by the results?\nResponse: yes, are in accordance with demonstrated results.\nMajor Reviews Introduction:\nQuestion: Are there studies involving some biological activities, more related to the present study, described for Eucalyptus genus? If positive, include in the introduction and discussion.\n\nWhy evaluate these biological activities for Eucalyptus? Improve this point.\nMethods: should be better detailed to allow replication.\nQuestion: Is there are an exsiccate number of this plant? Botanical Identification?\n\nHow much seeds and of powdered were used?\n\nSentence: ...”The ground seeds were then homogenized in 0.02 M sodium phosphate buffer, pH 6.0” - what proportion?\n\nSentence: “crude extract (EB)” - correct acronym.\n\nSentence: “method described by Bradford (Bradford, 1976) with adaptations for microplate assay” – cite reference of the microplate method.\n\nhemagglutination unit (U.H.) - correct acronym.\n\nHemagglutinating activity: cite reference of the method.\n\nHemolytic activity: cite reference of the method;\n\nInclude statistical analysis;\nResults and Discussion Table 1:\nEB, F1 and F2: to present the yields;\n\nAre there carbohydrates in these samples? It is important to dose.\n\nDiscuss why the difference in specific activity between F1 and F2 (i.e. yield, total protein, degree of purity);\n\nIs important to add a negative control;\n\nInclude in discussion the relation between the results obtained in SDS-PAGE and trypsin inhibitory activity;\nTable 2:\nImprove the presentation of Table 2, the current format is confusing;\n\nInclude the blood clotting time obtained in presence of EB, F1 and F2;\n\nInclude negative and positive controls;\n\nThrombin time (common pathway) and prothrombin (extrinsic pathway) has not been evaluated? Discuss it.\nFigure 1:\nInclude MW of the others markers protein;\n\nWhich component present in the samples is under BSA? Identify its MW in figure;\n\nEB and F1 electrophoretic profile is different from F2 fraction? Explain.\n\nInclude in discussion what’s relation between F1 electrophoretic profile and its higher specific inhibitory activity?\n\nElectrophoresis: these MW obtained is characteristic of other protease inhibitors?\nFigures 2 and 3:\nWhat’s the importance of enzymatic treatment with papain and trypsin? Why the results were higher with treated blood? Include in discussion;\n\nF2 was not tested in haemagglutinating activity? Include in the discussion the justification for this;\n\nWhy did the blood type \"O\" treated with trypsin not increase HU? Only treated with papain, is there any reason for that?\nReferences:\nStandardize according to the rules of F1000Research.\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? Partly\n\nAre the conclusions drawn adequately supported by the results? Yes",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-28
|
https://f1000research.com/articles/7-1754/v1
|
06 Nov 18
|
{
"type": "Opinion Article",
"title": "Are we aiming to miss in translational autoimmunity treatments?",
"authors": [
"Gisela M. Vaitaitis",
"David H. Wagner",
"Gisela M. Vaitaitis"
],
"abstract": "Autoimmunity treatments, fruitfully pioneered in mouse models, can be disappointing or result in immunosuppression and opportunistic infections in translational trials. Many possible reasons exist, but one major, overlooked reason may be the treatment timing in relation to circadian oscillations of the immune system. Mice and humans both have immunological circadian clocks and experience the same circulatory oscillations of immune cells with regards to their sleep/wake phases, but have opposite sleep/wake phases with regard to the daylight cycle. Therefore, researchers mainly study mice and potential autoimmunity treatments during the murine sleep/rest phase, which is when pro-inflammatory mediators and more adaptive immune cells are prevalent in the circulation. In translational trials, however, treatment administration happens primarily during a patient’s wake/activity phase, during the daytime, which is when more local and acute immune responses are active in the circulation. Therefore, we believe that the most opportune window for autoimmunity treatment may be missed in translational trials. Shifting the timing, and adjusting dosing to target only immune cells that are active at that time, may result in higher success with minimized immunosuppression or toxicities.",
"keywords": [
"autoimmunity",
"treatment",
"translational",
"circadian",
"timing",
"dosing"
],
"content": "Translational treatments of autoimmunity often do not reflect findings in mouse models\n\nMouse models of autoimmunity are widely used and, while they are not perfect, they are often the best available tools for research in potential treatments for a number of human autoimmune diseases such as type 1 diabetes, rheumatoid arthritis, and multiple sclerosis (MS), to name but a few. In such mouse models, many potential therapeutics demonstrate great effects in preventing the development of disease, ameliorating the disease symptoms, or even curing the fulminant autoimmunity1–5. However, when applying therapeutics in translational medicine, clinical trials often are disappointing, demonstrating low or short-lived efficacy6–8. Many therapeutics come with toxicities that, in some cases, can be mitigated. However, once toxicities are addressed, the efficacy often remains at a point of ameliorating symptoms to some degree, but not in all patients, and rarely do the therapeutics cure the patients6,9. There are successful treatments that ameliorate human autoimmune disease; however, those treatments can be broadly immune suppressive, causing opportunistic infections in the patient10,11.\n\nDifferences in success of autoimmunity treatments in mice compared to humans could certainly be due to that autoimmunity in humans is different from the mouse models. The mouse models likely address only some of the possible etiologies of human autoimmune disease when the human disease consists of many different facets and etiologies, genetics, environmental differences, relative exposures etc., that converge on the same types of symptoms. Therefore, those etiologies are not all treatable with one strategy that may have been elucidated in the mouse model. However, when directing the autoimmunity therapy at events upon which all the etiologies may converge, such as killing or tolerizing T cells that are attacking self-tissues, why do therapies not have better outcomes? Could timing be at the core?\n\nIn this opinion paper, we wish to highlight one major overlooked possibility for some of the discrepancies between the findings in mouse and human autoimmunity treatments. We are by no means dismissing other possibilities since there are likely many reasons for the discrepancies; however, we feel that the particular possibility presented in this piece has yet to be considered.\n\n\nCircadian rhythm of the immune system and autoimmunity treatments\n\nIn 2017, the Nobel Prize in Physiology or Medicine was awarded for work on understanding circadian rhythm12. That work has led to the understanding that not only is there a “master clock” that governs the biorhythm of humans and animals alike, but there are many organ specific “clocks” that turn individual organ processes on and off, perhaps several times in a 24 hour period. This turns out to be true for the immune system as well13. Interestingly, there are different circadian rhythms for different parts of the immune system. Basal plasma levels of pro-inflammatory cytokines such as IL-1β, TNFα, IFNγ, and IL-6 are higher during the sleep/rest phase and are paralleled by an abundance of memory and naïve T cells in the circulation13. Contrarily, anti-inflammatory cytokines, such as IL-4 and IL-10, increase upon awakening and CD8+ effector T cells as well as natural killer T cells peak during the wake/active phase13. This results in more local cytotoxic activities during the active phase, when it is also more likely that wounding and acute pathogen and antigen exposure will occur13.\n\nThe diurnal oscillation in T and B cells in the circulation is paralleled by an opposite oscillation phase in the lymph nodes14,15. These oscillations have implications in different disease courses as well as for vaccinations. For example, the disease course in the experimental autoimmune encephalomyelitis (EAE) mouse model of MS is significantly more severe if the disease induction regimen is given during the light cycle (when mice sleep/rest) than if given during the dark cycle (when mice are awake/active)14. Similarly, the magnitude of Leishmania infection is dependent on the circadian time of infection16. In humans, the timing of vaccination has been demonstrated to have an impact on the efficacy of the vaccine, where giving the influenza vaccine to patients in the morning enhanced the antibody response compared to when giving it in the afternoon17. On the flip-side of this, severe disturbances in the sleep/wake cycle, such as during shift work, has been shown to cause significant health problems, including increased association with autoimmune disease18. Given the circadian oscillations of the immune system and the impact of those oscillations on infection and vaccination/disease induction, it is not farfetched to consider that the efficacy of immune system targeting treatments would also be affected by the timing of treatment administration in relation to such oscillations.\n\nHumans and mice have an opposite sleep/rest and wake/activity schedule in regards to the daylight cycle, but have the same oscillations in the innate and adaptive immune system in regards to the sleep/rest and wake/activity phases13. Researchers tend to study mice primarily during the murine sleep/rest cycle, as that happens when the researcher is awake. Therefore, test treatments in mice affect immunological processes that are prevalent during the sleep/rest cycle, i.e. pro-inflammatory activities and memory cell formation13. When translating the findings to humans, however, the opposite occurs. Human subjects are treated primarily during the wake/activity cycle, again for convenience. Therefore, patients are treated when anti-inflammatory mediators and local cytotoxic immune responses are active in the circulation but not when adaptive responses are prevalent. Since pro-inflammatory activities drive autoimmunity, and that type of activity occurs in the circulation mostly during the sleep/rest phase, the best window for treatment in humans may very well be missed (Figure 1). Depending on the stability of a therapeutic, much of it may have been degraded or metabolized by the time the actual intended target is present in the circulation. In the case of therapeutic antibodies, which are generally considered to have good stability, the antibodies may be engulfed through endocytosis or pinocytosis19 by cells other than the target cells and therefore become less available by the time the target cells are present. In addition, depending on what the particular therapeutic target molecule is, cells other than the intended ones, which also express that particular molecule and are present in the circulation during the wake/activity phase, may be targeted thus causing a broader immune suppression than intended.\n\nHumans and mice have different wake/activity and sleep/rest phases with regard to the daylight cycle. Because of this, researchers perform autoimmunity treatments at very different immune system peak times in humans compared to mice, which may account for some disappointing treatment trial outcomes in human autoimmunity.\n\n\nTiming and dosing in relation to toxicities and success of autoimmunity treatments\n\nRecently it has become more apparent that there is a biologically and medically relevant impact of time-of-day in administering pharmaceuticals or in encountering environmental toxicants, where the time-of-day significantly modulates the efficacy and toxicity of the administered drug or encountered toxicant20. The timing and dose of therapeutics administration in autoimmunity may also be tied to toxicities or immunosuppression. Often the thinking is that if a therapeutic with a reasonably long half-life is given in a high (but tolerated) dose then there will be a maximum effect for the patient. However, it is possible that by doing so, the therapeutic ends up targeting a much broader range of immune cells that may share expression of the same target molecule as the intended target cells, causing unnecessary immune suppression. For example, if a large dose of a T cell inhibitory antibody (targeting a molecule expressed on most T cells) is given during the wake/activity phase, it will first target CD8 effector and natural killer T cells since they are present in the circulation at that time. Subsequently, it will target the intended memory and naïve T cells once the sleep/rest phase is entered. This could happen several times over several days or weeks until the antibody is no longer available and may unnecessarily render the patient unable to respond to acute pathogen exposure. If the antibody instead is given in a smaller dose during the sleep/rest phase, it may adequately target the intended cells (memory and naïve T cells) but not be available to target other cells (CD8 effector and natural killer T cells) once the wake/activity phase is entered, thereby allowing for adequate acute pathogen response.\n\nIn the case of attempts to tolerize autoreactive T cells to self-antigens21–24, a similar scenario may be at play. Restoring tolerance with small doses of antigen works well in the setting of IgE driven allergy25,26. However, in autoimmunity there has thus far been very limited success8. One hypothesis could be that, in allergies, the target cells/molecules that initiate the allergic response are those that are active during the wake/activity phase, which is also the period that human subjects normally encounter allergens. Therefore, administering a small dose of antigen during the wake/activity phase works to target the intended immune cells and is then, because of the low dose, no longer available by the time the sleep/rest phase commences. In autoimmunity, however, the intended target cells, B cells, CD4 helper T cells, etc. are more prominent during the sleep/rest phase. Therefore, administering the antigen to humans during the daytime may target the wrong cells and the small dose of antigen will be more or less exhausted by the entry into sleep/rest phase, resulting in inadequate targeting of the intended cells.\n\n\nConcluding remarks\n\nObviously, the circadian rhythm of the immune system may not be at play in the efficacy of all autoimmunity treatments. However, considering the oscillations of immune cells in the circulation in some autoimmunity therapies may be prudent. Determining when the target molecules/immune cells are present in the circulation and administering small doses of the therapeutic at that point may maximize the effect and lessen unintended consequences. Currently, researchers are serendipitously targeting the culprit cells/molecules in autoimmune mouse model studies because of the difference in mice’s sleep/wake phases and ours. While it may be difficult to treat human autoimmunity during the sleep/rest phase, it may pay off with an increase in the therapeutic effect. To address this question, experiments can be done on autoimmune mice housed in an altered light cycle such that their sleep/wake cycle coincides with ours and therefore treatments can easily be applied when anti-inflammatory mediators and local cytotoxic immune responses are active in their circulation. Thus it can be ascertained whether the efficacy of the treatments lessen when applied during the wake/activity phase. While it may be more difficult to accomplish, in human trials of autoimmunity treatments it would be useful to assign some subjects to groups receiving treatment at different times during a 24-hour period, especially during the sleep/rest phase. Certainly, it would be worthwhile to revisit treatments that previously demonstrated great efficacy in mouse models but only slight improvements in translational treatments.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nXiao X, Guo P, Shiota C, et al.: Endogenous Reprogramming of Alpha Cells into Beta Cells, Induced by Viral Gene Therapy, Reverses Autoimmune Diabetes. Cell Stem Cell. 2018; 22(1): 78–90.e4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVaitaitis GM, Olmstead MH, Waid DM, et al.: A CD40-targeted peptide controls and reverses type 1 diabetes in NOD mice. Diabetologia. 2014; 57(11): 2366–73. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenedek G, Chaudhary P, Meza-Romero R, et al.: Sex-dependent treatment of chronic EAE with partial MHC class II constructs. J Neuroinflammation. 2017; 14(1): 100. PubMed Abstract | Publisher Full Text | Free Full Text\n\nElliott DM, Singh N, Nagarkatti M, et al.: Cannabidiol Attenuates Experimental Autoimmune Encephalomyelitis Model of Multiple Sclerosis Through Induction of Myeloid-Derived Suppressor Cells. Front Immunol. 2018; 9: 1782. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhang Q, Dehaini D, Zhang Y, et al.: Neutrophil membrane-coated nanoparticles inhibit synovial inflammation and alleviate joint damage in inflammatory arthritis. Nat Nanotechnol. 2018. PubMed Abstract | Publisher Full Text\n\nRosenblum MD, Gratz IK, Paw JS, et al.: Treating human autoimmunity: current practice and future prospects. Sci Transl Med. 2012; 4(125): 125sr1. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCreusot RJ, Postigo-Fernandez J, Teteloshvili N: Altered Function of Antigen-Presenting Cells in Type 1 Diabetes: A Challenge for Antigen-Specific Immunotherapy? Diabetes. 2018; 67(8): 1481–94. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBleich D, Wagner DH: Challenges to Reshape the Future of Type 1 Diabetes Research. J Clin Endocrinol Metab. 2018; 103(8): 2838–2842. PubMed Abstract | Publisher Full Text\n\nLi P, Zheng Y, Chen X: Drugs for Autoimmune Inflammatory Diseases: From Small Molecule Compounds to Anti-TNF Biologics. Front Pharmacol. 2017; 8: 460. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPrior DE, Nurre E, Roller SL, et al.: Infections and the relationship to treatment in neuromuscular autoimmunity. Muscle Nerve. 2018; 57(6): 927–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTudesq JJ, Cartron G, Rivière S, et al.: Clinical and microbiological characteristics of the infections in patients treated with rituximab for autoimmune and/or malignant hematological disorders. Autoimmun Rev. 2018; 17(2): 115–24. PubMed Abstract | Publisher Full Text\n\nSehgal A: Physiology Flies with Time. Cell. 2017; 171(6): 1232–5. PubMed Abstract | Publisher Full Text\n\nGeiger SS, Fagundes CT, Siegel RM: Chrono-immunology: progress and challenges in understanding links between the circadian and immune systems. Immunology. 2015; 146(3): 349–58. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDruzd D, Matveeva O, Ince L, et al.: Lymphocyte Circadian Clocks Control Lymph Node Trafficking and Adaptive Immune Responses. Immunity. 2017; 46(1): 120–32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSuzuki K, Hayano Y, Nakai A, et al.: Adrenergic control of the adaptive immune response by diurnal lymphocyte recirculation through lymph nodes. J Exp Med. 2016; 213(12): 2567–74. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKiessling S, Dubeau-Laramée G, Ohm H, et al.: The circadian clock in immune cells controls the magnitude of Leishmania parasite infection. Sci Rep. 2017; 7(1): 10892. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLong JE, Drayson MT, Taylor AE, et al.: Morning vaccination enhances antibody response over afternoon vaccination: A cluster-randomised trial. Vaccine. 2016; 34(24): 2679–85. PubMed Abstract | Publisher Full Text | Free Full Text\n\nToth LA, Trammell RA, Liberati T, et al.: Influence of Chronic Exposure to Simulated Shift Work on Disease and Longevity in Disease-Prone Inbred Mice. Comp Med. 2017; 67(2): 116–26. PubMed Abstract | Free Full Text\n\nKeizer RJ, Huitema AD, Schellens JH, et al.: Clinical pharmacokinetics of therapeutic monoclonal antibodies. Clin Pharmacokinet. 2010; 49(8): 493–507. PubMed Abstract | Publisher Full Text\n\nDallmann R, Okyar A, Lévi F: Dosing-Time Makes the Poison: Circadian Regulation and Pharmacotherapy. Trends Mol Med. 2016; 22(5): 430–45. PubMed Abstract | Publisher Full Text\n\nLudvigsson J, Krisky D, Casas R, et al.: GAD65 antigen therapy in recently diagnosed type 1 diabetes mellitus. N Engl J Med. 2012; 366(5): 433–42. PubMed Abstract | Publisher Full Text\n\nDiabetes Prevention Trial--Type 1 Diabetes Study Group: Effects of insulin in relatives of patients with type 1 diabetes mellitus. N Engl J Med. 2002; 346(22): 1685–91. PubMed Abstract | Publisher Full Text\n\nSkyler JS, Krischer JP, Wolfsdorf J, et al.: Effects of oral insulin in relatives of patients with type 1 diabetes: The Diabetes Prevention Trial--Type 1. Diabetes care. 2005; 28(5): 1068–76. PubMed Abstract | Publisher Full Text\n\nWriting Committee for the Type 1 Diabetes TrialNet Oral Insulin Study Group, Krischer JP, Schatz DA, et al.: Effect of Oral Insulin on Prevention of Diabetes in Relatives of Patients With Type 1 Diabetes: A Randomized Clinical Trial. JAMA. 2017; 318(19): 1891–902. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMatsuoka T, Shamji MH, Durham SR: Allergen immunotherapy and tolerance. Allergol Int. 2013; 62(4): 403–13. PubMed Abstract | Publisher Full Text\n\nArasi S, Corsello G, Villani A, et al.: The future outlook on allergen immunotherapy in children: 2018 and beyond. Ital J Pediatr. 2018; 44(1): 80. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "40307",
"date": "12 Nov 2018",
"name": "Chris Kevil",
"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 is an opinion article addressing the question whether missed opportunities have occurred due to inconsistent immune intervention studies between mice and men involving differential time of intervention. The proposed hypothesis is interesting and potentially plausible given the recent insights into circadian regulation of immune responses. However, the current opinion piece is not fully substantiated in certain areas requiring revision.\nAre arguments adequately supported by evidence from the literature?\nWhile the authors cite some reports highlighting circadian regulation of immunity, insufficient discussion was provided regarding specific differential immune responses and how they may actually impact tolerance strategies. Further discussion of this area would strengthen the work.\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? Yes",
"responses": [
{
"c_id": "4240",
"date": "16 Nov 2018",
"name": "Gisela Vaitaitis",
"role": "Author Response",
"response": "Dr. Kevil, Thank you very much for taking the time to review our paper. We greatly appreciate your input and agree that we neglected to discuss the specific immune responses that are relevant in autoimmunity. We will update the manuscript to reflect this once we have the comments from other reviewers as well."
}
]
},
{
"id": "40305",
"date": "20 Nov 2018",
"name": "Li Wen",
"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\nVaitaitis and Wagner published this timely opinion article regarding an obvious overlook of the circadian control of immune cells and the unsatisfactory outcome of clinical trials. The authors rightly pointed out that most, if not all, immune-therapies for human autoimmune diseases have been based on animal model, especially mouse model, studies. The mouse has a completely opposite circadian rhythm to humans. This could be one of the reasons that many clinical trials of immune therapy cannot be faithfully translated to humans. Pre-clinical studies, if not specific for circadian research, are all carried out in the daytime, when the mouse normally sleeps, whereas clinical studies in humans are conducted in the daytime, which is the human waking phase. In addition to the difference in circadian regulation of immune cell functions, importantly, the circadian regulation of metabolism of immune cells needs to be taken into consideration in the interpretation of the “failed” immunotherapies in humans, which have had positive outcomes in mouse model studies. The authors have made a very important point in highlighting this major overlooked possibility for some of the discrepancies in efficacy of treating autoimmune diseases, such as type 1 diabetes, between the mouse and human studies.\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": "4246",
"date": "20 Nov 2018",
"name": "Gisela Vaitaitis",
"role": "Author Response",
"response": "Dr. Wen, Thank you very much for taking the time to review our manuscript. We greatly appreciate your time, interest and comments."
}
]
},
{
"id": "40306",
"date": "22 Nov 2018",
"name": "Hubert M. Tse",
"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\nVaitaitis and Wagner have written an interesting opinion article regarding the shortcomings of mouse models of autoimmunity and their assessment of immunotherapies to delay disease progression in human clinical trials. Mouse models are instrumental in increasing our understanding of genetics and immune responses involved in disease pathogenesis. However, the identification of numerous interventions that can delay autoimmunity in mice are ineffective in humans. One potential scenario of the lack of success in translational studies may be due to opposite circadian clocks and sleep/wake phases in mice and humans. The authors provide an interesting opinion that is overlooked by immunologists and warrants attention in future studies. The opinion article is well written, but could be strengthened by providing examples of how some immunotherapies including anakinra and rituximab have shown to be efficacious in mice, but not in human clinical trials with respect to opposite circadian clocks.\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? Yes",
"responses": [
{
"c_id": "4254",
"date": "26 Nov 2018",
"name": "Gisela Vaitaitis",
"role": "Author Response",
"response": "Dr. Tse, Thank you very much for reviewing our article. We are grateful for your time, input, and comments. We will take your comments into account when revising the manuscript to create version 2."
}
]
},
{
"id": "40304",
"date": "07 Dec 2018",
"name": "Jon Piganelli",
"expertise": [
"Reviewer Expertise Autoimmunity",
"T cells",
"Redox-dependent signaling",
"immunometabolism"
],
"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\nVaitaitis and Wagner present an interesting hypothesis regarding the inability of effective immunotherapies for autoimmunity in mice to translate to human clinic. They describe the difference in sleep/wake cycle as a confounding factor in the translation. Where in mice, the majority of work is conducted during the rodent sleep cycle where as the authors state there is an increase in adaptive immune cells, those which are often targeted by these types of therapies. However these same treatments are given in the daylight/wake period of humans, where the adaptive immune system is less active. They propose to give these types of treatment during the sleep cycle in humans to facilitate the optimal targeting of the adaptive immune response to down-modulate the ensuing self-reactive response.\nAlthough this makes sense there are a few caveats that must also be considered. For example, the administration route and dose may have a profound impact on the overall immune response. Also for dosing and and timing we assume that tissue distribution is limited if at all, and that the agent is not getting to the target tissue. This is likely not the case, as then many agents would fail to have any real therapeutic value at all. I understand the logic but as we know pharmacokinetics is extremely complex for drug delivery. Notwithstanding, the complexity of this is far greater still, since the interaction of the circulating immune cells, for example CD4-helper cells must interact with APC/DC in the draining lymph node, where the APC has ferried antigen where it can present to T cells, This interaction then puts the APC at the center of the issue and therefore the whereabouts of these cells at the time of antigen exposure becomes critical.\nOverall, this article brings us out of the dark and into the light for just how much we take for granted in these types of experiments when moving to the clinic. Overall the author’s suggestions to assess these differences warrant that these types of experiments are investigated. For example, the design of sustained delivery of agents that can be given in more conducive time frames that then release their payload during the sleep cycle may be impactful.\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": "4289",
"date": "07 Dec 2018",
"name": "Gisela Vaitaitis",
"role": "Author Response",
"response": "Dr. Piganelli, Thank you very much for reviewing our manuscript. We greatly appreciate your time. Your comments and insights are well taken and we agree that there are many levels of complexity in autoimmunity treatments. We also agree that the explanation proposed in our manuscript for the discrepancy in success of autoimmunity treatments between mice and humans does not necessarily fit all situations. Nonetheless, adjusting the timing of administration of treatments may be very important for some therapeutics with resulting increase in success."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1754
|
https://f1000research.com/articles/8-25/v1
|
07 Jan 19
|
{
"type": "Research Note",
"title": "Accuracy of linear measurements using low dose cone beam computed tomography protocol versus direct skull linear measurements: An in vitro study",
"authors": [
"Nora Aly Al Abbady",
"Reham Mohamed Hamdy",
"Sahar Hosny El Dessouky",
"Reham Mohamed Hamdy",
"Sahar Hosny El Dessouky"
],
"abstract": "Background: Cone beam computed tomography (CBCT) imaging has been widely used for different dental applications over the last few years. It delivers a high dose of radiation compared to conventional imaging modalities. This study aimed to compare the accuracy of linear measurements conducted using a low dose CBCT protocol in comparison with direct skull linear measurements. Methods: Ten dry human skulls were included in the study. 12 linear measurements were measured directly on each skull between 23 chosen anatomical landmarks using a digital calliper. Radio-opaque markers were then glued on these anatomical landmarks. Each skull was then scanned using low dose CBCT protocol operated at 90 kVp, 7.1 mA, for 9 sec. Results: There was no statistically significant difference in the accuracy of linear measurements conducted using the low dose CBCT protocol when compared with direct linear measurements. Relative Dahlberg Error value ranged from 0.8% to 1.9%. Conclusion: Reducing mAs using a low dose CBCT protocol does not affect the accuracy of linear measurements used in craniofacial imaging tasks as compared with those taken directly on the skull by a digital calliper.",
"keywords": [
"Cone Beam CT",
"Linear Measurements",
"Low Dose."
],
"content": "Introduction\n\nTwo-dimensional (2D) imaging techniques have been used in dentistry since 1896. Despite its long clinical success, 2D imaging possesses a number of problems, including superimposition and magnification, which may result in interpretation problems of the images, whether it actually represents the anatomical structures and/or pathological conditions1.\n\nLately, dental imaging techniques have advanced with the introduction of tomography. Cone beam computed tomography (CBCT) is the most recently introduced tomography that greatly approximates the accepted standard for three-dimensional (3D) maxillofacial imaging that can guide diagnosis, treatment planning, and follow-up2,3. CBCT produces accurate images, leading it to be utilized for many dental fields, such as surgical, endodontics, prosthodontics, and orthodontics4,5.\n\nThe radiation dose imparted by a CBCT examination varies as it depends on many variables, such as the type of the CBCT machine, the chosen field of view (FOV), the number of basis images, the mode of exposure (continuous or pulsed), and the exposure parameters used for scanning6. Varying the machine’s exposure parameters will result in considerable reductions in radiation dose, which is considered advantageous from a biological point of view. However, theoretically reductions in radiation dose may possibly lead to under sampling artifacts or quantum noise that could adversely affect the diagnostic quality of the images, thus affecting the accuracy of measurements obtained from these images6.\n\nThis study aimed to compare the accuracy of linear measurements conducted using a low dose CBCT protocol in comparison with direct skull linear measurements.\n\n\nMethods\n\nThe current study was conducted on ten dry human skulls that were obtained from the Anatomy Department, Faculty of Medicine, Cairo University.\n\nThe study was approved by the Research Ethics committee of Faculty of Dentistry, Cairo University.\n\nUsing a blue permanent marker, 23 anatomical landmarks were identified on each skull by marking small points representing each landmark, then 12 linear measurements were conducted directly on the skull between these anatomical landmarks using an electronic digital calliper (Allendale Electronics Ltd, Hertfordshire, UK) (Figure 1). These linear measurements are shown in Table 1.\n\nGutta-percha cones, size 80, were cut into 2 mm rod and were glued over the drawn anatomical landmarks on the skull, to be used as radiopaque radiographic markers. After gutta-percha application, the skulls were covered with block of pink wax of 10-12 mm thickness, which was adapted carefully on the facial surface of the skull from the inferior border of the mandible till above the frontonasal suture for soft tissue simulation7,8.\n\nCBCT examinations were performed using Planmeca ProMax 3D Mid CBCT unit (Planmeca, Helsinki, Finland).\n\nThe skulls were mounted on the machine, and the laser beams were adjusted to centralize the skull within the scanning field. The skulls were then scanned with a low dose protocol of 90 kVp, 7.1 mA, 9 sec, 600 µm voxel size and 20×20 cm FOV.\n\nAfter scanning each skull, the reconstructed images were viewed on the computer screen using Romexis Viewer 4.4.O.R software. The same linear measurements conducted on the skulls were conducted on CBCT orthogonal images (Figure 2; Table 1).\n\nThe CBCT and linear measurements were conducted by two observers; the first observer repeated the reading two times with a time interval of one month between each reading.\n\nStatistical analysis was performed using SPSS (version 17), and Microsoft office Excel was used for data handling and graphical presentation. For assessment of the agreement between all measurements Dahlberg error (DE) and Relative Dahlberg Error (RDE) were used together with Intra-class Correlation Coefficients (ICC), including the 95% confidence limits of the coefficient calculated assuming analysis of variance two-way mixed model ANOVA with absolute agreement on SPSS. For both inter and intra observer reliability analysis, DE and RDE were used with ICC, including the 95% confidence limits of the coefficient. Significance level was set at P < 0.05 and two tailed test assumption was applied.\n\n\nResults\n\nThere was no statistically significant difference between the low dose CBCT measurements when compared to direct skull measurements. Using the low dose protocol, mean DE was recorded as 0.83. RDE ranged from 0.8% to 1.9% for almost all measurements (Table 2).\n\nSD, standard deviation; DE, Dahlberg error; RDE, Relative Dahlberg error; ICC, Intraclass Correlation Coefficient.\n\nThe ICC for inter- and intra-observer reliability analysis for all measurements were excellent, ranging from 0.96–1.00.\n\n\nDiscussion\n\nFor many CBCT machines, it is possible to optimize one or more of the investigated exposure parameters and therefore reduce the patient's radiation dose, while maintaining diagnostic image quality and accuracy for some diagnostic tasks9,10. Accordingly, the current study aimed to investigate the effect of reducing the mA and exposure time on the accuracy of CBCT linear measurements as compared with a digital calliper.\n\nThe results of the present study revealed that the RDE of almost all the measurements ranged from 0.8% to 1.9%, which didn't exceed 5%. This percentage has been considered as clinically acceptable and permissible relative error in the medical field, as reported by Tarazona-Álvarez et al.11 and Rokn et al.12.\n\nThe highest accepted RDE was recorded on the MORw-MORw horizontal linear measurements, 3.4%. The inverse relation between the RDE and the mean of the gold standard of MORw-MORw among all conducted measurements clearly explained its high RDE relative to its low mean of gold standard, which was 21.37%.\n\nThe results of Hidalgo et al.13 was in accordance with this study as it showed that the coefficient of variation for measurements was between 1.0% and 1.3% using different tube voltage and tube current. They concluded that a low dose protocol of 80 kV and 3 mA could be used for clinical practice, which represented as much as a 50% dose reduction compared with manufacturer’s recommendations, while giving the operator the freedom to adjust the mA by +0.5 mA on the basis of their judgments of the patient’s size.\n\nFurther confirmation was obtained from Vasconcelos et al.14. They concluded that there was no association between the increase in milliamperage and the reliability of the measurements, and recommended the use of low dose protocols when the purpose of the examination is to obtain linear measurements. They added that the 2 mA and 4 mA should be avoided because they could cause degradation to the image and could affect the visualization of the mandibular cortical bone.\n\nIn accordance with the current study, Al-Ekrish6 results revealed that on decreasing exposure time the reliability and dimensional accuracy of linear measurements for implant site evaluation were not affected.\n\nIn conclusion, the results of this study support the idea that decreasing mA and/or exposure time will not affect the accuracy of linear measurements when craniofacial imaging tasks is required.\n\n\nData availability\n\nOpen Science Framework: Dataset 1. Measurements for all 12 linear measurements for both methods: digital calliper and CBCT for both repeats of the experiment, and inter/intra observer measurements, https://doi.org/10.17605/OSF.IO/NH8FE15.\n\nOpen Science Framework: Dataset 1. Additional images of the process of linear measurements using both the digital calliper and CBCT, https://doi.org/10.17605/OSF.IO/NH8FE15.\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 author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nAlamri HM, Sadrameli M, Alshalhoob MA, et al.: Applications of CBCT in dental practice: a review of the literature. Gen Dent. 2012; 60(5): 390–400; quiz 401–2. PubMed Abstract\n\nDawood A, Patel S, Brown J: Cone beam CT in dental practice. Br Dent J. 2009; 207(1): 23–28. PubMed Abstract | Publisher Full Text\n\nBrown J, Jacobs R, Levring EJ, et al.: Basic training requirements for the use of dental CBCT by dentists: a position paper prepared by the European Academy of DentoMaxilloFacial Radiology. Dentomaxillofac Radiol. 2014; 43(1): 20130291. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNemtoi A, Czink C, Haba D, et al.: Cone beam CT: a current overview of devices. Dentomaxillofac Radiol. 2013; 42(8): 20120443. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShukla S, Chug A, Afrashtehfar KI: Role of Cone Beam Computed Tomography in Diagnosis and Treatment Planning in Dentistry: An Update. J Int Soc Prev Community Dent. 2017; 7(Suppl 3): S125–S136. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAl-Ekrish AA: Effect of exposure time on the accuracy and reliability of cone beam computed tomography in the assessment of dental implant site dimensions in dry skulls. Saudi Dent J. 2012; 24(3–4): 127–34. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCaldas Mde P, Ramos-Perez FM, de Almeida SM, et al.: Comparative evaluation among different materials to replace soft tissue in oral radiology studies. J Appl Oral Sci. 2010; 18(3): 264–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMichetti J, Basarab A, Tran M, et al.: Cone-Beam Computed Tomography contrast validation of an artificial periodontal phantom for use in endodontics. Conf Proc IEEE Eng Med Biol Soc. 2015; 2015: 7905–7908. PubMed Abstract | Publisher Full Text\n\nChoi JW, Lee SS, Choi SC, et al.: Relationship between physical factors and subjective image quality of cone-beam computed tomography images according to diagnostic task. Oral Surg Oral Med Oral Pathol Oral Radiol. 2015; 119(3): 357–365. PubMed Abstract | Publisher Full Text\n\nGoulston R, Davies J, Horner K, et al.: Dose optimization by altering the operating potential and tube current exposure time product in dental cone beam CT: a systematic review. Dentomaxillofac Radiol. 2016; 45(3): 20150254. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTarazona-Álvarez P, Romero-Millán J, Peñarrocha-Oltra D, et al.: Comparative study of mandibular linear measurements obtained by cone beam computed tomography and digital calipers. J Clin Exp Dent. 2014; 6(3): e271–4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRokn AR, Hashemi K, Akbari S, et al.: Accuracy of Linear Measurements Using Cone Beam Computed Tomography in Comparison with Clinical Measurements. J Dent (Tehran). 2016; 13(5): 333–339. PubMed Abstract | Free Full Text\n\nHidalgo Rivas JA, Horner K, Thiruvenkatachari B, et al.: Development of a low-dose protocol for cone beam CT examinations of the anterior maxilla in children. Br J Radiol. 2015; 88(1054): 20150559. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVasconcelos TV, Neves FS, Queiroz de Freitas D, et al.: Influence of the milliamperage settings on cone beam computed tomography imaging for implant planning. Int J Oral Maxillofac Implants. 2014; 29(6): 1364–8. PubMed Abstract | Publisher Full Text\n\nABBADY NAA: Dataset 1. OSF. 2018. http://www.doi.org/10.17605/OSF.IO/NH8FE"
}
|
[
{
"id": "44500",
"date": "15 Feb 2019",
"name": "Plauto Christopher Aranha Watanabe",
"expertise": [
"Reviewer Expertise Dental radiology area"
],
"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 \"aimed to compare the accuracy of linear measurements conducted using a low dose CBCT protocol in comparison with direct skull linear measurements\". I have seious doubts if \"CBCT protocol operated at 90 kVp, 7.1 mA, for 9 sec\" is really low dose CBCT. The authors cite Hidalgo et al.13 that used 80 kV and 3 mA and Vasconcelos et al.14 that used 2, 4, 6.3, 8, 10, 12, 15 mA and 60 kV, 10.8 seconds. Another problem, it was that the skulls not were placed in a polystyrene box filled with water before the CBCT examination to simulate soft tissue attenuation. This situation would cause more artifacts on the tomographic images. So, the conclusion is not fully real.\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? 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? No",
"responses": []
},
{
"id": "48149",
"date": "07 Jun 2019",
"name": "Maman Hermana",
"expertise": [
"Reviewer Expertise I have a good background in physics and experienced in medical physic and medical imaging work"
],
"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\nDirect measurement is not comparable technically to the slicing measurement method of landmark point.\n\nDefinition of statistical significance needs to be explained in more detail from statistical data.\n\nDose/exposure is not a variable of this work. Needs additional work by varying the dose/exposure to support the conclusion.\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? 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? No",
"responses": [
{
"c_id": "5274",
"date": "02 Mar 2020",
"name": "Nora Al Abbady",
"role": "Author Response",
"response": "1 - The direct and the cbct measurements were taken at the same points 2 - Didn’t get what explanation you need more for statistical significance ...would pls clarify more3 - We used different exposure one normal and the other low dose and we have another paper with ultra low doseCould you pls explain to me what results u need more to support conclusionThanks in advance"
}
]
}
] | 1
|
https://f1000research.com/articles/8-25
|
https://f1000research.com/articles/8-21/v1
|
07 Jan 19
|
{
"type": "Software Tool Article",
"title": "restfulSE: A semantically rich interface for cloud-scale genomics with Bioconductor",
"authors": [
"Shweta Gopaulakrishnan",
"Samuela Pollack",
"BJ Stubbs",
"Hervé Pagès",
"John Readey",
"Sean Davis",
"Levi Waldron",
"Martin Morgan",
"Vincent Carey",
"Shweta Gopaulakrishnan",
"Samuela Pollack",
"BJ Stubbs",
"Hervé Pagès",
"John Readey",
"Sean Davis",
"Levi Waldron",
"Martin Morgan"
],
"abstract": "Bioconductor's SummarizedExperiment class unites numerical assay quantifications with sample- and experiment-level metadata. SummarizedExperiment is the standard Bioconductor class for assays that produce matrix-like data, used by over 200 packages. We describe the restfulSE package, a deployment of this data model that supports remote storage. We illustrate use of SummarizedExperiment with remote HDF5 and Google BigQuery back ends, with two applications in cancer genomics. Our intent is to allow the use of familiar and semantically meaningful programmatic idioms to query genomic data, while abstracting the remote interface from end users and developers.",
"keywords": [
"Bioinformatics",
"REST APIs",
"HDF5",
"BigQuery",
"Bioconductor"
],
"content": "Introduction\n\nAnalyses of multiomic archives like The Cancer Genome Atlas (TCGA) and single-cell transcriptomic experiments such as the 10x 1.3 million mouse neuron dataset typically begin with downloads of large files and conversion of file contents into formats based on local preferences. In this paper we consider how targeted queries of large remote genomic data resources can be conducted using methods available for Bioconductor’s SummarizedExperiment class. For large data archives that have been centralized in cloud storage, use of this approach can help diminish effort required to manage local storage, and can facilitate interactive analysis of data subsets in familiar programming idioms, without downloading entire datasets. Clients for HDF5 or Google BigQuery are available in numerous languages; our Bioconductor interface permits access to remote archives of genomic data with familiar and semantically meaningful programmatic idioms, while abstracting the remote interface from end users and developers.\n\n\nMethods: Data structures and remote back ends\n\nLet Q denote a matrix of quantifications arising from a genome scale assay with G assay features measured on N experimental samples. The elements of Q are the numbers qij, i = 1, … , G, j = 1, …, N. Bioconductor’s SummarizedExperiment structure manages feature quantifications with associated metadata about assay features and samples.\n\nIn the 10x mouse neuron dataset, G = 27998 and N = 1.3 million. Each of the G features is a gene, and it is useful to have handy a number of feature annotations like gene name, location, functional role; suppose each gene has F such features recorded. When these quantifications and associated annotations are managed in a Bioconductor SummarizedExperiment X, the matrix Q is programmatically bound to a G × F table of feature-level metadata accessible by the rowData method, and to an N × R table of sample-level metadata accessible by colData, where R denotes the number of sample-level metadata features recorded (Huber et al.1). See Figure 1.\n\nColored regions of panels within the schematic are linked with command examples in colored text beneath the panels. For example, the purple command subsetByOverlaps(se, roi) would produce a restricted RangedSummarizedExperiment instance with features limited to those colored purple. The sizeFactors component is specific to a subclass for single cell data.\n\nIn the context of R programming, let K denote a vector of feature identifiers, S denote a vector of sample identifiers. The standard subsetting idiom X[K,S] expresses filtering of the all the information in Q and the associated metadata to features K and samples S. A GRanges instance (Lawrence et al.2) defining genomic coordinates for features may be bound to X, facilitating queries defined by genomic location (using, for example, subsetByOverlaps) to isolate features coincident with or near the elements of a set of query genomic ranges (eg., binding peaks). This outline of genomic data representation and analysis is characteristic of Bioconductor.\n\nGoogle BigQuery. The Institute for Systems Biology Cancer Genomics Cloud project (ISB-CGC) (ISB3) uses Google BigQuery to provide access to various public cancer genomics resources including TCGA and the PanCancer Atlas (Hoadley et al.4). The pancan_SE function of restfulSE constructs queries that derive SummarizedExperiment instances using quantifications and annotations for PanCancer atlas experiments managed in BigQuery tables.\n\nHDF Scalable Data Service (HSDS). An AWS S3-based distributed data object model for HDF5 datasets, including a RESTful API to structure, populate, and query HDF5 archives, has been implemented by the HDF Group. A number of datasets of interest in bioinformatics are served through HDF Kita Lab in the /shared/bioconductor folder.\n\nThe restfulSE package provides interfaces to BigQuery and HSDS so that the numerical content housed in these services satisfies the API of the Bioconductor DelayedArray (Pagès and Hickey5). Any DelayedArray instance can serve as the assay component of a SummarizedExperiment instance. Thus the capacities of SummarizedExperiment to bind semantically rich metadata to genome-scale assays are extended implicitly to data resources for which no standards exist for associating substantive metadata.\n\nIn conjunction with the rhdf5client and bigrquery packages, restfulSE functions translate filtering and selection operations which are readily defined using rowData, rowRanges, colData into formal queries resolvable by the HDF5 and BigQuery services. Numerical results are transmitted from server to client only when needed.\n\n\nResults\n\nThe RESTful SummarizedExperiment representation allows complicated research queries to be obtained in a concise, fast, convenient and robust fashion, as illustrated by the following examples.\n\nThe following code chunk, which generates Figure 2, illustrates the use of the restfulSE protocol with the ISB-CGC BigQuery back end.\n\n\n\nOur interest is in replicating part of Figure 5C of Bailey et al.6. In that paper, it is shown that microsatellite instability (MSI) is associated with different expression signatures of immune cell infiltration for adenocarcinomas of colon (COAD) and stomach (STAD), and uterine corpus endometrial carcinoma (UCEC). The MSI scores developed using MSIsensor are found in Table S5 of Ding et al.7. These scores are not available in BigQuery, but can be combined with the assay data using standard R programming, leading to a hybrid data/annotation strategy.\n\nFunctions in the BiocOncoTK package (Carey8) build on restfulSE functionality to a) authenticate the user to the BigQuery platform, b) select a tumor type (COAD) and assay for SummarizedExperiment construction, c) bind Ding et al.’s MSI values as sample-level data variable msiTest, d) acquire and transform the PD-L1 and CD8A (Entrez IDs 29126 and 925) expression values, and e) form the stratified boxplot. The basic findings of Bailey et al. are replicated. Enhancement of the code to produce a display covering more genes and tumor types is demonstrated in the BiocOncoTK package vignette. Note that in this example, expression values are only downloaded for the genes requested, without altering the end user programming paradigm of working with a SummarizedExperiment instance.\n\nFigure 3 demonstrates use of a RESTful SummarizedExperiment, with assay data provided in the object /shared/bioconductor/darmgcls.h5 at hsdshdflab.hdfgroup.org. Briefly, as a prelude to single-cell RNA-sequencing of glioblastoma (GBM) tumors from four patients, Darmanis et al.9 used immunopanning to increase the proportion of non-neoplastic cells that constitute the “migrating front” of progression of glioblastoma. Antibody to CD45 was used to capture microglial cells. Figure 3 provides code to compare the distribution of CD45 expression among the classes of cells as labeled in the metadata of GSE84465, the NCBI GEO archive from which the quantifications were derived. In this example, data on one gene from all cells is retrieved when the statement defining vector vals is executed. The display can be recapitulated for other genes by substituting different symbols in the statement computing ind. The DelayedArray framework leveraged here enables basic computations of this kind without loading the entire matrix into memory.\n\n\n\n\nPerformance\n\nWe focus on pursuit of reliability, expressivity, and scalability using restfulSE.\n\nReliability: The restfulSE, rhdf5client and BiocOncoTK packages are accompanied by detailed unit tests that compare retrievals to known values. In the case of BigQuery table queries, the test suite composes random queries in both BigQuery SQL and in the SummarizedExperiment idiom. Results are checked for elementwise equality.\n\nExpressivity: The code segments for Figure 2 and Figure 3 are complex but easy to break down. The joining and reshaping of pancan-atlas tables in BigQuery corresponding to the code in Figure 2 can be checked through the query history in the BigQuery interface. The acquisition of expression values employed five nested SELECT statements; the query for assay quantifications was 6000 characters in length. The R code is less than 500 characters including comments.\n\nScalability. BigQuery is intrinsically auto-scaling, but charges accrue with the amount of data scanned, so query design can have effects on throughput and cost. We rely on the bigrquery (Wickham10) and dbplyr (Wickham and Ruiz11) packages for efficient translation of R-oriented data manipulations to BigQuery SQL. Throughput with the HDF Scalable Data Service is dependent upon the configuration of the object server, the relationship of numerical data layout to prevalent access patterns, and the degree to which queries capitalize on API efficiencies like chunk-based retrieval. For both back ends, proper design and deployment of the querying client can lead to throughput that scale with client-side resources.\n\n\nConclusions\n\nCloud-scale storage and retrieval strategies are of significant interest for genome science. The SummarizedExperiment class unifies assay data with substantive sample- and experiment-level metadata, and its API for managing and interrogating genome-scale experiment archives is used in numerous analytic packages. The restfulSE package exposes high-performance cloud-resident data stores to users and algorithms as SummarizedExperiments. Continued improvements in efficiency of representation and query resolution for assay data and metadata will help to achieve the potential of a federated data ecosystem for enhanced discovery in biology through interactive genome-scale analysis.\n\n\nSoftware availability\n\nrestfulSE package available from: https://bioconductor.org/packages/3.9/restfulSE Source code available from: https://github.com/shwetagopaul92/restfulSE Archived source code as at time of publication: DOI: 10.18129/B9.bioc.restfulSE12 License: Artistic-2.0",
"appendix": "Grant information\n\nSupport for the development of this software was provided by NIH grants NCI U01 CA214846 (Carey, PI), NCI U24 CA180996 (Morgan, PI), and NHGRI 1U24HG010263-01 (J Taylor, PI), and Chan Zuckerberg Initiative DAF 2018-183436 (Carey, PI).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nHuber W, Carey VJ, Gentleman R, et al.: Orchestrating high-throughput genomic analysis with Bioconductor. Nat Methods. Nature Publishing Group. 2015; 12(2): 115–121. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLawrence M, Huber W, Pagès H, et al.: Software for computing and annotating genomic ranges. PLoS Comput Biol. 2013; 9(8): e1003118. PubMed Abstract | Publisher Full Text | Free Full Text\n\nISB: ISB Cancer Genomics Cloud 1.0.0 Documentation. 2018; Accessed: 2018-08-17. Reference Source\n\nHoadley KA, Yau C, Hinoue T, et al.: Cell-of-Origin Patterns Dominate the Molecular Classification of 10,000 Tumors from 33 Types of Cancer. Cell. 2018; 173(2): 291–304.e6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPagès H, Hickey P: DelayedArray: Delayed operations on array-like objects. R package version 0.7.28. 2018.\n\nBailey MH, Tokheim C, Porta-Pardo EP, et al.: Comprehensive Characterization of Cancer Driver Genes and Mutations. Cell. 2018; 173(2): 371–385.e18. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDing L, Bailey MH, Porta-Pardo E, et al.: Perspective on Oncogenic Processes at the End of the Beginning of Cancer Genomics. Cell. 2018; 173(2): 305–320.e10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCarey V: BiocOncoTK: Bioconductor components for general cancer genomics. R package version 1.1.16. 2018. Reference Source\n\nDarmanis S, Sloan SA, Croote D, et al.: Single-Cell RNA-Seq Analysis of Infiltrating Neoplastic Cells at the Migrating Front of Human Glioblastoma. Cell Rep. 2017; 21(5): 1399–1410. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWickham H: bigrquery: An Interface to Google’s ’BigQuery’ ’API’. R package version 1.0.0. 2018. Reference Source\n\nWickham H, Ruiz E: dbplyr: A ’dplyr’ Back End for Databases. R package version 1.2.1. 2018. Reference Source\n\nCarey V, Gopaulakrishnan S: restfulSE: Access matrix-like HDF5 server content or BigQuery content through a SummarizedExperiment interface. R package version 1.4.0. 2018. Publisher Full Text"
}
|
[
{
"id": "43093",
"date": "04 Feb 2019",
"name": "Dennis J. Hazelett",
"expertise": [
"Reviewer Expertise Bioinformatics",
"regulatory genomics",
"cancer genomics and epigenomics"
],
"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 restfulSE software package for bioconductor purports to extend a very useful data structure, the SummarizedExperiment to handle very large datasets wherein dynamic download of the full dataset is neither necessary nor practical. Therefore, Gopaulakrishnan et al. have created restfulSE to make this data structure interactive with remote databases on an as-needed basis.\n\nThis is a very useful idea from the Bioconductor core team, and likely to be impactful as datasets grow larger, cheaper to produce, and it becomes increasingly necessary for bioinformaticians to leverage available data against local experiments.\nThe tool is technically sound, built on Google BigQuery and HDF5, and the paper is well written and clear. The manuscript includes code examples making it simple to get a quick start and see how the software works.\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": "42652",
"date": "25 Feb 2019",
"name": "Sheila Reynolds",
"expertise": [
"Reviewer Expertise Computational biology",
"cloud-computing",
"integrative analyses of heterogeneous and large-scale cancer data sets"
],
"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 restfulSE interface described in this article by Gopaulakrishnan et al. is a very useful extension to the SummarizedExperiment class which provides a convenient approach to storing and manipulating rectangular matrices of experimental results, along with associated meta-data. This new extension allows users to query remote data, eliminating the common “download” step that still precedes many large-scale analyses.\n\nAs these datasets grow, and are more commonly made available in cloud-hosted technologies such as Google or AWS object stores or data warehouses such as Google BigQuery, tools that allow users to easily access and query these datasets become critical. The restfulSE interface permits targeted queries of such remote datasets.\nAs background information, the article includes a nice summary of the SummarizedExperiment class and related methods, for researchers (such as this reviewer) who had not come across this package before. The authors go on to describe two separate remote back ends: one which accesses PanCancer Atlas TCGA, hosted in Google BigQuery by the ISB-CGC; and the other which access HDF5 data hosted in AWS S3. Both of these backends further make use of the DelayedArray package, which implements delayed or block-processing operations to facilitate working with large datasets that cannot be stored in-memory. This enables “lazy” data retrieval, with numerical results transmitted from server to client only when needed.\nThe authors provide two concrete examples, illustrating the usage of both remote back ends. This reviewer ran into some issues trying to run these examples and reached out to the authors who provided additional information in video and Jupyter notebook form. Making additional tutorial resources available with this article will render this information useful and usable by a wider audience and is strongly encouraged.\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": []
}
] | 1
|
https://f1000research.com/articles/8-21
|
https://f1000research.com/articles/8-17/v1
|
04 Jan 19
|
{
"type": "Research Article",
"title": "An interpretable machine learning model of biological age",
"authors": [
"Thomas R. Wood",
"Christopher Kelly",
"Megan Roberts",
"Bryan Walsh",
"Christopher Kelly",
"Megan Roberts",
"Bryan Walsh"
],
"abstract": "Background: Assessments of biological (rather than chronological) age derived from patient biochemical data have been shown to strongly predict both all-cause and disease-specific mortality. However, these population-based approaches have yet to be translated to the individual. As well as using biological age as a research tool, by being able to better answer the question “why did we get this result?”, clinicians may be able to apply personalised interventions that could improve the long-term health of individual patients. Methods: Here, the boosted decision tree algorithm XGBoost was used to predict biological age using 39 commonly-available blood test results from the US National Health and Nutrition Examination Survey (NHANES) database. Results: Interrogation of the algorithm produced a description of how each marker contributed to the final output in a single individual. Additive explanation plots were then used to determine biomarker ranges associated with a lower biological age. Importantly, a number of markers that are modifiable with lifestyle changes were found to have a significant effect on biological age, including fasting blood glucose, lipids, and markers of red blood cell production. Conclusions: The combination of individualised outputs with target ranges could provide the ability to personalise interventions or recommendations based on an individual’s biochemistry and resulting predicted age. This would allow for the investigation of interventions designed to improve health and longevity in a targeted manner, many of which could be rooted in targeted lifestyle modifications.",
"keywords": [
"Aging",
"Machine Learning",
"Age"
],
"content": "Introduction\n\nOne of the fastest-growing areas at the intersection of clinical medicine and data science is the investigation of human aging1, with multiple avenues being explored to find biomarkers of aging that could be used to inform efforts to enhance human longevity2–4. If robust and easily-accessible biomarkers of aging are identified, they could assist in the rapid assessment of promising interventions aimed at increasing longevity, without the need to perform clinical trials that last decades. For instance, epigenetic modifications on DNA are increasingly being used to determine biological (rather than chronological) age, including how environmental determinants may affect an epigenetic signal for longevity4.\n\nAn individual’s biological age can be described based on the assumption that cellular aging processes, which are highly-influenced by the environment5, occur at different rates in different people with the same chronological age. As these ageing processes are associated with changes in routine biochemical measures6, algorithmic determination of biological or phenotypic age using widely-available indices such as those from blood test results is therefore becoming increasingly common. This has previously been done using both machine learning (ML) and statistical techniques3,6.\n\nOne important aspect for the utility of biological age measures is that a given output can be interpreted in order to guide individualized interventions. ML-based predictions of biological age have the potential to elucidate and describe complex, non-linear, and unintuitive patterns in biochemical data, which may provide greater predictive power compared to other statistical techniques. To date, published approaches to generate predicted biological age from biochemical data have used deep neural networks (DNNs), with the output being directly associated with mortality risk3. However, while individual outputs from DNNs are interpretable7, it is currently not possible to interrogate the effects of the entire training dataset on the model output, which may be important for determining how one may intervene given an individual’s output.\n\nAs a result of the issues with interpreting certain ML algorithms, the field of explainable artificial intelligence is developing rapidly8. If such approaches can be successfully applied to determining biological age from commonly available data, biological signatures of aging could be more rapidly discovered and tracked, including the ability to personalise interventions based on the outputs of the model. Here, we describe the development of an explainable ML model using blood marker data from the National Health and Nutrition Examination Survey (NHANES) database to predict biological age, as well as provide individual weighting for how each biomarker affected the final output. By determining how markers affect the model globally, potential target reference ranges associated with lower biological age can also be determined.\n\n\nMethods\n\nData from a total of 46,739 participants (n=22,545 males and n=24,194 females) in the NHANES database were included, with a mean (range) age of 48.5 (19.0–85.0) years. A total of 39 common blood markers were used: complete blood count (CBC) with differential, lipids, fasting glucose, iron panel, and a comprehensive metabolic panel (including electrolytes, and liver and kidney function). Descriptive data for the dataset is listed in Table 1.\n\nNHANES data (all available individuals with the 39 markers listed in Table 1 from years 1999–2015) was downloaded as .xpt files from the NHANES website using their in-built web search engine. The data was then concatenated, cross-tabulated, and stratified by gender. A random split in the data set was created to withhold 20% of participants (n=4,509 males and n=4,839) for model validation. The remaining 80% of the dataset was used to train an XGBRegressor model (XGBoost version 0.81) using chronological age and the 39 biochemical input markers. For the remaining 20% of the data, the 39 markers were provided to the algorithm9 with the chronological age withheld, and the resulting dependent variable “predicted age” defined as a measure of biological age. Age predictions for the withheld data were plotted against actual age using jointplot from the seaborn Python library (version 0.9.0).\n\nFor individual predictions, the weight of each marker was extracted using ELI5 (version 0.8.1), and graphed using a waterfall chart (version 3.8). For a given age prediction, each marker was individually weighted with regard to how it contributed to the final output. Shapley additive explanations plots (SHAP, version 0.26.0) were constructed to describe how each individual marker affects the predicted age output within the laboratory normal range.\n\nTo provide an individual output example based on data not seen by the algorithm9 previously, author C.K. had the necessary input markers measured by Quest Laboratories (Santa Cruz, CA). As C.K. is an author who ran his own data through the algorithm9 he trained during development of the manuscript, institutional ethical approval was not sought for publication of this data. C.K. approved the publication of his data in this manner.\n\n\nResults\n\nLinear regression analysis (Figure 1) showed a significant correlation between predicted (biological) and actual (chronological) age (r=0.77 and 0.75 in females and males, respectively; p<0.0001 for both). However, discrepancies between the biological and chronological age could be considered clinically relevant, as they would allow for the generation of a signature of premature biological aging.\n\nData shown for women (A) and men (B) using the 20% withheld data (n=4,509 males and n=4,839). A significant correlation between predicted and actual age (r=0.77 and 0.75 in females and males, respectively) was seen in both sexes (p<0.0001).\n\nSHAP summary plots (Figure 2) were used to determine which markers have the greatest influence on predicted biological age. The top 20 markers in terms of importance are shown. In females, blood urea nitrogen (BUN) had the greatest influence on biological age, with albumin the most influential marker in men. Fasting glucose was the second most influential marker in both sexes (Figure 2). SHAP plots for each of the 20 most influential markers are available on GitHub and Zenodo9. Based on each of these 20 markers, the level at which an inflection point was seen in the SHAP plot (i.e. when a further change in a marker would result in a net increase in predicted biological age) was determined, as well as the estimated range over which each marker would be associated with the lowest biological age (Table 2 and Table 3). Using the five most influential markers as an example, the lowest predicted age in women would be associated with a BUN 6–11 mg/dl, fasting glucose 71–86 mg/dl, bicarbonate (carbon dioxide) 19–22 mmol/l, total cholesterol 130–150 mg/dl, and mean corpuscular volume (MCV) 80–85 fl. In men, the lowest predicted age would be associated with albumin 4.6–4.8 g/dl, fasting glucose 70–88 mg/dl, BUN 6–12 mg/dl, red blood cell (RBC) 5.0–5.7 ×103/µl, and RBC distribution width (RDW) 11.0–12.5%.\n\nData shown for women (A) and men (B). Each plot is made up of thousands of individual points from the training dataset such with a higher value being more red, and a lower value being more blue. This is depicted by the “feature value” bar on the right of each plot. Therefore, if the dots on one side of the central line are increasingly red or blue, that suggests that increasing values or decreasing values, respectively, move the predicated age in that direction. For instance, lower BUN values (blue dots) are associated with lower predicted age in both men and women.\n\nRanking of markers affecting predicted age in women, in order of importance, as determined by the SHAP summary outputs. Visual examination of the individual SHAP plots for each marker was used to estimate the range over which each marker would result in the lowest predicted age, and the magnitude of the adjustment in years. The final column is the value at which a marker changes from a net negative to net positive effect on biological age.\n\nRanking of markers affecting predicted age in women, in order of importance, as determined by the SHAP summary outputs. Visual examination of the individual SHAP plots for each marker was used to estimate the range over which each marker would result in the lowest predicted age, and the magnitude of the adjustment in years. The final column is the value at which a marker changes from a net negative to net positive effect on biological age.\n\nFor a given individual, the model output allows for each marker to be individually weighted with regard to how it contributed to the final output (Figure 3). The average age in the training dataset (BIAS) is given as a starting point, with each marker subsequently increasing or decreasing predicted age by a number of years. This allows for the most influential markers for the individual to be determined. The example shown is for one of the study authors (C.K.), the data for whom is available on Zenodo9. Bias (48.3 years) is sequentially adjusted, with the five markers contributing most to an increase in biological age were BUN (+3.5 years), total cholesterol (+2.8 years), potassium (+1.7 years), phosphorus (+1.2 years), and LDH (+0.9 years). The five markers contributing most to a decrease in biological age were lymphocytes (-1.2 years), RBCs (-2.3 years), albumin (-2.7 years) fasting glucose (-3.1 years), and triglycerides (-3.9 years). The final predicted biological age was 43.0 years.\n\nBias (first column, 48.3 years) is the mean age in the input population. The five markers contributing most to an increase in biological age (columns 2–6 from the left) were BUN, total cholesterol, potassium, phosphorus, and LDH. The five markers contributing most to a decrease in biological age (columns 2–6 from the right) were lymphocytes, RBCs, albumin, glucose and triglycerides. The final predicted biological age (43.0 years) is in the last column.\n\n\nDiscussion\n\nBiomarkers of aging are increasingly important in the development and investigation of interventions with which to slow aging processes, which may also have the ability to aid in the treatment or prevention of aging-associated chronic disease. One such marker is the individual’s biological or phenotypic age, as reflected by patterns of biochemical markers in the blood, which have previously been shown to be associated with risk of mortality2,3,6. While there are a number of approaches to this problem in the published literature, we provide an alternative using a tree-based ML model that a) is fully interpretable, b) can be completely individualized for a given patient, and c) allows the development of target ranges associated with a potential signature for slowed biological aging.\n\nOne issue surrounding the utility of algorithmically-derived biological age is the response to any associated interventions or therapeutics. As this field is relatively new, it is uncertain how much an improvement in predicted biological age resulting from a given therapeutic approach will translate into improvements in longevity. Even if a given marker decreases predicted biological age, this also does not guarantee that manipulating the value will increase longevity. For instance, in our models, increasing ALT and decreasing total cholesterol were associated with lower predicted biological age; however, there are a number of scenarios where lower total cholesterol and higher ALT may be associated with increased mortality despite a lower predicted biological age10,11. Despite this, these models are at least able to generate hypotheses that can be tested in both the preclinical and clinical setting. Our approach also provides an example that other groups may use to produce fully-interpretable and personalisable outputs.\n\nThough the current analysis does not include confirmation of the ability to predict mortality risk, certain outputs from the algorithm9 do provide some confidence that the output is likely to be associated with individual health outcomes. For instance, the greatest increase in predicted age associated with fasting glucose level occurs in the range 90–100 mg/dl, which is strikingly similar to the blood glucose level associated with the largest increase in mortality risk in multiple population studies12,13. Similar associations are seen with many of the target ranges derived from the algorithm9, such as for albumin, RDW, and ferritin (especially in men)14–16.\n\nIf modulation of certain markers does indeed contribute to the reversal of cellular aging processes, the combination of an individual output with the population SHAP plots for a given marker could therefore allow for targeted therapeutic interventions aimed at improving biological age based on an individual’s specific output. For instance, elevated fasting blood glucose could be decreased by addressing diet, exercise, micronutrient deficiencies, and reducing inflammation or psychosocial stress17. Similar approaches are also likely to improve cholesterol, RDW, and MCV, confirming that lifestyle factors should play a key role in the pursuit of health and longevity15,18,19. A personalised approach is important, because the markers contributing most strongly to biological age in the whole dataset are not necessarily the same markers that most strongly contribute to a prediction in a single individual (see example in Figure 3).\n\nThe current approach does have some limitations. The dataset may only be applicable in the United States, as different countries and ethnic backgrounds might display variations in both baseline biochemistry and predicted longevity3. Expanding available input data and allowing for stratification based on nationality and ethnic background will be the focus of future work. Larger and more expanded datasets will also allow for the analysis of biological aging in association with other potentially important factors such as genetics and the microbiota20,21. It is also worth mentioning that NHANES is designed to capture data that is representative of the US population. Therefore, this data comes from participants that represent a population that has some of the highest metabolic and cardiovascular disease prevalence in the Western world22,23, which may distort the results. Additionally, the current outputs would benefit from being correlated with disease outcomes or mortality in order to determine how well predicted biological age acts as an accurate biomarker of health and longevity.\n\nBy using well-understood and robust biomarkers that are available to almost any clinician, methods such as those described in this study can be used immediately as adjuncts to research investigating the outcomes of interventions designed to increase human longevity. As multiple methods are currently available with which to predict biological or phenotypic age, the field should also collaborate in an attempt to compare methods such that we can find the approach that results in an accurate output that can most easily be used in both the research and clinical settings.\n\n\nData availability\n\nAll NHANES data used to produce the models is accessible through the CDC website (listed by NHANES study year): https://wwwn.cdc.gov/nchs/nhanes/search/default.aspx.\n\nData access, tabulation, and concatenation is automated by the “01-download-preprocess” Jupyter notebook file within our Zenodo repository; DOI: https://doi.org/10.5281/zenodo.24402039. This repository also includes the original Quest laboratory test results from author C.K., which were used to provide the worked example (Figure 3).\n\n\nSoftware availability\n\nThe algorithm developed here, including the associated libraries and the necessary versions, are available on Zenodo: https://doi.org/10.5281/zenodo.24402039.\n\nLicense: GNU General Public License version 3\n\nNotes: The algorithm itself can be trained and tested by running the “02-train-test-explain” Jupyter notebook. Note that each time the algorithm runs, a new random split in the dataset is generated in order to train and test the algorithm. Therefore, the resulting outputs might be slightly different.",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nPyrkov TV, Slipensky K, Barg M, et al.: Extracting biological age from biomedical data via deep learning: too much of a good thing? Sci Rep. 2018; 8(1): 5210. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu Z, Kuo PL, Horvath S, et al.: Phenotypic Age: a novel signature of mortality and morbidity risk. bioRxiv. 2018: 363291. Publisher Full Text\n\nMamoshina P, Kochetov K, Putin E, et al.: Population Specific Biomarkers of Human Aging: A Big Data Study Using South Korean, Canadian, and Eastern European Patient Populations. J Gerontol A Biol Sci Med Sci. 2018; 73(11): 1482–1490. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLevine ME, Lu AT, Quach A, et al.: An epigenetic biomarker of aging for lifespan and healthspan. Aging (Albany NY). 2018; 10(4): 573–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFeil R, Fraga MF: Epigenetics and the environment: emerging patterns and implications. Nat Rev Genet. 2012; 13(2): 97–109. PubMed Abstract | Publisher Full Text\n\nBelsky DW, Caspi A, Houts R, et al.: Quantification of biological aging in young adults. Proc Natl Acad Sci U S A. 2015; 112(30): E4104–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMontavon G, Samek W, Müller K-R: Methods for interpreting and understanding deep neural networks. Digital Signal Processing. 2018; 73: 1–15. Publisher Full Text\n\nLundberg SM, Nair B, Vavilala MS, et al.: Explainable machine-learning predictions for the prevention of hypoxaemia during surgery. Nat Biomed Eng. 2018; 2: 749–60. Publisher Full Text\n\nKelly C: cck197/ml-bio-age: Initial release (Version v1.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.2440203\n\nPetursson H, Sigurdsson JA, Bengtsson C, et al.: Is the use of cholesterol in mortality risk algorithms in clinical guidelines valid? Ten years prospective data from the Norwegian HUNT 2 study. J Eval Clin Pract. 2012; 18(1): 159–68. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKunutsor SK, Apekey TA, Seddoh D, et al.: Liver enzymes and risk of all-cause mortality in general populations: a systematic review and meta-analysis. Int J Epidemiol. 2014; 43(1): 187–201. PubMed Abstract | Publisher Full Text\n\nYi SW, Park S, Lee YH, et al.: Association between fasting glucose and all-cause mortality according to sex and age: a prospective cohort study. Sci Rep. 2017; 7(1): 8194. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBjørnholt JV, Erikssen G, Aaser E, et al.: Fasting blood glucose: an underestimated risk factor for cardiovascular death. Results from a 22-year follow-up of healthy nondiabetic men. Diabetes Care. 1999; 22(1): 45–9. PubMed Abstract | Publisher Full Text\n\nFulks M, Stout RL, Dolan VF: Albumin and all-cause mortality risk in insurance applicants. J Insur Med. 2010; 42(1): 11–7. PubMed Abstract\n\nZurauskaite G, Meier M, Voegeli A, et al.: Biological pathways underlying the association of red cell distribution width and adverse clinical outcome: Results of a prospective cohort study. PLoS One. 2018; 13(1): e0191280. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKadoglou NPE, Biddulph JP, Rafnsson SB, et al.: The association of ferritin with cardiovascular and all-cause mortality in community-dwellers: The English longitudinal study of ageing. PLoS One. 2017; 12(6): e0178994. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKolb H, Martin S: Environmental/lifestyle factors in the pathogenesis and prevention of type 2 diabetes. BMC Med. 2017; 15(1): 131. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKelley GA, Kelley KS, Roberts S, et al.: Comparison of aerobic exercise, diet or both on lipids and lipoproteins in adults: a meta-analysis of randomized controlled trials. Clin Nutr. 2012; 31(2): 156–67. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAslinia F, Mazza JJ, Yale SH: Megaloblastic anemia and other causes of macrocytosis. Clin Med Res. 2006; 4(3): 236–41. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBiagi E, Franceschi C, Rampelli S, et al.: Gut Microbiota and Extreme Longevity. Curr Biol. 2016; 26(11): 1480–5. PubMed Abstract | Publisher Full Text\n\nGovindaraju D, Atzmon G, Barzilai N: Genetics, lifestyle and longevity: Lessons from centenarians. Appl Transl Genom. 2015; 4: 23–32. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenjamin EJ, Blaha MJ, Chiuve SE, et al.: Heart Disease and Stroke Statistics-2017 Update: A Report From the American Heart Association. Circulation. 2017; 135(10): e146–e603. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBhupathiraju SN, Hu FB: Epidemiology of Obesity and Diabetes and Their Cardiovascular Complications. Circ Res. 2016; 118(11): 1723–35. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "42590",
"date": "15 Jan 2019",
"name": "Alex Zhavoronkov",
"expertise": [
"Reviewer Expertise aging research",
"machine learning"
],
"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\nWhile the study is not novel and the technical sophistication is considerably low, the study addresses one of the most important challenges in biomedicine and I recommend accepting it if the authors agree to make substantial improvements to the manuscript, try out the other machine learning methods, research the prior art, and expand the methodology.\nFirstly, the study does not provide an overview of the other interpretable biomarkers of aging developed using the multiple data types. Some of the prior clocks are described here: Zhavoronkov et al.1. It is very similar to the study published in 2016 (Putin et al.2) which not only introduced the concept but also provided a comparison with the other machine learning methods including GBM, RF, DT, LR, kNN, ElasticNet, SVM and DNNs and an online testing platform for the hematological aging clocks.\n\nAll of these machine learning methods allow for the various feature selection and feature importance techniques that provide very different results and pick the most important features differently. This paper explains the differences in how the different machine learning techniques prioritize different genes using the transcriptomic age predictor (Mamoshina et al.3). This is not a recommendation for citing these papers but an example of the work that needs to be done. As it stands, the study looks like a student machine learning data processing exercise and application of the out-of-the-box of python library on the NHANES dataset rather than a complete research paper. The conclusion that SHAP library is a good tool for interpreting the results from a machine learning model is not surprising at all. The paper can be hardly called a methodological paper because it lacks novelty of both methods of age prediction and comparison with classical methods of age prediction using a common blood test.\nThere is a number of issues I noticed that need to be addressed:\nThe paper is lacking the information on how the train and test set were selected along with the age by sex distribution. Was the training and optimization of models performed without cross-validation? At the same time, NHANES data also contains people with various conditions including diabetes and kidney disease. Were those individuals excluded from the training process? These important questions are not clear from the paper and need to be clarified.\n\nRelated to comment #1: how does the model perform on individuals with chronic diseases?\n\nIt is not clear why the predicted age is referred to as ‘biological age’. Biological age should be predictive of mortality. The observed difference between predicted and actual age should be associated with outcome in terms of morbidity or mortality. This should be explored in details with respect to the interpretation of the age predictor results. NHANES data has information about mortality that can be used for this type of analysis. At this point, the analysis suggests that selected blood parameters are associated with age and so predictive of chronological age. This type of analysis was performed in one of the referenced papers utilizing the NHANES dataset but not in this paper. It needs to be performed in order for the paper to be published.\n\nThe baseline is lacking. What would the performance be if you predict all samples as a median age for the population? Would it be higher or would it be the same as the test set error?\n\nIn line with the above comments, because the performance evaluation is not rigorous and no hyperparameter selection was performed, it is not clear why this age prediction method was selected. One of the commonly used and extensively validated models is Klemera and Doubal. (Klemera P, Doubal S. A new approach to the concept and computation of biological age 4). I would suggest exploring KD age prediction model in terms of interoperability of the blood test markers. Would be the machine learning model better? If so, why?\n\nAs mentioned above, there is no baseline model, comparison of different models or hyperparameters tuning. Without the interpretation of the difference between the predicted and actual chronological age (association with mortality or diseases for example), this difference is just an error of the model. How this error of the model would affect the results? Would the results change if the model is trained on samples that were initially predicted accurately? What about the samples predicted with a greater error? This need to be explored.\n\nRelated to the point, age distribution plots of those randomly selected samples are needed. How would different age groups contribute to the results?\n\nInstead of using k-fold cross-validation authors used just random 80/20 train/test split, so results presented at figure 2 (SHAP summary plots) cannot be interpreted as stable. E.g. for men the first 5 biomarkers are very similar in terms of importance for age prediction, so the order of these five biomarkers probably will be changed using different random data split.\n\nPreprocessing is rather scarce. E.g. outlier analysis was not provided. Were they excluded from the analysis? If not, why and how they would contribute the SHAP summary plots?\n\nI would like to see the comparison of the estimated reference ranges with commonly accepted reference ranges.\n\nA linear fit line on figure 1 is barely visible because dots and line are plotted using the same color\n\nIt is always a good practice to provide figures optimized color blind readers. Figure 2 colors are hardly distinguishable.\n\nFigure 3 is lacking the actual chronological age of the individual analyzed.\nMy recommendation is to address these points and explore the prior art. Biological age prediction using machine learning is a very interesting and important field and the studies need to be consistent and comparable.\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? 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? No",
"responses": []
},
{
"id": "43612",
"date": "12 Feb 2019",
"name": "Peter O. Fedichev",
"expertise": [
"Reviewer Expertise aging research",
"biomarkers of aging",
"theory of aging",
"aging therapeutics"
],
"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 concerns quantification of aging by means of biological age (BA) model trained as a predictor of chronological age from the widely available blood markers (complete blood cell counts and biochemistry). Understanding capabilities and the biology behind such biomarkers are among the key issues in fundamental aging studies and could be very helpful for practical applications.\n\nThe manuscript, however, falls short to provide the necessary characterization of the proposed BA model. I believe that the presentation could be improved by addressing the issues listed below so that the results of the study could be eventually indexed in a revised form.\n\nThe authors introduced and documented the performance of the particular ML pipeline (XGBoost flavor of decision tree algorithms) trained to predict the chronological age from the blood markers provided by the National Health and Nutrition Examination Survey (NHANES). The rationalizations behind the approach were two-fold. First, the biological age predictor could be (at least according to previous studies) associated with all-cause and disease-specific mortality. Second, the proposed algorithm could produce a better interpretation of the biological age model output in a form, eventually suitable for personalized recommendations.\n\nUnfortunately, the results presented in the manuscript are not sufficient to fully judge the merits of the model.\nMajor issues:\nThis is not the first work concerning the biological age estimation from the blood markers in general or in NHANES in particular. I would expect more references to previous work and different machine learning techniques (from principal components analysis to deep learning). I would take a log-linear mortality model from Levine 20181 and a deep learning model from Putin 20165 as state of the art modern implementations The results should be compared with a reference model. I would not expect anything sophisticated, but there must be a comparison. For example, would the novel XGBoost method perform better than a linear regression to chronological age? What is the correct measure of the model's performance? A biological age should not be judged by the quality of the chronological age prediction only. Iеt has been shown that improvements in the accuracy of this class of BA models may lead to a degradation of the association with chronic diseases and mortality (Levine 20181, Pyrkov 2018a2. The open access part of the NHANES database contains enough death events and clinical diagnosis. I propose to demonstrate how strongly the proposed BA is associated with the remaining lifespan (Cox-regression significance test )? Is there an association of the biological age (after adjustment for age and sex) with lifestyles (such as smoking, see Pyrkov 2018b3, Mamoshina 20194, etc). Are the effects of smoking reversible in cohorts of individuals, who quit smoking (see Pyrkov 2018b3)? What is the aging acceleration in years associated with smoking (see Mamoshina 20194)? How is it related to the actual lifespan depreciation associated with smoking? Is the biological age associated with chronic diseases? A linear model, such as a (regularized) regression to age, a log-linear proportional hazard model, would also provide the biological age estimation with contributions associated with the specific markers. Without comparison with a reference linear model, it would be difficult to argue that a more sophisticated approach is easier to interpret.\n\nLet me list a number of minor points, recommendations for the discussion (not necessarily calculations!):\nIt would be reasonable to discuss hyperparameters involved in the XGBoost model tuning. How those parameters were selected? There is a log-linear proportional hazard model predicting mortality in NHANES (Levine 20181). Is there a way to see if the XGBoost model is better? Is it possible to produce a prophetic statement? Could the authors speculate if their model is more or less statistically powerful than the phenoage? In the authors' opinion, what are advantages or disadvantages of XGBoost over deep learning models, such as Zhavoronkov? Is there a way to improve the biological age assessment with XGBoost in combination with proportional hazards models?\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?\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/8-17
|
https://f1000research.com/articles/8-16/v1
|
04 Jan 19
|
{
"type": "Research Article",
"title": "Correlation between luteinizing hormone receptor gene expression in human granulosa cells with oocyte quality in poor responder patients undergoing in vitro fertilization: A cross-sectional study",
"authors": [
"Budi Wiweko",
"M. Luky Satria",
"Kresna Mutia",
"Pritta Ameilia Iffanolida",
"Achmad Kemal Harzif",
"Gita Pratama",
"R. Muharam",
"Andon Hestiantoro",
"Pritta Ameilia Iffanolida",
"Achmad Kemal Harzif",
"Gita Pratama",
"R. Muharam",
"Andon Hestiantoro"
],
"abstract": "Background: This study was performed to evaluate the role of luteinizing hormone (LH) and granulosa cell LH receptor (LH-R) in poor responder patients who underwent controlled ovarian stimulation. Expression levels of LH-R mRNA in granulosa cells was investigated and compared with oocyte morphology, oocyte maturity and fertilization rate. Methods: Granulosa cells were obtained from 30 patients who underwent in vitro fertilization (IVF) at Dr. Cipto Mangunkusumo Hospital, Jakarta. The patients were divided into two groups: group I (n=10) poor responders; and group II (n=20) non-poor responders. After the extraction of total RNA from granulosa cells, semi-quantitative RT-PCR was performed and the amount of LH-R mRNA was quantified. The relative values were calculated as the ratio of LH-R mRNA and actin beta mRNA. Statistical analysis was performed using Mann-Whitney test and Spearman correlation. Results: The relative value of LH-R mRNA was higher in group I compared with group II (27.37[0.00-28939.37] vs 0.00[0.00-7196.12]). Oocyte maturity (r=0.267) and morphology (r=0.267) in group I consistently showed a positive correlation with LH-R mRNA; in group II a negative correlation with LH-R mRNA was shown for oocyte maturity (r= -0.552) and morphology (r= -0.164). Group I had a positive correlation between LH-R expression with fertilization rate (r=0.430), and group II showed a negative correlation (r=-0.340). Conclusions: The expression of LH-R mRNA has a positive correlation with oocyte quality in poor responder patients and a negative correlation in non-poor responders. Our study suggests an optimal expression of LH- R mRNA in granulosa cells during controlled ovarian stimulation to obtain good quality oocytes.",
"keywords": [
"Granulosa Cells",
"LH-Receptor",
"Oocytes",
"Poor Responder",
"qRT-PCR"
],
"content": "Introduction\n\nIn 2012, the World Health Organization reported that 80 million reproductive-aged couples, which constitutes 10% of the total number of couples globally, have issues related to infertility1. In Indonesia, 12–15% of reproductive-aged couples have infertility issues2,3. One way to manage infertility issues is using assisted reproductive technology (ART). One method of ART that is commonly used is in vitro fertilization (IVF). The In Vitro Fertilization World Report 2000 showed that the average number of pregnancy and births post-IVF is 26.7% and 18.6%, respectively4. According to a 2008 report, the percentage of pregnancy post-IVF in Indonesia is 20–52.9%5. As reported by the Society of Assisted Reproductive Technology, the success rate of IVF in women under 35 years of age is 41–43%6. However, this success rate declines as a woman’s age increases, especially for women who are older than 35 years old and those who are not adequately affected by gonadotropin treatment (also known as a poor responders)7.\n\nAccording to the The European Society of Human Reproduction and Embryology (ESHRE) 2011 consensus in Bologna, a poor responder is defined by the presence of 2 of 3 of the following criteria: (1) more than 40 years of age; (2) ovary reserve test such as the basal antral follicle count (BAF) <6-8 follicles/ovary, or level of anti-Müllerian hormone (AMH) <0.5-1.1 ng/mL; and (3) history of ovary stimulation producing < 3 oocytes8. In the USA, approximately 80.3% cancellations of an IVF cycle are caused by an inadequate number of eggs following ovarian stimulation9. Poor responders have a lower pregnancy rate compared with normal responders. Poor responders have a pregnancy rate varying from 7.6 to 17.5% compared with normal responders, varying from 25.9 to 36.7%10. The failure rate among the poor responder group is quite high, caused by the very low amount of oocytes and the low quality of oocytes, which eventually affects embryo quality. Low embryo quality will cause low implantation rate and high miscarriage rate11,12. If the poor responders eventually becomes pregnant, the risk of having pregnancy complications, such as hypertension and pre-eclampsia, increases13.\n\nLuteininzing hormone (LH) is an important glycoprotein hormone that regulates gonadal function that is subsequently involved in menstrual cycle physiology. LH works through LH receptor (LH-R), which are expressed in theca, granulosa, and cumulus cells. LH will bind to LH-R that is present in the cell membrane. Although the role of LH in the non-poor responder cycle is undisputed, the role of LH in ovarian stimulation during IVF is still debatable. LH supplementation for patients that respond positively toward gonadotropin releasing hormone (GnRH) agonists does not increase the number of pregnancies14. Several studies show an advantage of LH supplementation on poor responders who were using GnRH agonist15,16. In a study with Asian women who were using GnRH agonist, LH supplementation was recommended for poor responders during previous IVF cycles for slow follicle growth during days 6–8 of stimulation. This study also suggested LH supplementation for women at risk of suboptimal response, primarily those who are >35 years of age17. However, like several other studies, LH supplementation in this particular group of women did not significantly affect pregnancy outcomes17–19. For poor responders who were using the GnRH antagonist protocol, LH supplementation showed a better IVF outcome20–22. However, Konig et al. claimed no significant difference between LH supplementation for women > 35 years of age who were using the GnRH agonist protocol7. The difference between the effect due to administration of follicle stimulating hormone (FSH) and LH may be due to the difference in hormone receptor expression on oocyte cells that play a role in the maturation of the follicle22.\n\nA study by Humaidan et al. in 2002 reported that women who had LH levels of <0.5mlU/ml and >1.51mlU/ml on day 8 of stimulation had a lower implantation rate compared with women with LH levels between 0.5mlU/ml and 1.51 mlU/ml23. This particular study showed that LH has an optimal lower and upper threshold to reach adequate growth and maturation of egg cells. However, another study by Humaidan et al. in 2004 showed that women with LH level > 1.99 mlU/ml also had good results after LH supplementation24. This implies that there was an inadequate LH bioactivity; therefore, even though LH endogen level was within optimum range, it may not give an optimum effect. Alviggi et al. further suggested that an LH polymorphism (v-bethaLH) resulted in a group of women showing inadequate response after FSH administration, despite having enough ovarian reserve25.\n\nMany theories have tried to explain the etiology of poor ovarian response towards gonadotropin administration. Age and low ovarian reserve are the most common factors used to explain the presence of a poor responder patient group. However, some poor responders are still young, thus the etiology of poor responders seems to be multifactorial, including decreased blood flow to the ovary, decreased aromatase activity, FSH and LH receptor polymorphisms, and autoimmunity towards the ovary26. Understanding LH-R gene expression in humans is important to increase the success rate of IVF. The importance of LH during follicular phase and the optimum dosage of LH for IVF patients are still debatable27. Genetic studies have had important roles in understanding the pathogenesis of diseases and development of therapy. By understanding genetic studies with a focus on gene and gene product, IVF specialists can decide the appropriate therapy for patients18. Therefore, this study investigates the correlation between LH-R granulosa expression in poor responder patients and non-poor responder patients who are going through an IVF program and compares the oocyte quality outcome, fertilization rate, and pregnancy rate.\n\n\nMethods\n\nThis cross-sectional study was conducted to find the correlation between LH-receptor gene expression in granulosa cells with oocyte quality in poor responder patients undergoing IVF. The study took place at Yasmin Clinic, Dr. Cipto Mangunkusumo Hospital, Jakarta, Indonesia, between January and June 2015. This study was approved by the Ethics Committee of the Faculty of Medicine, Universitas Indonesia (now called the Health Research Ethics Committee, Universitas Indonesia and Dr. Cipto Mangunkusumo Hospital (HREC-FMUI/CMH) (approval number, 631/UN2.F1/ETIK/2014).\n\nWomen attending the Yasmin Clinic for IVF procedures were selected according to Bologna criteria based on anamnesis, ultrasound, and laboratory examination8. The patients underwent ovarian stimulation, which is part of the IVF procedure, continuing to the ovum pick-up (OPU) procedure. Prior to OPU, patients were offered to participate in the study. An explanation of the research, objectives, procedures, benefits, risks and expected study outcomes were provided, along with an informed consent form. The subjects who were willing to participate in the study were asked to sign the consent form. Patients with incomplete baseline data and who failed the OPU procedure were not included in this study. In total, 30 patients were recruited from January to June 2015 in the Yasmin IVF Clinic Dr. Cipto Mangunkusumo General Hospital, Jakarta. The patients were divided into two groups: I poor responders (n=10); and II non-poor responders (n=20). The poor responder group (group I) matched minimum 2 of these following criteria: (1) more than 40 years of age; (2) ovary reserve test such as the basal antral follicle count (BAF) <6-8 follicles/ovary, or level of anti-Müllerian hormone (AMH) <0.5-1.1 ng/mL; and (3) history of ovary stimulation producing < 3 oocytes. Patients who do not have those criteria, entered into group II.\n\nAt the time of OPU, the doctor extracted intrafolicular fluid from the patient under anesthesia. The intrafolicular fluid containing oocytes and granulosa cells was processed by the embryologist, and then the oocyte is processed in the next stage of IVF. The granulosa cells were stored at -20°C before use.\n\nThe RNA of the granulosa cells was extracted using High Pure RNA Isolation Kit (Roche, Mannheim, Germany). Measurement of total RNA concentration was made using NanoVue spectrophotometer (General Electric). Subsequently, cDNA synthesis was performed. The positive control used was mRNA from the Transcriptor First Strand cDNA synthesis kit (Roche). A light cycler fast start DNA MasterPLUS SYBR green I (Roche) was used for real time polymerase chain reactions (PCR), with the following profile: pre-incubatiton (1 cycle at 95°C for 10 minutes), quantification (45 cycles each at 95°C for 10 seconds, 65°C for 10 seconds and 72°C for 25 seconds), melting curve (1 cycle each at 95°C for 0 seconds, 65°C for 60 seconds and 95°C for 0 seconds). All the procedures are according to the manufacturer’s instruction.\n\nIn this study, we used one microgram of complementary DNA (cDNA) per reaction in a 10 microliter reaction volume. Beta-actin RNA was chosen as a suitable nonpoorization control gene. LH-R gene quantification was done using Light Cycler Fast Start DNA MasterPLUS SYBR green I kit (Roche). The real time PCR was performed using Light- Cycler 2.0 Instrument (03531414001, Roche). Primer sequences can be seen in Table 1. Output data of qRT-PCR were used to calculate the ratio of gene LH-R is the value of delta Rn. The value is the result of fluorescence detection by qRT-PCR machine and translated by using LightCycler Software Version 4.1 (Roche). The Rn value was stored in file comma delimited file (CSV) using the Kingsoft Spreadsheet program Version 2013. The data were then processed using the R Studio program Version 2.11.1, which had been added to the qpcR software package. The result of script application is the ratio value, Cp, and the efficiency of each reaction.\n\nFor the oocyte morphology data, we used modification scoring system based on the Xia criteria, including first polar body, perivitelline space and cytoplasmic granulation (Table 2)28. Morphology scores were measured to all oocyte retrieved with minimum score of 0 and maximum scores of 6. Mean value of morphology scores per patient were used for analysis. The percentage of fertility rate were measured from the total of fertilized divided by total embryos.\n\nData of expression of LH-R gene, oocyte maturity, morphology, and fertilization rate were analyzed. The normality of the data was tested using Shapiro-Wilk test. Normally distributed data were then tested with the unpaired T-test, but if the data had an abnormal distribution, the Mann-Whitney test was performed to compare the differences in expression of LH-R gene between groups of poor responder and non-poor responder patients. Furthermore, the Spearman’s test was performed to determine the correlation between LH-R gene expression and oocyte maturity, morphology, and fertilization rate. A p-value of less than 0.05 was considered significant. Statistical analysis was performed using IBM SPSS (Statistical Package for Social Sciences) version 22.\n\n\nResults\n\nTable 3 shows the characteristics of groups I and group II, poor and non-poor responders, respectively. We found that granulosa LH-R expression is higher in the poor responders (27.37 (0.00-28939.37) arbitrary unit) than in non-poor responders (0.00 (0.00-7196.12)), but this was not statistically significant between the groups (p=0.169).\n\nTable 4 shows a statistically significant negative correlation between oocyte maturity in the poor responders and granulosa LH-R expression (p=0.003; r=0.552). There is no correlation between granulosa LH- R expression and oocyte morphology. There is negative correlation between granulosa LH-R expression and fertility rate in the poor responders (p=0.215; r=0.430) and non-poor responders (p=0.142; r=-0.340), which was not statistically significant.\n\n\nDiscussion\n\nFrom this study, granulosa LH-R expression in poor responders is higher than in non-poor responders, although it is not statistically significant. In a study by Thiruppathi et al.29, which compared gonadotrophin receptor expression in poor responders and non-poor responders, the results also showed that granulosa LH-R expression is higher in poor responders. This may suggest that there is a disruption in LH-R processing and trafficking or it could be caused by accelerated release of LH in poor responders29.\n\nIn this study, granulosa LH-R expression had a positive correlation with morphology, oocyte maturity and fertility rate in the poor responder group and a negative correlation in non-poor responder group. This showed that the poor responder group needs LH for oocyte growth and maturation, while in the non-poor responder group high granulosa LH-R expression would affect maturity and morphology of oocytes and fertility rate. Maman et al.18 found that in non-poor responders, granulosa LH-R expression increased in the antral-phase follicle and the highest expression happened pre-ovulation. There was a correlation between LH-R expression and fertility output. In that study, low LH-R expression correlated with low oocyte maturity, but excessive LH-R expression correlated with a low fertility rate18. From these studies, it seems that an optimal granulosa LH-R expression is needed to mature oocyte and yield a good fertility rate.\n\nA literature study by Shoham19 suggested that there is a therapeutic window with a threshold and ceiling in LH supplementation. If the LH level is below the needed ceiling, estradiol production will not be adequate, while if the LH level is higher than the threshold, there will be a negative impact on follicle growth. Shoham discovered that LH supplementation in patients with hypogonadotropic hypogonadism would generate more follicles and adequate estradiol levels to generate good endometrium growth. However, excessive LH supplementation in patients with hypogonadotropic hypogonadism or polycystic ovaries would cause negative effects, causing follicles to become atretic19. The study is supported by Humaidan et al.29, who found that LH level must be at an adequate level, not too high or too low, to generate good quality oocytes. Optimal LH levels measured in the 8th day of stimulation will decrease the required FSH doses, hence making the stimulation duration shorter and growth of good follicles faster29. However, it must be remembered that this study was done in non-poor responder patients, using a GnRH agonist protocol, where patients’ gonadotropin is stimulated with GnRH agonist administration before stimulation. Therefore, if endogenous LH levels decrease too much because of that suppression, follicular growth will be disrupted, because the androgen production in theca cells, which will be converted into estradiol in granulosa cells, is decreasing. Because of that, non-poor responder patients who receive an agonist GnRH protocol should use an adequate dose that is not too high. Besides the dose, the administration mode also influences mid-follicular LH levels. In the administration of intranasal buserelin, the decrease of mid-follicular LH is not too low, and the pregnancy rate is better compared to subcutaneous administration29.\n\nIn the present research, all subjects received the antagonist protocol, so there was no excessive endogenous LH suppression as seen in the agonist protocol. It seems that the stimulation protocol type may not influence granulosa LH-R expression. This is supported by a microarray analysis of gene expression study in rFSH and hMG stimulation, where LH-R expression showed no differences30. Therefore, it may be inferred that the main contributors of granulosa LH-R expression are follicle size and follicle maturation stage18. Granulosa LH-R is expressed in the early antral phase of follicle growth even when granulosa LH-R expression is still very low18,31. Then, by increasing follicle maturation under FSH influence, granulosa LH-R expression will also increase. Therefore, granulosa LH-R expression can be increased by exposing granulosa to adequate FSH before18. However, how much FSH to get optimal LH-R expression must be investigated. The most common problem in poor responder patients, other than the inadequate number and bad quality of oocytes produced, is that the FSH dose needed for stimulation is too high. This has been shown in the present study (Table 2). In this study, the total FSH dose used in the poor responder group was higher than that in the nonpoor responder group (3660 IU vs 2822 IU). In the Alviggi study, a higher FSH dose was also found in polymorphism LH patients (v-betaLH). According to Alviggi, homozygote and heterozygote v-betaLH patients have a poor ovarian response to gonadotropin and the need of total FSH dose is higher than in wild-type LH patients. The oocytes are also generated in a smaller number. In his hypotheses, Alviggi stated that it was caused by the difference in the bioactive effects between v-betaLH and wild-type LH25,32. v-betaLH has shorter t-half but has a more potent efficacy at receptor level compared to the wild-type25. In the present study, analysis of v-beta LH was not done. Besides LH polymorphism, in the Alviggi study, there was also a patient who did not have v-betaLH but had an inadequate response to gonadotropin. The possible explanation of that phenomenon is due to the increment in FSH consumption without the presence of v-betaLH, which is caused by LH-R and FSHR polymorphisms32. However, this still needs to be further investigated.\n\n\nConclusions\n\nThis study showed that granulosa LH-R expression in poor responders is higher than non-poor responders. Statistical analysis showed a positive correlation between granulosa LH-R expression with oocyte quality and fertility rate in the poor responders, and a negative correlation between granulosa LH-R expression with oocyte quality and fertility rate in the nonpoor responders.\n\n\nData availability\n\nUnderlying data F1000Research: Dataset 1. Raw data for all variables reported in the study, https://doi.org/10.5256/f1000research.17036.d23085933",
"appendix": "Grant information\n\nThis study received a research from Research Center Revitalization Grant, Universitas Indonesia 2014.\n\n\nAcknowledgements\n\nThe authors thank Indonesian Reproductive Medicine Research and Training Center (INAREPROMED) and Yasmin Clinic Rumah Sakit Dr. Cipto Mangunkusumo Hospital teams for assistance and support in this study.\n\n\nReferences\n\nOrganization WH: Infertility. 2012. Reference Source\n\nRespati G: Keberhasilan progran fertilisasi in vitro di klinik infertilitas FK UNDIP-RS Dr Kariadi dan RS Telogorejo Semarang. Semarang: Universitas Diponegoro Semarang; 2005. Reference Source\n\nSoebijanto S: Pengaruh beberapa kondisi ginekologis/klinis infeksi TORSH-KM lampau dan kondisi spermatozoa pada fertilisasi in vitro. universitas Indonesia; 1993. Reference Source\n\nInternational Committee for Monitoring Assisted Reproductive Technology, Adamson GD, de Mouzon J, et al.: World collaborative report on in vitro fertilization, 2000. Fertil Steril. 2006; 85(6): 1586–622. PubMed Abstract | Publisher Full Text\n\nPERFITRI: Indonesian Association for In Vitro Fertization’s National Report 2017. IAIVF; 2018. Reference Source\n\nInformation NCFB: In vitro fertilization 2012.\n\nKonig TE, van der Houwen LE, Overbeek A, et al.: Recombinant LH supplementation to a standard GnRH antagonist protocol in women of 35 years or older undergoing IVF/ICSI: a randomized controlled multicentre study. Hum Reprod. 2013; 28(10): 2804–12. PubMed Abstract | Publisher Full Text\n\nFerraretti AP, La Marca A, Fauser BC, et al.: ESHRE consensus on the definition of ’poor response’ to ovarian stimulation for in vitro fertilization: the Bologna criteria. Hum Reprod. 2011; 26(7): 1616–24. PubMed Abstract | Publisher Full Text\n\nPromotion NCoCDPaH: Assisted Reproductive Technology, National Summary Report 2010. Reference Source\n\nOudendijk JF, Yarde F, Eijkemans MJ, et al.: The poor responder in IVF: is the prognosis always poor?: a systematic review. Hum Reprod Update. 2012; 18(1): 1–11. PubMed Abstract | Publisher Full Text\n\nLiu K, Case A, REPRODUCTIVE ENDOCRINOLOGY AND INFERTILITY COMMITTEE: Advanced reproductive age and fertility. J Obstet Gynaecol Can. 2011; 33(11): 1165–75. PubMed Abstract | Publisher Full Text\n\nLiu KE, Case A: No. 346-Advanced Reproductive Age and Fertility. J Obstet Gynaecol Can. 2017; 39(8): 685–95. PubMed Abstract | Publisher Full Text\n\nvan Disseldorp J, Eijkemans R, Fauser B, et al.: Hypertensive pregnancy complications in poor and normal responders after in vitro fertilization. Fertil Steril. 2010; 93(2): 652–7. PubMed Abstract | Publisher Full Text\n\nFerraretti AP, Gianaroli L, Magli MC, et al.: Exogenous luteinizing hormone in controlled ovarian hyperstimulation for assisted reproduction techniques. Fertil Steril. 2004; 82(6): 1521–6. PubMed Abstract | Publisher Full Text\n\nDe Placido G, Alviggi C, Perino A, et al.: Recombinant human LH supplementation versus recombinant human FSH (rFSH) step-up protocol during controlled ovarian stimulation in normogonadotrophic women with initial inadequate ovarian response to rFSH. A multicentre, prospective, randomized controlled trial. Hum Reprod. 2005; 20(2): 390–6. PubMed Abstract | Publisher Full Text\n\nWong PC, Qiao J, Ho C, et al.: Current opinion on use of luteinizing hormone supplementation in assisted reproduction therapy: an Asian perspective. Reprod Biomed Online. 2011; 23(1): 81–90. PubMed Abstract | Publisher Full Text\n\nCédrin-Durnerin I, Grange-Dujardin D, Laffy A, et al.: Recombinant human LH supplementation during GnRH antagonist administration in IVF/ICSI cycles: a prospective randomized study. Hum Reprod. 2004; 19(9): 1979–84. PubMed Abstract | Publisher Full Text\n\nMaman E, Yung Y, Kedem A, et al.: High expression of luteinizing hormone receptors messenger RNA by human cumulus granulosa cells is in correlation with decreased fertilization. Fertil Steril. 2012; 97(3): 592–8. PubMed Abstract | Publisher Full Text\n\nShoham Z: The clinical therapeutic window for luteinizing hormone in controlled ovarian stimulation. Fertil Steril. 2002; 77(6): 1170–7. PubMed Abstract | Publisher Full Text\n\nMarrs R, Meldrum D, Muasher S, et al.: Randomized trial to compare the effect of recombinant human FSH (follitropin alfa) with or without recombinant human LH in women undergoing assisted reproduction treatment. Reprod Biomed Online. 2004; 8(2): 175–82. PubMed Abstract | Publisher Full Text\n\nDe Placido G, Mollo A, Clarizia R, et al.: Gonadotropin-releasing hormone (GnRH) antagonist plus recombinant luteinizing hormone vs. a standard GnRH agonist short protocol in patients at risk for poor ovarian response. Fertil Steril. 2006; 85(1): 247–50. PubMed Abstract | Publisher Full Text\n\nHill MJ, Levens ED, Levy G, et al.: The use of recombinant luteinizing hormone in patients undergoing assisted reproductive techniques with advanced reproductive age: a systematic review and meta-analysis. Fertil Steril. 2012; 97(5): 1108–14.e1. PubMed Abstract | Publisher Full Text\n\nHumaidan P, Bungum L, Bungum M, et al.: Ovarian response and pregnancy outcome related to mid-follicular LH levels in women undergoing assisted reproduction with GnRH agonist down-regulation and recombinant FSH stimulation. Hum Reprod. 2002; 17(8): 2016–21. PubMed Abstract | Publisher Full Text\n\nHumaidan P, Bungum M, Bungum L, et al.: Effects of recombinant LH supplementation in women undergoing assisted reproduction with GnRH agonist down-regulation and stimulation with recombinant FSH: an opening study. Reprod Biomed Online. 2004; 8(6): 635–43. PubMed Abstract | Publisher Full Text\n\nAlviggi C, Pettersson K, Longobardi S, et al.: A common polymorphic allele of the LH beta-subunit gene is associated with higher exogenous FSH consumption during controlled ovarian stimulation for assisted reproductive technology. Reprod Biol Endocrinol. 2013; 11: 51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLuborsky JL, Thiruppathi P, Rivnay B, et al.: Evidence for different aetiologies of low estradiol response to FSH: age-related accelerated luteinization of follicles or presence of ovarian autoantibodies. Hum Reprod. 2002; 17(10): 2641–9. PubMed Abstract | Publisher Full Text\n\nPapamentzelopoulou M, Mavrogianni D, Partsinevelos GA, et al.: LH receptor gene expression in cumulus cells in women entering an ART program. J Assist Reprod Genet. 2012; 29(5): 409–16. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXia P: Intracytoplasmic sperm injection: correlation of oocyte grade based on polar body, perivitelline space and cytoplasmic inclusions with fertilization rate and embryo quality. Hum Reprod. 1997; 12(8): 1750–5. PubMed Abstract | Publisher Full Text\n\nThiruppathi P, Shatavi S, Dias JA, et al.: Gonadotrophin receptor expression on human granulosa cells of low and normal responders to FSH. Mol Hum Reprod. 2001; 7(8): 697–704. PubMed Abstract | Publisher Full Text\n\nBrannian J, Eyster K, Mueller BA, et al.: Differential gene expression in human granulosa cells from recombinant FSH versus human menopausal gonadotropin ovarian stimulation protocols. Reprod Biol Endocrinol. 2010; 8: 25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJeppesen JV, Kristensen SG, Nielsen ME, et al.: LH-receptor gene expression in human granulosa and cumulus cells from antral and preovulatory follicles. J Clin Endocrinol Metab. 2012; 97(8): E1524–31. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlviggi C, Clarizia R, Pettersson K, et al.: Suboptimal response to GnRHa long protocol is associated with a common LH polymorphism. Reprod Biomed Online. 2009; 18(1): 9–14. PubMed Abstract | Publisher Full Text\n\nWiweko B, Satria ML, Mutia K, et al.: Dataset 1 in: Correlation between luteinizing hormone receptor gene expression in human granulosa cells with oocyte quality in poor responder patients undergoing in vitro fertilization: A cross-sectional study. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.17036.d230859"
}
|
[
{
"id": "46201",
"date": "01 Apr 2019",
"name": "Yuval Yung",
"expertise": [
"Reviewer Expertise I'm researcher in the field of reproduction and embryolog."
],
"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 by Wiweko et al. asks, does the LHR expression levels in mural granulosa cells correlate with embryo quality in poor/non-poor responder patients?\nTheir results show a positive correlation between fertilization and embryo morphology in poor responder patients and a negative correlation between fertilization and embryo morphology in non-poor responder patients. These findings are novel and are well presented.\nOne point of interest is missing in the discussion; LHR undergoes a sharp decrease after the LH surge (Nair et al., 20021 and Ophir et al., 20142) and the granulosa cells examined in this study were obtained around 34 h post LH surge when LHR levels should be under suppression. This may explain the lower LHR levels found in non-poor responders that responded better to the LH surge. Further discussion in this regard should be added.\n\nThe manuscript after minor revision may be suitable for indexing.\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": "46684",
"date": "29 Apr 2019",
"name": "Yanping Kuang",
"expertise": [
"Reviewer Expertise Endocrine disorder",
"Assisted reproduction technology"
],
"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 was to evaluate the role of LH and granulosa cell LH-R in poor responder patients who underwent COS. Then the authors analyzed the expression of LH-R in the granular cells from poor responder and non-poor responders and found the expression of LH-R mRNA had a positive correlation with oocyte quality in poor responder patients and a negative correlation in non-poor responders. The aim of this study was vague. The conclusion was not consistent with the purpose of this study and made with insufficient supporting data. The correlations between the LH-R expression and oocyte quality was established based on Spearman’s test analysis. However, the oocyte quality might be correlated with age, BMI, basal FSH level, basal LH level and other factors. To analyse the correlation between the expression of LH-R and oocyte quality, multivariate logistic regression should be performed to quantify the effect of all related factors on the oocyte quality.\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? No",
"responses": []
},
{
"id": "47813",
"date": "17 May 2019",
"name": "Peng-Hui Wang",
"expertise": [
"Reviewer Expertise Gynaecologic Oncology",
"sialylation"
],
"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 provided a cross section study to evaluate the predictor of oocyte quality 30 poor responders undergoing in vitro fertilization (IVF) and found the expression of luteinizing hormone receptor (LHR) was important to predict the oocyte quality of poor responders or non-poor responders, but the former is a positive correlation and the latter is a negative correlation. The current study showed oocyte quality could be predicted by oocyte-surrounding granulosa cells. Although the study is interesting, there are some issues worthy of further discussion.\nFor the definition of poor responders, it is very clear. The key factor of the criteria is a poor ovarian reservation (Chern et al., 20181, Li et al., 20182, Tsui et al., 20173, Lin et al., 20174, Lin et al., 20175, Lin et al., 20176, Lin et al., 20157, Tsui et al., 20158, Tsui et al., 20149). Even though there was an adequate protocol for ovary stimulation, the total number of the oocytes retrieved is limited. By contrast, high responders are often risky for the development of ovary hyperstimulation syndrome, and the total number of oocytes retrieved is too many. In clinical practice, polycystic ovary syndrome (PCO) might be one of the best examples. Although the ratio of LH/FSH is abandoned in the diagnostic criteria of PCO, we can predict the serum level of LH is higher in the high responders, which might be secondary to the relative “hyperandrogenism”. Although the authors used the non-poor responders as control, it is not clear that the similar finding could be applied in high responders.\n\nThe study material of granulosa cells might be interesting. Is there any difference of LHR expression between different populations of granulosa cells (oocyte surrounding granulosa cell (cumulus cells) or follicle-type granulosa cells)? (Lin et al., 20174, Tsui et al., 20149).\n\nSince the significant difference of dosage of LH trigger between poor responders and non-poor responders was noted, it is rational to suppose the LHR expression will be different between two groups (Lin et al., 20176). The correlation between the ligand and receptor should be introduced in much more detail. Membrane-type receptor and nuclear-type receptor expression might be varied greatly or conflicted dramatically when the different dosage of ligands is provided. Down regulation and overexpression might be the secondary effect. In addition, if the oocyte quality is positively correlated with LHR expression in poor responders, the absolute qualification of the amount of LHR might be needed. It is relatively unusual that the same amount of the hormone profiles will result in the different clinical patterns. If this condition is real, how would the normal range of hormone profile in the clinical practice be obtained? The authors should consider it with much more concern.\n\nOocyte quality is often recognized by morphological classification. If more markers could be used in much more scientific or quantitative patterns, it is welcome. However, the easy-to-use rule and reproducibility might be important.\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": []
}
] | 1
|
https://f1000research.com/articles/8-16
|
https://f1000research.com/articles/7-686/v1
|
31 May 18
|
{
"type": "Research Article",
"title": "Challenges in the use of atomistic simulations to predict solubilities of drug-like molecules",
"authors": [
"Guilherme Duarte Ramos Matos",
"David L. Mobley",
"Guilherme Duarte Ramos Matos"
],
"abstract": "Background: Solubility is a physical property of high importance to the pharmaceutical industry, the prediction of which for potential drugs has so far been a hard task. We attempted to predict the solubility of acetylsalicylic acid (ASA) by estimating the absolute chemical potentials of its most stable polymorph and of solutions with different concentrations of the drug molecule. Methods: Chemical potentials were estimated from all-atom molecular dynamics simulations. We used the Einstein molecule method (EMM) to predict the absolute chemical potential of the solid and solvation free energy calculations to predict the excess chemical potentials of the liquid-phase systems. Results: Reliable estimations of the chemical potentials for the solid and for a single ASA molecule using the EMM required an extremely large number of intermediate states for the free energy calculations, meaning that the calculations were extremely demanding computationally. Despite the computational cost, however, the computed value did not agree well with the experimental value, potentially due to limitations with the underlying energy model. Perhaps better values could be obtained with a better energy model; however, it seems likely computational cost may remain a limiting factor for use of this particular approach to solubility estimation.\n\nConclusions: Solubility prediction of drug-like solids remains computationally challenging, and it appears that both the underlying energy model and the computational approach applied may need improvement before the approach is suitable for routine use.",
"keywords": [
"solubility",
"molecular crystals",
"free energy calculations",
"chemical potentials",
"solvation"
],
"content": "Introduction\n\nSolubility is a critical property for pharmaceutical drug discovery; problems with solubility can frustrate drug discovery efforts and prevent treatments from working. The bioavailability of a drug depends on the solubility difference between different crystal structures (polymorphs), dose, drug permeability and formulation1, so solubility plays a key role. Solubility problems can be unexpected and can pose crucial obstacles that even threaten the administration of care. For example, a well-documented case occurred in the late 1990s, when ritonavir, an HIV-protease inhibitor marketed as Norvir, failed dissolution requirements2. Since ritonavir is not bioavailable in its solid form, it was administrated in capsules containing solutions designed not to be saturated with respect to the originally known molecular crystal (form I)2. The newly identified polymorph, form II, was unusually stable and unusually hard to crystallize; the preparation protocol of Norvir was inadequate to make capsules from the new polymorph, which severely threatened the supply of the drug in the market and endangered the lives of many HIV-poaitive patients2. Considerable effort has already been devoted to the methods to predict crystal polymorphs3–9, but much less attention has been given to methods to predict solubilities, with or without likely polymorphs as input.\n\nDue to the importance of aqueous solubilities in different industrial processes and environmental applications, a scientific challenge, consisting of the prediction of 32 solubilities given a database of 100 reliable measurements10,11, was created with the goal of comparing the outcomes of different solubility prediction techniques. Participants employed methods such as artificial neural networks12, quantitative structure-property relationship (QSPR)13, and deep learning14 to predict the aqueous solubilities of drug-like molecules. All of the employed methods were empirical and trained on existing measurements. The limitation of these methods, however, is the dependence of a training set of data that limits their applicability to compounds similar to those in the training of the set and impairs its transferability.\n\nSome newer methods attempt to predict solubilities based on a physical description of the interactions in solution and in the solid state, yielding results that are in principle rigorous given an accurate energy model and an adequate method. In these approaches, molecular systems are described using force fields, i.e, potential energy functions that contain parameters describing bonds, atoms, electrostatic and non-electrostatic interactions. Molecular dynamics or Monte Carlo simulations are commonly used to sample different configurations of the system described by an energy model called a force field. The simulations then allow the estimation of physical properties, such as internal energy, free energy, and enthalpy, under different conditions. The quality of the results of such methods depends on how well the force field describes the system under study and how good the sampling method is. Thus, some researchers have recently estimated aqueous solubilities using simulations of thermodynamic cycles encompassing the crystal, the ideal gas, and an infinitely dilute solution of a given molecule15,16. When the structure of the solid is unknown, some studies have substituted simulations of solid melts in place of a structure of the solid17–20.\n\nWhile these physical methods for predicting solubilities have received some attention in the literature, most are still in their infancy, with only a handful of studies applying them, and it is not yet clear how broadly applicable they will be17–20, and others have only been suggested or demonstrated in proof-of-principle tests16,21–23. Our view is that the time is ripe for physical methods to predict solubility, especially given the routine nature of solvation free energy calculations at present24–29, which comprise essentially half of the solubility problem (see the Theory section). Polymorph and crystal structure prediction successes also mean that we may often have a suitable crystal structure of the compound as an input3–5,8,9,30–35, so what remains is to predict the solubility given a crystal structure and simulations of the relevant phases.\n\nHere, we focus on adapting and testing an existing approach for solubility prediction in the hope that it will prove to be a generally applicable method for solubility prediction that can be applied routinely. This method uses all-atom molecular dynamics simulations to estimate absolute chemical potentials and predict aqueous solubilities of molecular solids, given the crystal structure (or an estimate thereof) as input.\n\n\nTheory\n\nSolubility is defined as the maximum concentration of solute that can be dissolved in a selected bulk solvent. Chemical potentials (µ) of the solid-state solute and the solution are by definition equal at the solubility point, when the solution is in equilibrium with the solid.\n\n\n\nSolid particles precipitate in concentrations higher than the solubility point because the solid phase becomes more stable in these conditions. In principle, we can predict at which concentration a molecule precipitates in solution if we calculate the chemical potentials of the components:\n\n\n\nwhere µi is the chemical potential of component i; A is the Helmholtz free energy; G is the Gibbs free energy; Nj, j≠i is the number of molecules of each component in the mixture; V is the volume of the system; T its temperature; and P its pressure. Calculations from systems under a constant V and T yield A; G is obtained from simulations under constant P and T conditions. In order to estimate the chemical potential of one component in solution and in its molecular solid, however, we need to know the absolute free energy of the system in these states. We calculated absolute free energies using alchemical free energy calculations.\n\nThe absolute free energy of a system can be determined if we know its partition function (Q), a function that connects microscopic properties of the system with macroscopic thermodynamic quantities. Unfortunately, it is very hard to calculate the absolute free energy of real systems because we don’t know their partition functions. Free energy calculations allow us to bypass this problem, but require at least two states: a reference state whose free energy can be analytically or numerically found, and a final state of interest36,37. We chose to calculate the free energy difference using alchemical free energy calculations, a method in which we simulate a series of non-physical intermediates between the end states38.\n\nEach intermediate state in the alchemical path is described by a Hamiltonian ℋ (q, p; λ), i.e, the energy of the state as a function of atomic positions (q), momenta (p) and a coupling parameter (λ):\n\n\n\nwhere ℋinitial and ℋfinal respectively are the Hamiltonians of the initial and the final state; and f (λ) and g(λ) are functions used to mix the Hamiltonians, and are usually set such that ℋ = ℋinitial at λ = 0 and ℋ = ℋfinal at λ = 1.\n\nA variety of different estimators can be used to analyze alchemical free energy calculations, and have different strengths and weaknesses, as well as different data requirements. Here, we employ several different estimators we introduce briefly in the following.\n\nOne way to calculate the free energy difference (∆A) between the end states is Thermodynamic Integration (TI)39:\n\n\n\nin which a set of discrete λ values correspond to states along the alchemical path. 〈〉 means that we are have to calculate the ensemble average of the derivative between the brackets. TI performs as well as more efficient methods if the integrand is smooth, but breaks down if this condition is not satisfied40–42.\n\nAn alternate free energy estimation method computes ∆A directly via:\n\n\n\nwhere the ensemble average is calculated over the configurations of the initial state, and β is the reciprocal of kBT, the product between the Boltzmann constant and the absolute temperature. We call this approach exponential averaging43 (EXP).\n\nMost free energy calculations involve many intermediates associated with the coupling parameter (λ), allowing simulation of intermediate states in between the two end states of interest. The free energy change between the end points of a path defined by N intermediates is:\n\n\n\nwhere ∆An→n+1 is the free energy difference between (n+1)-th and the n-th intermediate states. Equation 5 can be used to calculate the free energy difference between each adjacent pair of states and yields the exact result at the limit of very large samples, but it is inefficient for most applications38.\n\nThe Bennett acceptance ratio44 (BAR) provides an estimator that is superior for most purposes. It calculates the free energy difference between the n-th and the (n + 1)-th states from the following relationship:\n\n\n\nwhere Nn and Nn+1 are the number of statistically independent samples in states n and n + 1, respectively, and ∆ℋn→n+1 = −∆ℋn+1→n are the Hamiltonian differences between n and n + 1. BAR is more efficient than EXP45,46 and minimizes the free energy uncertainty44. Multistate Bennett acceptance ratio41 (MBAR) is an extension of BAR that takes in consideration the degree of configuration space overlap between a given state and all other states in the transformation, whereas BAR only uses the information of neighboring states. MBAR and BAR perform similarly when the spacing between the intermediate states is moderate, but MBAR is the most well-performing free energy estimator42.\n\nIn this work, we seek to predict the solubilities of molecular solids. Part of this problem requires predicting the free energy or chemical potential of the solid. One way this has been attempted in the past is via the Einstein crystal method (ECM), which calculates the absolute free energy of a solid using an Einstein crystal as a reference state. In this method, the crystal lattice is made of atoms restrained to their positions by a harmonic potential; additionally, the center of mass of the system is held fixed47.\n\nIn the ECM, and in this work, the absolute free energy of the molecular solid is found by designing a path where force field terms are progressively turned on, and the harmonic potential position restraints are turned off. The fixed center of mass is important to avoid a quasi-divergence issue when calculating the free energy term of releasing the system from the harmonic position restraints, but the contribution of the fixed center of mass needs to be included in the cycle to obtain the correct absolute free energy for the system (Figure 1(a))47–49.\n\n(a) Thermodynamic cycle representing the Einstein Crystal Method. (b) Thermodynamic cycle representing the Einstein molecule method (EMM). Note that the EMM requires only two free energy calculations despite being a bigger thermodynamic cycle. The canceling terms in (b) correspond to the free energies of fixing and releasing one atom in the crystal lattice48.\n\nIn ECM, the free energy is calculated by:\n\n\n\nwhere AFCMEC is the free energy of the Einstein crystal (EC) with a fixed center of mass (FCM); ∆AEC→IEC is the free energy difference between the Einstein crystal (EC) and the interacting Einstein crystal (IEC), i.e., the free energy difference in a transformation where the force field is progressively turned on throughout the calculation path. ∆AIEC→SFCM is the free energy difference between the IEC and the solid with a fixed center of mass (SFCM), i.e, turning off the harmonic restraints; and ∆Arelease CM is the free energy of release of the center of mass (CM).\n\nECM can be difficult to implement because of the need for a fixed center of mass, so our work here is based on an alternative approach that is easier to implement. When particles move in ECM, the lattice needs to be moved because the center of mass is fixed47,48,50. Our method of choice, the Einstein Molecule Method (EMM, see Figure 1(b)), fixes a single atom in the lattice instead of the center of mass and is more easily implemented than ECM because of the relative difficulty of introducing center of mass restraints into existing simulation packages22,48,50–52. EMM has been used to predict phase diagrams of TIP4P and SPC/E water models48, free energies of ice polymorphs, solid methanol and toy systems49,52, and the solubilities of potassium and sodium chlorides22,51.\n\nIn EMM, the free energy of a solid is:\n\nAsolid = AEM+∆AEM→IEM+∆AIEM→solid (9)\n\nwhere AEM is the free energy of the ideal Einstein molecule; ∆Aid→IEM is the free energy difference between the ideal Einstein molecule and the interacting Einstein molecule (i.e, turning on the force field); and ∆AIEM→solid is the free energy difference between the interacting Einstein molecule and the solid (i.e, turning off the harmonic restraints). The advantage of EMM over ECM is the absence of the need to calculate a free energy term associated with releasing the fixed reference point48.\n\nHere, as per Equation 9, we compute the free energy of the solid by combining the absolute free energy of the ideal Einstein molecule with two terms that we calculate via alchemical free energy calculations—∆AEM→IEM and ∆AIEM→solid; these involve alchemically changing the interactions in the system. Numerical integration of Equation 10 allows the calculation of the ideal term, AEM52:\n\nAEM=−1βlnQEM=1βlnNΛ3V−1βln∫e−βUEM,1(Ω1)dΩ1−(N−1)βln∫1Λ3e−βUEM,2(r2,Ω2)dr2dΩ2(10)\n\nwhere AEM and QEM are the free energy of the Einstein molecule and its partition function; UEM,1(Ω1) is the potential energy of the fixed particle 1; UEM,2(r2Ω2) is the potential energy of a non-fixed particle at a distance r2 of particle 1; Ω1 and Ω2 are all the possible orientations the molecules can have in the lattice; Λ, V, N, and β respectively are the de Broglie wavelength, the system’s volume, its number of particles, and the reciprocal of kBT, the product of the Boltzmann constant and the absolute temperature.\n\nAnother critical component of computing the solubility of a compound is estimating the chemical potential of a solute in solution, since the solubility point is the concentration at which the chemical potentials of compound in the two phases are equal.\n\nThe chemical potential of a component i in solution, µi, has an ideal and and excess component:\n\nμi=−1βln qi+1βlnΛi3NiV−1βln〈e−β[U(Ni+1)−U(Ni)]〉initial(11)\n\nwhere qi is the internal partition function of a single molecule of the solute, U(Ni) is the potential energy of the system with Ni particles, Λ is the de Broglie thermal wavelength, and V is the system’s volume53. 〈〉initial means that the term was obtained from an ensemble average over the configurations of the initial state (see Equation 5). The first two terms of the equation above correspond to the ideal component of µi; the last one, μiex, corresponds to the excess component of µi, and is associated with all non-ideal interactions of the extra component i with the solution (i.e. physical interactions that differ from those given by the ideal gas law). We obtained excess chemical potentials from solvation free energy calculations; the solute molecule is inserted in the solution by progressively turning on its interactions with the surrounding environment24,28,54.\n\nThe challenge associated with the calculation of µi is the calculation of the standard chemical potential of i, μi0, the first term of Equation 11. qi, the internal partition function, includes the rotation, vibrational, electronic and nuclear partition functions of a single molecule53 and is unknown. Here, we found a way of calculating μi0 without the knowledge of qi by alchemically transforming a single solute molecule into a single Einstein molecule, whose absolute free energy we know how to calculate48–50.\n\nWe are aware of three main approaches to compute the solubility of solids in solution using physical approaches: ECM-based methods21,23, EMM-based methods22,51,55, and the approach of Michael Schnieders and collaborators which computes sublimation and solvation free energies and uses these in an alternate thermodynamic cycle to obtain solubility estimates15,56.\n\nMany of the applications of these approaches have been to the solubility of ionic solids, with both ECM-21 and EMM-based approaches22,51,55 having some success. However, molecular solids introduce substantial additional complexities for both of these approaches.\n\nThe ECM has seen an initial test on solubility estimation. Li et al.23 used the ECM to estimate the solubility of napthalene, but made several approximations such as assuming that the internal partition function component of the solute cancels between environments (perhaps justified given napthalene’s low solubility).\n\nWe are not aware of any work applying the EMM to solubility estimation of molecular solids; to our knowledge our work is the first to make such an attempt, though EMM has been used before to estimate the free energy of simple molecular solids49,52 but not the solubility. This explains our need to find our own approach to estimate μi0 for a single solute molecule.\n\nThe Schneiders approach is an orthogonal one that we do not examine here.\n\n\nMethods\n\nHere, we chose three systems to study: An argon crystal for some small initial tests, α-methanol to help establish our protocol, and acetylsalicylic acid (ASA) as our main object of study. ASA is a known anti-inflammatory whose most stable polymorph, form I57, has an aqueous solubility of approximately 0.038% mole fraction at 298 K58. We also used α-methanol at 150 K and a toy facecentered cubic (fcc) argon crystal59 to help us find an optimal protocol to calculate the absolute free energy of a molecular solid. α-methanol was chosen because it had been used before in a study that applied the EMM to calculate the absolute free energy of the solid52.\n\nAll simulations were run in GROMACS 4.6.760–63. With one exception, all simulations used the General Amber force field (GAFF) version 1.7 with AM1-BCC charges64,65; the exception was α-methanol, because we ran these simulations using the input files – coordinates and force field parameters – provided by Aragonès et al., who used an united atom version of the OPLS force field52.\n\nWe simulated all solids and liquids using 5 ns Langevin dynamics simulations. ASA, α-methanol, and argon were simulated respectively at 298.15 K, 150.0 K, and 4.0 K. Our simulations had the same length as the simulations run by Aragonès et al. All solid state simulations were run in NVT conditions. Liquid state simulations were run in NPT conditions; pressure was kept constant at 101.335 kPa using the Parrinello-Rahman barostat66. We used the TIP3P water model67 for all our liquid state simulations. More simulation details and example input files with full details can be found in the Supporting Information.\n\nThe absolute free energies of the solids were calculated from trajectories of simulation boxes with 64 ASA molecules, 100 OPLS methanol molecules, and 864 argon atoms with periodic boundary conditions. ASA’s unit cell was obtained from Mercury CSD 3.868 and the fcc argon crystal was obtained from the literature59. Simulation box sizes were chosen to be approximately between 2 nm and 3 nm to ensure that box sizes were large enough that atoms and their periodic copies were not within cutoff distance of one another. α-methanol’s crystal was obtained from the Supporting information of Aragonès et al.52 We used Amber14’s ambertools69–72 and ParmEd73 to generate the ASA’s and argon’s solid state input files. All atoms but one were subjected to harmonic constraints in the x, y, and z coordinates. A single atom was kept fixed in space to act as the reference point for the calculations, as explained in the Introduction.\n\nMonte Carlo integration yielded AEM, the free energy of the Einstein molecule, as it was previously done for α-methanol in the literature52. ∆Aid→IEM and ∆AIEM→solid were estimated using TI39 and the multistate Bennett acceptance ratio (MBAR)74. We used force constants of 4000 kBT/Å2 to restrain atoms to their lattice positions in ASA and argon simulations because it allowed us to use a reasonable time step of 1.0 fs in all simulations. α-methanol simulations used the same force constant that had been previously used by Aragonès et al.52.\n\nWe used alchemical free energy calculations to obtain the difference in free energy between the reference Einstein molecule and the solid. This step was divided in two parts: (a) the force field parameters are alchemically turned on, and (b) the harmonic constraints are turned off.\n\nHere, we deviate from earlier work which calculated the absolute free energy of a solid using EMM by introducing additional intermediate states to improve accuracy, along with using a superior free energy estimator.\n\nFor the calculation of ∆Aid→IEM, we found it was crucial to introduce intermediate states; we also switched to using the MBAR estimator. The original EMM calculation of the absolute free energy of a solid22,48–52 estimated ∆Aid→IEM using exponential averaging (EXP) with just two states: the Einstein molecule (EM) and the interacting Einstein molecule (IEM)21,22,48–52,55. As EXP is known to have convergence issues and biases38,40,41,45, we switched to the superior MBAR free energy estimator74. Additionally, when we did so, we found that overlap of states (as measured by the overlap matrix75) was insufficient so we created a series of intermediate states connecting both ends of the transformation.\n\nFor ∆AIEM→solid., the original work used TI39. Here, we replaced TI with MBAR as our analysis method of choice. Generally, the literature shows that TI performs as well as more efficient methods like BAR and MBAR when the integrand is smooth38,40,41, but it is sensitive to the choice and number of intermediate states76. MBAR is the most consistently well-performing free energy estimator42 and exploits the overlap between states more thoroughly than its predecessor, the Bennett Acceptance Ratio (BAR) estimator74. Here, we chose to compare performance of MBAR and TI for calculation of ∆AIEM→solid for ASA and α-methanol; we also applied EXP as a comparison in the latter case only.\n\nThe chemical potential of a pure solid is its molar free energy:\n\nμ=AN(12)\n\nwhere N is the number of molecules in the solid, and A its Helmholtz free energy.\n\nThe chemical potential of a substance i in water is defined as the derivative of the free energy of the system with respect to the composition:\n\nμi=(∂G∂Ni)P,T,NH2O(13)\n\nwhere G is the Gibbs free energy, and Ni is the number of molecules of i in solution; P, T, and NH2O are the pressure, absolute temperature, and number of water molecules in solution, and are kept constant in the calculation.\n\nOne important aspect to discuss is the reason why we chose to calculate the Helmholtz free energy for the solid and Gibbs free energies for each solution. Solid state simulations with position restraints required running under constant temperature and constant volume conditions due to software limitations, therefore we were able to calculate A for the solids. At constant pressure, both kinds of free energy are related by:\n\n∆G = ∆A+ P∆V (14)\n\nSince solids are much less susceptible to volume changes than liquids, it is reasonable to consider that P∆V is negligible and ∆G ≈ ∆A. For instance, the difference in volume between the experimental ASA crystal structure and the simulation box after a constant pressure equilibration stage is 0.14 nm3. The P∆V term – i.e., the free energy difference discounting possible structure relaxation effects – would be much smaller than the simulation error.\n\nAs we explain in more detail in the Results section, successful absolute free energy calculations for molecular solids require a pathway involving a large number of alchemical intermediate states. The calculation of the absolute free energies of α-methanol at 150 K and ASA required 600 states. Since GROMACS reads each λ until its fourth decimal place and the states need to be spaced more closely together as as the harmonic restraints are turned off (see Supporting Information), we decided to split each free energy calculation into sets of 100 states.\n\nLiquid state simulation boxes were generated using the SolvationToolkit77, a Python package that uses packmol78, OpenMolTools (v0.6.7)79 and OpenEye Python Toolkits80–82. Excess chemical potentials were obtained with the same solvation free energy protocol used in previous studies28: Starting from a fully interacting system, we progressively decouple the interactions of a single solute molecule with the remaining of the system, which allows us to calculate the free energy difference between a solute molecule in vacuum and in solution (i.e., the solvation free energy).\n\nWe also used alchemical free energy calculations using a single Einstein molecule as a reference state to estimate the standard chemical potential of a substance, μi0:\n\n\n\nwhere μiFFoff and μirestraining respectively are the chemical potential associated with turning off the force field and chemical potential of restraining the atoms of the molecule to their lattice positions. μiideal is calculated using the Monte Carlo integration procedure that we used to calculate AEM to a single molecule.\n\n\nResults\n\nThe first step to predict aqueous solubilities with the aid of absolute free energy calculations was the assessment of the methodologies we chose to use. Since our method is the same one used by Aragonès et al.52 and we wanted to be sure that we could reproduce previous results, we ran simulations for α-methanol at 150 K and estimated the free energies of solids using MBAR. Turning off the harmonic restraints was the challenging step. Our MBAR calculation of ∆AIEM→solid for α-methanol using 18 intermediate states yielded −18(3) kBT, while our TI result was −18.421(5) kBT and the literature result was −17.33(3) kBT using 17 states52. The MBAR error was unusually high (3 kBT), which is usually a signal of overlap problems or other serious concerns.\n\nMBAR is a free energy estimation method that minimizes the free energy variance and considers the overlap between a given state and all the others in the transformation path41, which means that high uncertainties (±3kBT) suggest the presence of problems in the transformation’s path. TI’s uncertainty estimates are much lower, but we believe that this is an artifact. Error analysis for TI simply does not work the same way and does not give insight into whether exploration of phase space is adequate, unlike MBAR. Specifically, uncertainty estimates from TI usually factor in only the uncertainty in the integrand at each sampled lambda value and could potentially also factor in the smoothness of the integrand (i.e. numerical integration error) but do nothing to factor in whether the integrand will in fact vary smoothly in between lambda points; usually no data is available on this. BAR and MBAR, in contrast, factor in information about how well the intermediate states overlap in phase space and reflect high uncertainties when phase space overlap is poor. In our experience, TI would usually suffer from similar problems if additional intermediate states were added, but uncertainties in TI typically do not reflect this, as is the case here. Thus, the high uncertainty of the MBAR value indicates a sampling/convergence problem which warrants further exploration.\n\nTo explore the high uncertainty of our MBAR free energy estimates, we examined the degree of overlap the intermediate states had with each other. Phase space overlap analysis83–85 quantifies the probability that any given configuration of an intermediate state can be found in other states. A good rule of thumb for designing a set of free energy calculations spanning between two states is to ensure that the states along the path have significant overlap with their neighbors as shown in Figure 2. More overlap improves the quality of the MBAR free energy estimation: Figure 2b represents a set of restraining simulations where the free energy uncertainty can potentially be accurately estimated using BAR and MBAR; Figure 2a shows a case where it cannot. In our case we find that the α-methanol simulation using 18 intermediate states does not have adequate overlap (Figure 3)– specifically, the states 4 ≤ λi ≤ 17 do not have overlapping configurations with other states, which explains the 3 kBT uncertainty in our MBAR estimate.\n\nΓ represents the phase space that contains all the configurations for all the states in the path. λ0 and λ1 (left) or λN (right) represent the end states along the path, each shaded region represents a state in phase space and the red lines represent the configurations visited by the simulation run in the λ0 state. The restrained state is a subset of the unrestrained one. (a) and (b) represent simulations with different numbers of intermediate states along the path between a fully restrained state (λ1 (a) or λN (b)) and an unrestrained state (λ0). In (a), the simulation (red) only visits very few configurations consistent with the restrained state – i.e, there is poor phase space overlap – indicating a need for more intermediate states, otherwise any free energy estimates will be subject to very high uncertainties; in (b) there is still almost no overlap between the simulation and states consistent with λN, but there is overlap with the next shaded region, λ1, indicating the potential for overlap and accurate free energy estimates. Thus simulations run in each shaded region are more likely to have a bigger phase space overlap with λN than simulations run in λ0.\n\nThe sum of all the elements in a row should yield 1.0, a probability of 100%. A good free energy estimate is obtained when the states along the alchemical path contain configurations that can be found in other intermediate states. In these situations, the phase space overlap is non-zero, which results in non-zero off-diagonal elements. Here, however, the phase space overlap plot shows that there is no overlap between the states λi, 4 ≤ i ≤ 17 indicating poor free energy estimates will result.\n\nSince prior work had appeared to do this estimation successfully52, we were uncertain why we were encountering such overlap problems, so we studied an even simpler system. We calculated ∆AIEM→solid of fcc argon at 4 K with 18 states as in our α-methanol free energy estimation. MBAR yielded an error estimate of infinity, whereas TI estimated ∆AIEM→solid to be −1666.5(8) kBT, which, as we show below, is incorrect. This path resulted phase space overlap diagram without overlap between the states after state number 2 (Figure 4). Apparently as the harmonic potential that holds atoms in their lattice positions tends to zero, atoms become rather mobile, dramatically decreasing phase space overlap and leading to poor free energy estimates.\n\nA good free energy estimate is obtained when the states along the alchemical path contain configurations that can be found in other intermediate states. Here, however, the phase space overlap diagram shows that there is no overlap between the states λi, 3 ≤ i ≤ 17, which explains the poor quality of the free energy result.\n\nTo improve phase space overlap, we introduced more intermediate states along the path for removing the restraints (see Figure 2). We chose to break down the simulation in smaller parts, adding a significant amount of states near the point where the harmonic restraints are approximately zero. The MBAR estimate of ∆AIEM→solid for fcc argon is −1016.0(2) kBT using 300 states. TI’s corresponding value was −1017(1) kBT, differing by far from the (incorrect) value of −1666.5(8) kBT obtained above with fewer states. Phase space overlap diagrams showed significant improvement in the configuration overlap between the states (Supporting Information). Thus, increasing the number of states was an effective strategy, and we used it in all subsequent calculations.\n\nEven though our α-methanol results were similar to results published previously by other authors52, we need to emphasize that reliable free energies resulted from simulations with a large number of intermediate states, as can be seen in Table 1. Despite its conceptual simplicity, calculating the components of the absolute free energy of a solid to a point where there is significant phase space overlap between the intermediate states is computationally demanding. A 900-atom OPLS α-methanol system required 40 states to calculate ∆Aid→IEM, and 600 states for ∆AIEM→solid.\n\nWe chose these intermediate states in advance, and these ultimately led to free energy errors smaller than 0.1 kBT; the estimated TI and MBAR values differed by no more than 0.3 kBT. Our results for ASA using an optimal number of states can be seen in Table 2. The MBAR chemical potential of ASA at 298.15 K equals to −220.67(3) kBT.\n\nThe computational cost of calculating AASA was high; Each state required a separate simulation (of a 1344-atom ASA system), with 718 states in total. Simulations typically required 11 hours on a single CPU, so the calculation of a single absolute free energy of a molecular solid required approximately 7898 CPU-hours.\n\nEquation 11 states that the absolute chemical potential of a solution is determined by three quantities: μi0, the standard chemical potential; μiex, the excess chemical potential of the component at a concentration of χ; and a volume-dependent ideal gas component of kBT⋅ ln (Λi3⋅NASA/〈V〉solution). Calculation of μASA0 only required information regarding the internal structure of the molecule53, thus we estimated μASA0 by alchemically transforming a single solute molecule into a single Einstein molecule (Table 3), whose absolute free energy we know how to calculate. We used the same number of states that we chose for the solid state simulations and we found that μASA0 is equal to –150.7(2) kBT, as discussed in the last subsection of the Methods section.\n\nConcentrations, volumes and excess chemical potentials can be seen in Table 4. We obtained the excess chemical potentials from solvation free energy calculations24,28,54. Volumes were obtained from the state in the alchemical path where the solute was fully coupled to the rest of the system.\n\nThe experimental aqueous solubility of ASA is approximately 0.038% in water at 298 K58, but our model predicts that ASA is effectively insoluble in water (Figure 5). While all-atom simulations can yield solubility estimates given adequate simulation time and a correct method, the computed solubility will be that dictated by the underlying energy model or force field, and will not necessarily match experiment. Here, we use GAFF, a general-purpose force field with known limitations28,71,86,87; apparently, here, the right answer for the force field is not correct. Perhaps this is because of limitations in describing the solid state, as the force field is parameterized for liquid state simulations. Indeed, classical fixed charge force fields have shown severe limitations for polymorph prediction for these reasons5,31,33–35. Also, point partial atomic charges regularly used in molecular dynamics do not describe electrostatic interactions in a solid particularly well88. In the case of the ASA crystal, it is possible that its hydrogen bonds and π-stacking interactions add layers of complexity that are not properly described by GAFF.\n\n\nDiscussion\n\nDespite its theoretical rigor, solubility prediction from absolute free energy calculations is a difficult task: it is computationally expensive and, at least in the present approach, requires many different steps and a great deal of care. Here, we attempted to develop and test a general approach to compute the solubility of molecular solids by adapting the EMM to tackle this problem, as discussed above.\n\nTo tune our methodology, we initially decided to reproduce the absolute free energy of solid α-methanol, one of methanol’s polymorphs, at 150 K using EMM before doing the same calculations for our compound of choice, ASA. We verified that the free energy differences between the Einstein molecule and the interactive Einstein molecule (∆AEM→IEM) and between the latter state and the solid (∆AIEM→solid) were more reliably estimated with the MBAR. The absolute free energy of the crystal (as computed for united-atom OPLS α-methanol) agreed with results found in the literature, which suggested that we were on the right path. We did, however, require a very large number of intermediate alchemical states to obtain accurate free energy estimates, making these simulations fairly computationally demanding\n\nWe then chose to calculate the solubility of ASA, owing to its pharmacological importance and its relative complexity compared to previous molecular solids, whose absolute free energies have been computed via EMM previously52. As for α-methanol, this calculation required a large number of intermediate alchemical states and considerable computational cost – approximately 8000 CPU hours for a single absolute free energy calculation for the molecular solid, even with the crystal structure as input. Perhaps the number of intermediate states could be further optimized, but clearly a large number of intermediate simulations was required and thus considerable computational cost. Despite all of this, we still could not reproduce the experimental aqueous solubility of ASA; experimentally it is modestly soluble, whereas our work would suggest it is essentially completely insoluble in water, likely due to force field limitations.\n\nThe solubility of naphthalene was recently estimated using a similar methodology, the Extended Einstein Crystal Method23, but with additional approximations. Specifically, since naphthalene molecules interact very weakly with each other in the crystal lattice and with water molecules in solution, the differences between the internal partition function of a naphthalene molecule in the solid and in the solution were assumed to be negligible. This allowed the authors to drop some complexities in treatment of the solution-phase part of the calculation. However, that approach is only suitable for compounds that are only very weakly interacting in solution and in the crystal. ASA, in contrast, is a molecule that interacts strongly with other ASA molecules in its crystal lattice and with water molecules in solution via hydrogen bonds. For instance, an important crystalline feature that is not necessarily present in solution is the dimer structure, with two ASA molecules bound together via hydrogen bonds between the carboxylic acid groups. Differences between the internal partition functions of the molecule in the solid (qASAsolid) and in solution (qASAsolution) would probably not be negligible in this scenario, thus a more general approach is needed for treatment of such cases. Our work here provides one attempt in that direction.\n\nOverall, the present approach seems to have significant limitations – most notably that the computational expense is considerable, and the resulting estimated solubility is quite inaccurate. Perhaps both of these may be surmountable; GPU-based free energy calculations can be dramatically faster, potentially reducing an 8000 CPU-hour calculation to 80 GPU hours, which would amount to overnight on 8 GPUs, and perhaps this could be optimized via changes to simulation time and number of intermediate states. And with better force fields, perhaps accuracy could be improved; the AMOEBA-based approach of Schnieders shows considerable promise15. Alternatively, other approaches may be of interest. Solubility has been predicted by simulations using pseudocritical paths (i.e., paths were molecular crystals are transformed in tractable Einstein crystal-like states between the ending states of the transformation89–92) and a single experimental reference point91), and with the aid of a thermodynamic cycle formed by the molecular crystal, the molecule in vacuum, and the solvated molecule15.\n\nWe believe the time has come for routine physical methods for estimation of solubility, even if improved force fields prove necessary before results have significant accuracy for application to biomolecular design problems.\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": "Competing interests\n\n\n\nD.L.M. is a member of the Scientific Advisory Board for OpenEye Scientific Software.\n\n\nGrant information\n\nD.L.M. and G.D.R.M. appreciate the financial support from the National Science Foundation (CHE 1352608), and computing support from the UCI GreenPlanet cluster, supported in part by NSF Grant CHE-0840513. G.D.R.M. appreciates support from the Brazilian agency CAPES - Science without Borders program (BEX 3932-13-3).\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 would like to thank Dr. Gaetano Calabrò (OpenEye Software), Prof. Michael Shirts (University of Colorado, Boulder), Dr. Eric Dybeck (Pfizer), and Prof. Michael Schnieders (University of Iowa) for fruitful discussions on the project.\n\n\nSupplementary material\n\nSupporting Information. These files include GROMACS 4.6.7 input parameters for the simulation and all associated MDP files. Also included is a file containing the elements of the phase space overlap matrix of a ∆AEM→IEM, estimated from an alchemical path of 118 states.\n\nClick here to access the data.\n\n\nReferences\n\nPudipeddi M, Serajuddin AT: Trends in solubility of polymorphs. J Pharm Sci. 2005; 94(5): 929–939. PubMed Abstract | Publisher Full Text\n\nBauer J, Spanton S, Henry R, et al.: Ritonavir: an extraordinary example of conformational polymorphism. Pharm Res. 2001; 18(6): 859–866. PubMed Abstract | Publisher Full Text\n\nMotherwell WD, Ammon HL, Dunitz JD, et al.: Crystal structure prediction of small organic molecules: A second blind test. Acta Crystallogr B. 2002; 58(Pt 4): 647–661. PubMed Abstract | Publisher Full Text\n\nAaltonen J, Allesø M, Mirza S, et al.: Solid form screening--a review. Eur J Pharm Biopharm. 2009; 71(1): 23–37. PubMed Abstract | Publisher Full Text\n\nBardwell DA, Adjiman CS, Arnautova YA, et al.: Towards crystal structure prediction of complex organic compounds--a report on the fifth blind test. Acta Crystallogr B. 2011; 67(Pt 6): 535–551. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBraun DE, McMahon JA, Koztecki LH, et al.: Contrasting Polymorphism of Related Small Molecule Drugs Correlated and Guided by the Computed Crystal Energy Landscape. Cryst Growth Des. 2014; 14(4): 2056–2072. Publisher Full Text\n\nCruz-Cabeza AJ, Bernstein J: Conformational Polymorphism. Chem Rev. 2014; 114(4): 2170–2191. PubMed Abstract | Publisher Full Text\n\nNeumann MA, van de Streek J, Fabbiani FP, et al.: Combined crystal structure prediction and high-pressure crystallization in rational pharmaceutical polymorph screening. Nat Commun. 2015; 6: 7793. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBeran GJ: Modeling Polymorphic Molecular Crystals with Electronic Structure Theory. Chem Rev. 2016; 116(9): 5567–5613. PubMed Abstract | Publisher Full Text\n\nLlinàs A, Glen RC, Goodman JM: Solubility Challenge: Can You Predict Solubilities of 32 Molecules Using a Database of 100 Reliable Measurements? J Chem Inf Model. 2008; 48(7): 1289–1303. PubMed Abstract | Publisher Full Text\n\nHopfinger AJ, Esposito EX, Llinàs A, et al.: Findings of the Challenge To Predict Aqueous Solubility. J Chem Inf Model. 2009; 49(1): 1–5. PubMed Abstract | Publisher Full Text\n\nSalahinejad M, Le TC, Winkler DA: Aqueous Solubility Prediction: Do Crystal Lattice Interactions Help? Mol Pharm. 2013; 10(7): 2757–2766. PubMed Abstract | Publisher Full Text\n\nPalmer DS, Mitchell JB: Is Experimental Data Quality the Limiting Factor in Predicting the Aqueous Solubility of Druglike Molecules? Mol Pharm. 2014; 11(8): 2962–2972. PubMed Abstract | Publisher Full Text\n\nLusci A, Pollastri G, Baldi P: Deep architectures and deep learning in chemoinformatics: the prediction of aqueous solubility for drug-like molecules. J Chem Inf Model. 2013; 53(7): 1563–1575. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchnieders MJ, Baltrusaitis J, Shi Y, et al.: The Structure, Thermodynamics and Solubility of Organic Crystals from Simulation with a Polarizable Force Field. J Chem Theory Comput. 2012; 8(5): 1721–1736. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPalmer DS, Llinàs A, Morao I, et al.: Predicting Intrinsic Aqueous Solubility by a Thermodynamic Cycle. Mol Pharm. 2008; 5(2): 266–279. PubMed Abstract | Publisher Full Text\n\nWestergren J, Lindfors L, Höglund T, et al.: In silico prediction of drug solubility: 1. Free energy of hydration. J Phys Chem B. 2007; 111(7): 1872–1882. PubMed Abstract | Publisher Full Text\n\nLüder K, Lindfors L, Westergren J, et al.: In silico prediction of drug solubility: 2. Free energy of solvation in pure melts. J Phys Chem B. 2007; 111(7): 1883–1892. PubMed Abstract | Publisher Full Text\n\nLüder K, Lindfors L, Westergren J, et al.: In silico prediction of drug solubility. 3. Free energy of solvation in pure amorphous matter. J Phys Chem B. 2007; 111(25): 7303–7311. PubMed Abstract | Publisher Full Text\n\nLüder K, Lindfors L, Westergren J, et al.: In silico prediction of drug solubility: 4. Will simple potentials suffice? J Comput Chem. 2009; 30(12): 1859–1871. PubMed Abstract | Publisher Full Text\n\nFerrario M, Ciccotti G, Spohr E, et al.: Solubility of KF in water by molecular dynamics using the Kirkwood integration method. J Chem Phys. 2002; 117(10): 4947–4953. Publisher Full Text\n\nSanz E, Vega C: Solubility of KF and NaCl in water by molecular simulation. J Chem Phys. 2007; 126(1): 014507. PubMed Abstract | Publisher Full Text\n\nLi L, Totton T, Frenkel D: Computational methodology for solubility prediction: Application to the sparingly soluble solutes. J Chem Phys. 2017; 146(21): 214110. PubMed Abstract | Publisher Full Text\n\nKlimovich PV, Mobley DL: Predicting hydration free energies using all-atom molecular dynamics simulations and multiple starting conformations. J Comput Aided Mol Des. 2010; 24(4): 307–316. PubMed Abstract | Publisher Full Text\n\nShivakumar D, Williams J, Wu Y, et al.: Prediction of Absolute Solvation Free Energies using Molecular Dynamics Free Energy Perturbation and the OPLS Force Field. J Chem Theory Comput. 2010; 6(5): 1509–1519. PubMed Abstract | Publisher Full Text\n\nShivakumar D, Harder E, Damm W, et al.: Improving the Prediction of Absolute Solvation Free Energies Using the Next Generation OPLS Force Field. J Chem Theory Comput. 2012; 8(8): 2553–2558. PubMed Abstract | Publisher Full Text\n\nSkyner RE, McDonagh JL, Groom CR, et al.: A review of methods for the calculation of solution free energies and the modelling of systems in solution. Phys Chem Chem Phys. 2015; 17(9): 6174–6191. PubMed Abstract | Publisher Full Text\n\nMatos GDR, Kyu DY, Loeffler HH, et al.: Approaches for calculating solvation free energies and enthalpies demonstrated with an update of the FreeSolv database. J Chem Eng Data. 2017; 62(5): 1559–1569. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoulanger E, Huang L, Rupakheti C, et al.: Optimized Lennard-Jones Parameters for Druglike Small Molecules. J Chem Theory Comput. 2018. PubMed Abstract | Publisher Full Text\n\nPrice SL: The computational prediction of pharmaceutical crystal structures and polymorphism. Adv Drug Deliv Rev. 2004; 56(3): 301–319. PubMed Abstract | Publisher Full Text\n\nDay GM, Motherwell WD, Ammon HL, et al.: A third blind test of crystal structure prediction. Acta Crystallogr B. 2005; 61(Pt 5): 511–527. PubMed Abstract | Publisher Full Text\n\nWoodley SM, Catlow R: Crystal structure prediction from first principles. Nat Mater. 2008; 7(12): 937–46. PubMed Abstract | Publisher Full Text\n\nDay GM, Cooper TG, Cruz-Cabeza AJ, et al.: Significant progress in predicting the crystal structures of small organic molecules--a report on the fourth blind test. Acta Crystallogr B. 2009; 65(Pt 2): 107–125. PubMed Abstract | Publisher Full Text\n\nPrice SS: Computed crystal energy landscapes for understanding and predicting organic crystal structures and polymorphism. Acc Chem Res. 2009; 42(1): 117–126. PubMed Abstract | Publisher Full Text\n\nReilly AM, Cooper RI, Adjiman CS, et al.: Report on the sixth blind test of organic crystal structure prediction methods. Acta Crystallogr B Struct Sci Cryst Eng Mater. 2016; 72(Pt 4): 439–459. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChipot C, Pohorille A: Free Energy Calculations Theory and Applications in Chemistry and Biology. Springer, 2007. Publisher Full Text\n\nChipot C: Frontiers in free-energy calculations of biological systems. WIREs Comput Mol Sci. 2014; 4(1): 71–89. Publisher Full Text\n\nShirts MR, Mobley DL, Chodera JD: Alchemical Free Energy Calculations: Ready for Prime Time? In DC. Spellmeyer and R. Wheeler, editors, Annu Rep Comput Chem. Elsevier. 2007; 3: 41–59. Publisher Full Text\n\nKirkwood JG: Statistical Mechanics of Fluid Mixtures. J Chem Phys. 1935; 3(5): 300–313. Publisher Full Text\n\nYtreberg FM, Swendsen RH, Zuckerman DM: Comparison of free energy methods for molecular systems. J Chem Phys. 2006; 125(18): 184114. PubMed Abstract | Publisher Full Text\n\nShirts MR, Pande VS: Comparison of efficiency and bias of free energies computed by exponential averaging, the Bennett acceptance ratio, and thermodynamic integration. J Chem Phys. 2005; 122(14): 144107. PubMed Abstract | Publisher Full Text\n\nPaliwal H, Shirts MR: A Benchmark Test Set for Alchemical Free Energy Transformations and Its Use to Quantify Error in Common Free Energy Methods. J Chem Theory Comput. 2011; 7(12): 4115–4134. PubMed Abstract | Publisher Full Text\n\nZwanzig RW: High-Temperature Equation of State by a Perturbation Method. I. Nonpolar Gases. J Chem Phys. 1954; 22(8): 1420–1426. Publisher Full Text\n\nBennett CH: Efficient estimation of free energy differences from Monte Carlo data. J Comp Phys. 1976; 22(2): 245–268. Publisher Full Text\n\nWu D, Kofke DA: Asymmetric bias in free-energy perturbation measurements using two Hamiltonian-based models. Phys Rev E Stat Nonlin Soft Matter Phys. 2004; 70(6 Pt 2): 066702. PubMed Abstract | Publisher Full Text\n\nWu D, Kofke DA: Phase-space overlap measures. I. Fail-safe bias detection in free energies calculated by molecular simulation. J Chem Phys. 2005; 123(5): 54103. PubMed Abstract | Publisher Full Text\n\nFrenkel D, Ladd AJC: New Monte Carlo method to compute the free energy of arbitrary solids. Application to the fcc and hcp phases of hard spheres. J Chem Phys. 1984; 81(7): 3188–3193. Publisher Full Text\n\nVega C, Sanz E, Abascal JLF, et al.: Determination of phase diagrams via computer simulation: Methodology and applications to water, electrolytes and proteins. J Phys Condens Matter. 2008; 20(15): 153101. Publisher Full Text\n\nNoya EG, Conde MM, Vega C: Computing the free energy of molecular solids by the Einstein molecule approach: ices XIII and XIV, hard-dumbbells and a patchy model of proteins. J Chem Phys. 2008; 129(10): 104704. PubMed Abstract | Publisher Full Text\n\nVega C, Noya EG: Revisiting the Frenkel-Ladd method to compute the free energy of solids: the Einstein molecule approach. J Chem Phys. 2007; 127(15): 154113. PubMed Abstract | Publisher Full Text\n\nAragones JL, Sanz E, Vega C: Solubility of NaCl in water by molecular simulation revisited. J Chem Phys. 2012; 136(24): 244508. PubMed Abstract | Publisher Full Text\n\nAragones JL, Noya EG, Valeriani C, et al.: Free energy calculations for molecular solids using GROMACS. J Chem Phys. 2013; 139(3): 034104. PubMed Abstract | Publisher Full Text\n\nBen-Naim A: Molecular Theory of Solutions. Oxford University Press, Oxford, 2006. Reference Source\n\nMobley DL, Guthrie JP: FreeSolv: a database of experimental and calculated hydration free energies, with input files. J Comput Aided Mol Des. 2014; 28(7): 711–720. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBenavides AL, Aragones JL, Vega C: Consensus on the solubility of NaCl in water from computer simulations using the chemical potential route. J Chem Phys. 2016; 144(12): 124504. PubMed Abstract | Publisher Full Text\n\nPark J, Nessler I, McClain B, et al.: Absolute Organic Crystal Thermodynamics: Growth of the Asymmetric Unit into a Crystal via Alchemy. J Chem Theory Comput. 2014; 10(7): 2781–2791. PubMed Abstract | Publisher Full Text\n\nKim Y, Machida K, Taga T, et al.: Structure redetermination and packing analysis of aspirin crystal. Chem Pharm Bull (Tokyo). 1985; 33(7): 2641–2647. PubMed Abstract | Publisher Full Text\n\nYalkowsky SH, He Y, Jain P: Handbook of Aqueous Solubility Data. CRC Press, second edition, 2010. Reference Source\n\nHenshaw DG: Atomic Distribution in Liquid and Solid Neon and Solid Argon by Neutron Diffraction. Phys Rev. 1958; 111(6): 1470–1475. Publisher Full Text\n\nBerendsen HJC, van der Spoel D, van Drunen R: GROMACS: A message-passing parallel molecular dynamics implementation. Comput Phys Comm. 1995; 91(1): 43–56. Publisher Full Text\n\nVan Der Spoel D, Lindahl E, Hess B, et al.: GROMACS: fast, flexible, and free. J Comput Chem. 2005; 26(16): 1701–1718. PubMed Abstract | Publisher Full Text\n\nHess B, Kutzner C, van der Spoel D, et al.: GROMACS 4: Algorithms for Highly Efficient, Load-Balanced, and Scalable Molecular Simulation. J Chem Theory Comput. 2008; 4(3): 435–447. PubMed Abstract | Publisher Full Text\n\nPronk S, Páll S, Schulz R, et al.: GROMACS 4.5: a high-throughput and highly parallel open source molecular simulation toolkit. Bioinformatics. 2013; 29(7): 845–854. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJakalian A, Bush BL, Jack DB, et al.: Fast, efficient generation of high-quality atomic charges. AM1-BCC model: I. Method. J Comput Chem. 2000; 21(2): 132–146. Publisher Full Text\n\nJakalian A, Jack DB, Bayly CI: Fast, efficient generation of high-quality atomic charges. AM1-BCC model: II. Parameterization and validation. J Comput Chem. 2002; 23(16): 1623–1641. 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(12): 7182. Publisher 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\nMacrae CF, Edgington PR, McCabe P, et al.: Mercury: Visualization and analysis of crystal structures. J Appl Cryst. 2006; 39(3): 453–457. Publisher Full Text\n\nCase DA, Cheatham TE 3rd, Darden T, et al.: The Amber biomolecular simulation programs. J Comput Chem. 2005; 26(16): 1668–1688. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSalomon-Ferrer R, Case DA, Walker RC: An overview of the Amber biomolecular simulation package. WIREs Comput Mol Sci. 2013; 3(2): 198–210. Publisher Full Text\n\nCheatham TE 3rd, Case DA: Twenty-five years of nucleic acid simulations. Biopolymers. 2013; 99(12): 969–977. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCase DA, Babin V, Berryman JT, et al.: Amber 14. 2014. Reference Source\n\nSwails J, Hernandez C, Mobley DL, et al.: Parmed. (accessed October 9, 2015).\n\nShirts MR, Chodera JD: Statistically optimal analysis of samples from multiple equilibrium states. J Chem Phys. 2008; 129(12): 124105. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKlimovich PV, Mobley DL: A Python tool to set up relative free energy calculations in GROMACS. J Comput Aided Mol Des. 2015; 29(11): 1007–1014. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Ruiter A, Boresch S, Oostenbrink C: Comparison of thermodynamic integration and Bennett acceptance ratio for calculating relative protein-ligand binding free energies. J Comp Chem. 2013; 34(12): 1024–1034. PubMed Abstract | Publisher Full Text\n\nBannan CC, Calabró G, Kyu DY, et al.: Calculating Partition Coefficients of Small Molecules in Octanol/Water and Cyclohexane/Water. J Chem Theory Comp. 2016; 12(8): 4015–4024. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMartínez L, Andrade R, Birgin EG, et al.: PACKMOL: a package for building initial configurations for molecular dynamics simulations. J Comp Chem. 2009; 30(13): 2157–2164. PubMed Abstract | Publisher Full Text\n\nBeauchamp KA, Rustenburg AS, Rizzi A, et al.: OpenMolTools.\n\nInc: OpeneEye Scientific Software. OEChem Toolkit, 2010; (accessed June 16, 2015).\n\nInc: OpeneEye Scientific Software. QUACPAC 1.7.0.2.\n\nHawkins PC, Nicholls A: Conformer generation with OMEGA: learning from the data set and the analysis of failures. J Chem Inf Model. 2012; 52(11): 2919–2936. PubMed Abstract | Publisher Full Text\n\nFrenkel D, Smit B: Understanding Molecular Simulation: From Algorithms to Applications. Academic Press Inc. (London) Ltd., 1996; 1. Reference Source\n\nPohorille A, Jarzynski C, Chipot C: Good Practices in Free-Energy Calculations. J Phys Chem B. 2010; 114(32): 10235–10253. PubMed Abstract | Publisher Full Text\n\nKlimovich PV, Shirts MR, Mobley DL: Guidelines for the analysis of free energy calculations. J Comput Aided Mol Des. 2015; 29(5): 397–411. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMobley DL, Dumont E, Chodera JD, et al.: Comparison of charge models for fixed-charge force fields: small-molecule hydration free energies in explicit solvent. J Phys Chem B. 2007; 111(9): 2242–2254. PubMed Abstract | Publisher Full Text\n\nFennell CJ, Wymer KL, Mobley DL: A fixed-charge model for alcohol polarization in the condensed phase, and its role in small molecule hydration. J Phys Chem B. 2014; 118(24): 6438–6446. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKramer C, Spinn A, Liedl KR: Charge Anisotropy: Where Atomic Multipoles Matter Most. J Chem Theory Comput. 2014; 10(10): 4488–4496. PubMed Abstract | Publisher Full Text\n\nEike DM, Brennecke JF, Maginn EJ: Toward a robust and general molecular simulation method for computing solid-liquid coexistence. J Chem Phys. 2005; 122(1): 14115. PubMed Abstract | Publisher Full Text\n\nEike DM, Maginn EJ: Atomistic simulation of solid-liquid coexistence for molecular systems: application to triazole and benzene. J Chem Phys. 2006; 124(16): 164503. PubMed Abstract | Publisher Full Text\n\nPaluch AS, Jayaraman S, Shah JK, et al.: A method for computing the solubility limit of solids: application to sodium chloride in water and alcohols. J Chem Phys. 2010; 133(12): 124504. PubMed Abstract | Publisher Full Text\n\nDybeck EC, Schieber NP, Shirts MR: Effects of a More Accurate Polarizable Hamiltonian on Polymorph Free Energies Computed Efficiently by Reweighting Point-Charge Potentials. J Chem Theory Comput. 2016; 12(8): 3491–3505. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "34597",
"date": "25 Jun 2018",
"name": "Lillian T. Chong",
"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 report extensive computations of absolute chemical potentials to predict the solubility of the drug, acetylsalicylic acid (aspirin), using molecular dynamics simulations. The manuscript is clearly written, providing the relevant background for understanding their results, a non-trivial task for this subject matter. A non-expert in the field of statistical mechanics should have little trouble reading this paper due to such careful and effective writing. In addition, figures such as Figure 1 are well constructed and greatly aid in the understanding of the relevant theory. While the results are not ideal, the challenges and limitations that have been revealed are informative and important for moving forward in the field of drug discovery. This manuscript would be of broad interest to life scientists. I recommend indexing of this manuscript in F1000Research.\n\nI have only a few comments for minor revisions:\nIntroduction section: This section could be re-framed to accentuate the positive, informative aspects of this manuscript’s results. For instance, it would be worth mentioning the fact that the new thermodynamic cycle employed in this study was able to enhance solubility calculations of the methanol system, an important feature of this manuscript that should be highlighted early on. In addition, the example of Norvir in the first paragraph could be shortened considerably, and similar shortening of the Introduction could be more effective in presenting the broader impacts of this study.\n\nTheory section: This section would benefit from a clear definition of an Einstein crystal for life scientists at the very beginning. Also important would be to include early in the Theory section explanations of the ECM and EMM cycles and logically structuring the rest of the section from there, including more fundamental equations as needed. Also, it is not clear in equation (5) what is meant by averaging over the configurations of the initial state. Please clarify.\n\nDistinctives of this Work section: It would be beneficial to remind the reader of the unique aspects of the EMM method over ECM since the implementation of this method is the novel in the manuscript.\n\nMethods section, third paragraph: It was not clear to this reader which temperatures were used for the liquid simulations. Please clearly mention the temperatures. If the simulations were run at 4 K, then the authors should comment on the accuracy of the TIP3P water model at this very low temperature.\n\nDiscussion section, second to last paragraph: The authors mention that “with better force fields, perhaps accuracy could be improved.” Regarding “better force fields”, it would be worth mentioning two recent fixed-charge force fields, AMBER ff15ipq and AMBER ff15fb, that have been developed using sweeping optimizations of hundreds of parameters simultaneously using automated tools and could be worth considering before using more expensive polarizable force fields such as AMOEBA.\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": "3769",
"date": "28 Jun 2018",
"name": "David Mobley",
"role": "Reader Comment",
"response": "Thanks, Lillian! This is extremely helpful; we'll work to address these issues. (This is our first time experimenting with this platform and so far it's a huge success; I like having the feedback attached publicly to the actual article. Now I just have to figure out how revising works...)"
},
{
"c_id": "3773",
"date": "28 Jun 2018",
"name": "David Mobley",
"role": "Reader Comment",
"response": "I also wanted to respond to this point: > The authors mention that “with better force fields, perhaps accuracy could be improved.” Regarding “better force fields”, it would be worth mentioning two recent fixed-charge force fields, AMBER ff15ipq and AMBER ff15fb, that have been developed using sweeping optimizations of hundreds of parameters simultaneously using automated tools and could be worth considering before using more expensive polarizable force fields such as AMOEBA. I agree that other force fields might be worth trying before switching to polarizable approaches. However, ff15ipq and ff15fb are protein/nucleic acid force fields and don't cover general small molecules. We ARE working on better general small molecule force fields, though, and hopefully some day we can try those."
},
{
"c_id": "4283",
"date": "04 Jan 2019",
"name": "David Mobley",
"role": "Reader Comment",
"response": "We just submitted a revised version with changes along these lines. Thanks again for your feedback."
}
]
},
{
"id": "36225",
"date": "25 Jul 2018",
"name": "Eric C. Dybeck",
"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\nSummary and Overall Impressions This article seeks to explore the application (and challenges) of using atomistic simulations to compute the solubility of drug-like small molecules. The ability to predict the solubility of emerging drug candidates is indeed an important challenge for the pharmaceutical industry, as many drugs which come out of discovery are BCS class II or IV with low aqueous solubility. The method proposed in this work could in principle allow medicinal chemists and pharmaceutical scientists to evaluate drug solubility early in the development pipeline and accelerate product release. This method is sufficiently novel and well executed to deserve publication in this scientific journal. The authors have also done a fantastic job including the details and input files necessary for one well-versed in the art to reproduce their results. Minor revisions are suggested below to further improve the clarity and quality of the article.\n\nSuggested Revisions\nIn the original papers by Aragones, Noya, and Vega, the molecules were constrained to be completely rigid. In this work, the investigators appear use this method for both constrained and fully flexible molecular systems. The authors should consider highlighting this expanded capability, and perhaps discuss the tradeoffs in accuracy and simulation speed between flexible vs rigid molecular treatment for absolute solubility prediction.\n\nThe authors use both the term ‘restraints’ and ‘constraints’ to describe the harmonic potential being applied and removed from atoms in the system. It may be more clear to consistently refer to these alchemical harmonic potentials as ‘restraints’ and reserve the term ‘constraint’ for the subroutine used to keep molecules fully rigid.\n\nAn important feature in the Einstein molecule method utilized herein is the use of a frozen reference atom rather than the traditional full-system center-of-mass removal. In principle, the free energy to add restraints to a system with a frozen atom will be independent of the choice of reference. In practice, some choices of reference atom may lead to faster simulation convergence than others due to differences in the fluctuation magnitude of the atoms around their natural lattice positions. It would be useful to discuss best practices in how one chose the frozen reference atom, as well as to discuss the effect of different reference choices on the convergence of the various alchemical steps in this workflow.\n\nThe authors mention using a previously developed Monte Carlo code to compute the absolute free energy of the reference Einstein Molecule state. It would be useful to comment on the uncertainty inherent in this component of the overall free energy calculation. For example, how much variance would ten independent calls to the Monte Carlo program have for these compounds?\n\nOn page 7 of the paper, the authors briefly mention that “GROMACS only reads each lambda value up to the 4th decimal place”. In my own alchemical simulations with GROMACS, I have routinely used lambda values out to 8 decimal places. If this is a version-specific limitation, the authors should state this explicitly. Otherwise, this comment should be removed.\n\nThe investigators chose to use linear spacing for the lambda values in all alchemical processes including the addition of harmonic restraints to the physical system. They also note that adding harmonic restraints represented the most time-intensive part of the overall workflow and required splitting into 6 different steps of increasing restraint strength. Furthermore, they find that the overlap between neighboring lambda states during the restraint addition is quite low (Figure 2-4) and produces large uncertainties in their final free energy estimates. In my own investigations of adding harmonic restraints to solids from 2016 (cited in this work) I observed that either cubically- or quartically- spaced lambda values significantly improve the overlap along the thermodynamic path relative to linear spacing and reduce the total amount of simulation cost. This should be included as a potential remedy for the high simulation cost lamented in the discussion section of the paper.\n\nThe investigators chose to use a large value of >1,000,000 kJ/nm/mol for their final restraint state to add or remove inter-particle interactions. It is necessary to have strong restraints in order to remove stiff degrees of freedom such as bonds and angles. However, it is possible that the interaction removal could be achieved with a weaker value of the restraint constant, and this would in turn reduce the number of simulations to add or remove harmonic restraints. This should also be discussed in the context of ways to reduce the amount of simulation expense observed for these calculations.\n\nFinally, the authors should consider including the additional papers of Sellers et al. 20161 and Schilling and Schmid 20092 who also explore the use of atomistic simulation to compute absolute solid free energies. These articles also discuss how to apply restraints in a manner that preserves the indistinguishability of certain particles. It would be worth discussing ways to account for particle indistinguishablility in the method presented in this paper.\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": "4282",
"date": "04 Jan 2019",
"name": "David Mobley",
"role": "Reader Comment",
"response": "We've just submitted a revised version to deal with these comments, which we very much appreciate. I'll just respond to a couple of the comments here \"for the record\".To your point 1:We appreciate the your point that our use of this for “flexible” molecules (even if not particularly flexible) is potentially an extension of earlier work and have amended our manuscript accordingly. We also added a couple of sentences highlighting that this works on molecules which are somewhat flexible, though pointing out that the molecules here are not especially floppy and the method may not work as well on especially floppy molecules.It's worth noting that our molecules are not as flexible as the word might imply. Acetylsalicylic acid is made of an aromatic ring bonded to a carboxyl group and an acetyl group in the ortho position. In the crystal structure the ring and the carboxyl group are rather rigid in the same plane -- there are hydrogen bonds between two ASA molecules forming a dimer -- and the only flexible part of the molecule is the acetyl group. EMM would probably not perform well if the crystal contained very floppy carbon chains -- butyl, pentyl, hexyl, and so on. On this case, since the “core” of the molecule is reasonably rigid and the acetyl group on the side rather fixed in a position due to the spacial arrangement.To your point 3, as we discussed by e-mail, in this study, we chose the frozen reference in arbitrary manner. In principle the choice of reference atom does not matter. For acetylsalicylic acid we selected one of the carbon atoms in the aromatic ring. It is not uncommon in free energy calculations of various types, including binding free energy calculations, to have to make arbitrary choices in which atoms to restrain or other considerations, and several studies have demonstrated that such choices in practice are unimportant, so applying a similar approach here was not a cause for concern. We have not examined this issue carefully. The revision now addresses this.To your point 4, the integrator is not very robust. It requires a lot of parameter-tweaking, as it was outlined in Aragonès et al. For methanol, the uncertainty is +- 0.09. For ASA, the uncertainty is +- 3 kT (an orientation change parameters need to be optimized for each case). We added a couple of sentences discussing this to the paper. Presumably this could be a point for future optimization.We addressed all your other points by making changes/additions to the text, including a rather extensive new discussion for potential places for optimization."
}
]
}
] | 1
|
https://f1000research.com/articles/7-686
|
https://f1000research.com/articles/7-713/v1
|
08 Jun 18
|
{
"type": "Research Article",
"title": "HIV treatment and monitoring patterns in routine practice: a multi-country retrospective chart review of patient care",
"authors": [
"Baba M Musa",
"Everistus Ibekwe",
"Stanley Mwale",
"Daniel Eurien",
"Catherine Oldenburg",
"Gary Chung",
"Richard F Heller",
"Baba M Musa",
"Everistus Ibekwe",
"Stanley Mwale",
"Daniel Eurien",
"Catherine Oldenburg",
"Gary Chung"
],
"abstract": "Background: A study of patient records in four HIV clinics in three sub-Saharan African countries examined routine clinical care patterns and variations. Methods: Clinic characteristics were described, and patient data extracted from a sample of medical records. Data on treatment, CD4 count and viral load (VL) were obtained for the last visit in the records, dates ranging from 2015 to 2017, patient demographic data were obtained from the first clinic visit. Results: Four clinics, two in Nigeria, one in Zambia and one in Uganda, all public facilities, using national HIV treatment guidelines were included. Numbers of patients and health professionals varied, with some variation in stated frequency of testing for CD4 count and VL. Clinical guidelines were available in each clinic, and most drugs were available free to patients. The proportion of patients with a CD4 count in the records varied from 84 to 100 percent, the latest median count varied from 269 to 593 between clinics. 35% had a record of a VL test, varying from 1% to 63% of patients. Lamivudine (3TC) was recorded for more than 90% of patients in each clinic, and although there was variation between clinics in the choice of antiretroviral therapy (ART), the majority were on first line drugs consistent with guidelines. Only about 2% of the patients were on second-line ARTs. In two clinics, 100% and 99% of patients were prescribed co-trimoxazole, compared with 7% and no patients in the two other clinics. Conclusions: The wide variation in available clinic health work force, levels and frequency of CD4 counts, and VL assessment and treatment indicate sub-optimal adherence to current guidelines in routine clinical care. There is room for further work to understand the reasons for this variation, and to standardise record keeping and routine care of HIV positive patients.",
"keywords": [
"HIV",
"quality of care",
"variation",
"Public Health"
],
"content": "Introduction\n\nIn addition to choice of appropriate drugs, best practice management of patients with HIV requires monitoring response to treatment and disease progression. This includes tracking clinical immunological, and virological data on patients at diagnosis and on follow-up. There is a rich literature guiding HIV treatment, and guidelines are developed and updated regularly as new evidence comes to light1–5 (see http://www.who.int/hiv/pub/guidelines/en/ for an historic list of guidelines produced by the World Health Organisation). The implementation of standardized protocols for treatment and investigations may vary in resource constrained countries, with differences in resources available for such services. For example, the current guidelines for antiretroviral therapy (ART) requires all patients on ART to have viral load test at six months and 12 months5, implementation might vary based on availability of resources to support viral load testing. While a number of reports of current treatment and testing practice are available from resource-limited settings, these are mainly in the context of patients enrolled in research centres or in research studies6–8. Such settings often have substantially more resources than routine clinical practices, and as such practice patterns may differ from routine practice, where issues such as clinic and patient resources or drug stock-outs can affect care. This study was designed to explore variation in the evaluation, treatment, care and follow-up among patients diagnosed with HIV in routine care settings in low- to middle-income countries.\n\n\nMethods\n\nInterest in participating in this study was sought among the Masters of Public Health Alumni of Peoples-uni9. A protocol was developed by those who responded (Supplementary File 1), and a retrospective cohort was created via standardized record review.\n\nA data collection instrument was developed, based on previous research and publications and management guidelines1–5. A spreadsheet was created with coding instructions (Supplementary File 2). In addition, data on the characteristics of the setting for each facility were collected in early 2017 by each investigator in consultation with local clinic staff, and entered onto an online survey form (Supplementary File 3). This included country and city, hospital or other healthcare facility, number of patients seen, and what diagnostic, treatment, referral and follow-up facilities are available. Clinic and patient data were de-identified to maintain confidentiality. Four clinics from three countries in sub-Saharan Africa chose to participate in the study.\n\nFor clinic BA, a public hospital clinic in northern Nigeria, patients were selected in sequence of attendance as new patients, starting January 1, 2013 for whom records were also available over the next two years. Patient demographics were those obtained at the first visit, and CD4 count and ART treatments were recorded at each visit, with analysis relating to data recorded at the most recent visit.\n\nFor clinic EV, a public hospital clinic also in Nigeria, all existing patients were re-tested with an ELISA method in 2013/14, and patients were randomly selected for this study among those who tested positive at that time. CD4 count and treatments recorded were those at the latest visit, and patient demographics were those in the records from their first visit.\n\nFor clinic MW, a public community-based clinic in Zambia, patients were randomly selected from those present on the patient registry in 2015/16. At that time, patients were reviewed for the need to start on highly active antiretroviral therapy (HAART) based on CD4 count, and the ART regime recorded was that started as a result of that review at that time. Data on patient demographics were those present in the records, often this would be prior to the date of entry to this study.\n\nFor clinic AM, a public hospital based clinic in Uganda, the sampling frame was the register of all clients on active ART, and every third patient file was retrieved from files ordered according to clinic appointment date. The CD4 counts and ART regimes recorded were for the most recent measures in the records, and patient demographics were those in the records from their first visit.\n\nData on individual patients were extracted from medical records, including age, gender, and baseline clinical data at diagnosis and follow-up. Clinical information included clinical, immunological and virological information on patients at diagnosis and on follow-up, prescription of ART and other drug choices, important co-infections and/or co-morbidities.\n\nDetails of the tests ordered and their results, and treatments ordered were extracted from medical records. Individual patients were not contacted. Data were obtained from the records, and no attempt was made to validate the information. Missing data were recorded as missing, and not explored further.\n\nA pilot study tested the feasibility of the data collection and the method of recording on to a spreadsheet. Data were collected by a research assistant in each setting, using the spreadsheet, from examination of individual records.\n\nAs a retrospective patient records study, consent was not requested from individual patients. Ethics approval was sought and obtained in each setting from the appropriate authority (see Ethical approval and consent section). Considerable care was taken not to reveal the identity of any individual and all data were de-identified. The spreadsheet for the recording of data only had an identification number, and the key to the identity of each patient was kept separately to maintain confidentiality. To further this, the clinics names and exact locations have been removed.\n\nEach centre was asked to obtain information from at least 100 patients.\n\nDescriptive statistics were used to characterise the study population. Data distributions were assessed, checking for skewness and kurtosis. Data summary statistics were generated. Categorical variables were summarized using proportions while continuous variables were summarized using medians and interquartile ranges. Due to differences in data extraction across sites, no statistical analyses comparing across sites was conducted. All analyses were conducted in Stata 14.1 (StataCorp, College Station, X) and R version 3.3.0 (The R Foundation for Statistical Computing).\n\n\nResults\n\nTen alumni expressed interest in participating, three centres were able to pilot the data collection instrument and four centres in three countries participated in the data collection for the study. Overall, data were abstracted for 600 patients.\n\nTable 1 shows the characteristics of the 4 clinics. The number of patients seen per month varied from 500 to 4200, and the number of doctors, nurses, and allied health professionals available at the clinics varied from 1 to 50, 10 to 65, and 6 to 45 respectively. There was some variation in the stated frequency of testing for CD4 count (either every three or six months) and for measurement of viral load. Each clinic had availability of clinical guidelines, and most drugs were available free of charge to patients. All clinics have access to ART and co-trimoxazole.\n\nART antiretroviral therapy\n\nIDV indinavir\n\nATV atazanavir\n\nTable 2 shows the patient characteristics and test results. Age, gender and body mass index (BMI) were similar between clinics, with overall median age of 31 years, 62% were female and median BMI was 20. The proportion of patients with a CD4 count in the records varied from 84 to 100 percent, and the latest median count varied from 269 to 593 between the clinics. 209 (35%) had a record of a viral load test with 81% of then having a viral load of less than 50 copies/µL. Only 23 (19%) of 119 patients had records of AIDS defining illnesses. Figure 1 shows the variation in the distribution of the CD4 counts in each of the four clinics.\n\n1Missing BMI data: AM: 3/300, BA: 81/100, EV: 100/100, MW: 14/100 (total 198/600)\n\nTable 3 shows the treatment regimes. Lamivudine (3TC) was recorded for more than 90% of patients in each clinic, but otherwise there was considerable variation between clinics in the choice of ART. Tenofovir disoproxil fumarate (TDF) was prescribed for 317 (53.9%) of studied patients. Efavirenz (EFV) was prescribed for 326 (55%) of the patients, making combination of TDF/3TC/EFV the most common ART combination used by studied patients. In two clinics, 100% and 99% of patients were prescribed co-trimoxazole prophlaxis, compared with 7% and no patients in the two other clinics. There were 6 patients on lopinavir and 8 on atazanavir (second line treatment choices).\n\n\nDiscussion\n\nData were collected in a standardised way, with data on treatment regimes and CD4 counts relating to the latest information in the records (dates ranging from 2015 to 2017). While we initially planned to collect data on a cohort of patients enrolled two years before data collection in each clinic, this did not prove feasible and patients had presented to the clinics at variable times.\n\nThis study found modest documentation of clinical activities, with a wide variation in available clinic health work force, frequency of CD4 counts and levels and viral load assessment. While most patients had a CD4 count in the records, there was a wide spread of latest CD4 counts, with both within- and between-centre variation in the CD4 counts. Only one in three patients had a record of viral load at any time during the course of treatment, and most of these had an undetectable level (<50 copies/µL). Some of the absence of viral load might be poor record keeping, and the high proportion of undetectable results might reflect an absent rather than a low test result, and hence not an indicator of treatment success. Over the years the CD4 threshold for commencing ART had been lowered, as can be seen in Table 4, with the most current guideline requiring a “test and treat approach” ie eliminating CD4 threshold as a prerequisite for commencement of ART5. While this guideline is based on sound evidence, it adds an unanticipated number of potential eligible patients for ART care, with attendant strain on countries with an already fragile economy. The low frequency of viral load test results might also reflect resource limitations.\n\nART antiretroviral therapy\n\nAZT zidovudine : ZDV retrovir\n\n3TC lamivudine\n\nEFV (EFZ) efavirenz\n\nNVP nevirapine\n\nNRTI Nucleotide analog reverse transcriptase inhibitor\n\nTDF tenofovir disoproxil fumarate\n\nFTC emtricitabine\n\nNNRTI Non-nucleoside reverse transcriptase inhibitor\n\nHBV Hepatitis B virus\n\nLPV lopinavir\n\nMost patients were on TDF/3TC/EFV drug regimen, which is consistent with current guideline advice for first line treatment. Only 2% of the patients were on second-line ART, varying from 9% to less than 1% between clinics. It is likely that most of the studied patients are still on a potent first line ART judging by the low second-line ART usage. Use of a regimen with a low pill burden would enhance therapy adherence and might lead to reduced need for switching patients to second line ART10. TDF/3TC/EFV drug regimens have a low pill burden, and may contribute to the low usage of second-line ART regimens.\n\nThe variation in the use of co-trimoxazole, from almost universal in two clinics, to negligible in the other two clinics, is an extreme example of the variation we identified. The reason for this variation despite the guideline recommendation is not clear, but it is possible that it is a result of local clinic policy settings, possibly related to concerns about increasing antibiotic resistance. Variation in the use of co-trimoxazole has been found in other resource-limited settings, while efforts are being made to improve the rates of use11.\n\nWhile each clinic reported having access to treatment guidelines, those current at the time of the data collection do not appear to have been universally followed. The guidelines also do change regularly, as shown in Table 4. For example in relation to the frequency of CD4 count monitoring and the CD4 count threshold at which treatment should be started have changed since the time relating to the study data - new guidelines recommend starting treatment regardless of the CD4 count5,12. Some discrepancy between WHO guidelines and actual implementation in practice may arise from the time it takes to implement new guidelines, or due to lack of resources to immediately initiate all patients with HIV on ART. We see considerable between-clinic variation in a number of key management strategies reported by the clinics, from the recording of CD4 counts, median CD4 counts on treatment, and treatments used. This variation is consistent with a previous survey of stated management practices in 6 sub-Saharan countries13,14.\n\nThe management of patients in routine clinical care has been shown to differ from that seen in clinical trials15, as well as to lead to worse clinical outcomes, although it is beyond the scope of our study to explore outcomes. Findings of variation from standard management practice has previously been reported from routine care settings in other parts of sub-Saharan Africa including Ethiopia16, Uganda17 and Tanzania18, mostly in single isolated centres. Here we present combined data on health system related measures across multiple sub-Saharan African HIV treatment sites.\n\nIn the absence of standardised record keeping systems, it is difficult to make clear comparisons of management and outcome in routine clinical care. Our findings suggest that there is room for further work to understand the reasons for these record gaps, and to standardise the record keeping in routine care of HIV positive patients. The potential of electronic medical records to improve records could be explored.\n\nOur study has a number of limitations. First, this study was an analysis of data extracted from existing medical records, which are prone to error including missing data. We have found missing data where it should ordinarily not be missing. We cannot comment on other factors that may have contributed to missing data, such as whether tests were not done or if they were not recorded. Second some of the variation between centres may be due to differences in patient populations which we were unable to capture, even though we selected clinics involved in routine clinical care. Third, participation in the study was restricted to few countries, and to individual clinics, which may not be representative of the national picture in these countries. A prospective cohort study on representative samples would provide a more robust study design with more detailed quantitative data to delineate care dynamics and to explore clinical outcomes.\n\n\nConclusions\n\nWe demonstrate a wide variability in compliance with HIV treatment guidelines in four settings in sub-Saharan Africa, as well as gaps in the records available. The findings of this study may provide an explanation for heterogeneous HIV treatment outcomes across sub-Saharan Africa. In spite of the limitations, these data underscore the need for an in-depth study to address compliance with HIV treatment guidelines and best practice. While electronic medical record implementation might be a challenge for many HIV care points in sub-Saharan Africa, our findings emphasize the need for more robust interim paper-based medical record keeping.\n\n\nEthics approval and consent\n\nEthics approval was sought in each setting from the appropriate authority. For two of the centres, research ethics committees gave approval. In two of the centres, the ethics committees stated that they did not require formal approval from them, however approval to access records was obtained from the relevant District Health Office/r. Details for each clinic as follows:\n\nClinic BA: Ethics approval obtained from Aminu Kano Teaching Hospital Research Ethics Committee.\n\nClinic EV: Ethics approval obtained from Research and Ethics Committee of State House Medical Centre, Abuja.\n\nClinic MW: University of Zambia Biomedical Research Ethics Committee contacted and advised to notify the Lusaka District Health Office (LDHO) who gave approval.\n\nClinic AM: Eastern Uganda AIDS Support Organization (TASO) contacted who recommended no need for an approval but rather write to the Amuria District Health Officer (DHO) for permission to have access to the hospital records, which was given.\n\nInformed consent was not obtained from individual patients since this was a records study with appropriate institutional approval, strict confidentiality arrangements, and no patient contact, as stated in the manuscript.\n\n\nData availability\n\nDataset 1. De-identified data collected from clinical records used to create Table 2 and Table 3 and Figure 1. Scheme used to code the data is available as Supplementary File 2. 10.5256/f1000research.15169.d20633319\n\nAll data underlying the survey responses from clinic informants for Table 1 are available as part of the article.",
"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\nThanks to Rebecca Oyomopito for help with designing the data collection instrument, Ado Abubakar for help with piloting the instrument, and Krishna Shrinivas for help through Statistics Without Borders.\n\n\nSupplementary material\n\nSupplementary File 1 – Study protocol\n\nClick here to access the data.\n\nSupplementary File 2 – Data coding scheme\n\nClick here to access the data.\n\nSupplementary File 3 – Written from of online survey\n\nClick here to access the data.\n\n\nReferences\n\nWorld Health Organisation: Scaling up antiretroviral therapy in resource-limited settings: guidelines for a public health approach. September 2002. Reference Source\n\nWorld Health Organisation: Antiretroviral therapy for HIV infection in adults and adolescents: Recommendations for a public health approach (2006 revision). 2006. Reference Source\n\nWorld Health Organisation: Antiretroviral therapy for HIV infection in adults and adolescents: Recommendations for a public health approach: 2010 revision. Reference Source\n\nWorld Health Organisation: Consolidated Guidelines on the Use of Antiretroviral Drugs for Treating and Preventing HIV Infection: Recommendations for a Public Health Approach. 2013. PubMed Abstract\n\nWorld Health Organisation: Consolidated Guidelines on the Use of Antiretroviral Drugs for Treating and Preventing HIV Infection: Recommendations for a Public Health Approach. Second edition. 2016. PubMed Abstract\n\nPetersen M, Balzer L, Kwarsiima D, et al.: SEARCH test and treat study in Uganda and Kenya exceeds the UNAIDS 90-90-90 cascade target by achieving 81% population-level viral suppression after 2 years. AIDS. 2016; Durban, South Africa. Reference Source\n\nAyele TA, Worku A, Kebede Y, et al.: Choice of initial antiretroviral drugs and treatment outcomes among HIV-infected patients in sub-Saharan Africa: systematic review and meta-analysis of observational studies. Syst Rev. 2017; 6(1): 173. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSlaymaker E, McLean E, Wringe A, et al.: The Network for Analysing Longitudinal Population-based HIV/AIDS data on Africa (ALPHA): Data on mortality, by HIV status and stage on the HIV care continuum, among the general population in seven longitudinal studies between 1989 and 2014 [version 1; referees: 2 approved, 1 approved with reservations]. Gates Open Res. 2017; 1: 4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHeller RF, Machingura PI, Musa BM, et al.: Mobilising the alumni of a Master of Public Health degree to build research and development capacity in low- and middle-income settings: The Peoples-uni. Health Res Policy Syst. 2015; 13: 71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSahay S, Reddy KS, Dhayarkar S: Optimizing adherence to antiretroviral therapy. Indian J Med Res. 2011; 134(6): 835–849. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBardfield J, Agins B, Palumbo M, et al.: Improving rates of cotrimoxazole prophylaxis in resource-limited settings: implementation of a quality improvement approach. Int J Qual Health Care. 2014; 26(6): 613–22. PubMed Abstract | Publisher Full Text\n\nAbdool Karim SS: Overcoming Impediments to Global Implementation of Early Antiretroviral Therapy. N Engl J Med. 2015; 373(9): 875–876. PubMed Abstract | Publisher Full Text\n\nChurch K, Machiyama K, Todd J, et al.: Identifying gaps in HIV service delivery across the diagnosis-to-treatment cascade: findings from health facility surveys in six sub-Saharan countries. J Int AIDS Soc. 2017; 20(1): 21188. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAmbia J, Renju J, Wringe A, et al.: From policy to practice: exploring the implementation of antiretroviral therapy access and retention policies between 2013 and 2016 in six sub-Saharan African countries. BMC Health Serv Res. 2017; 17(1): 758. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLópez-Martínez A, O'Brien NM, Caro-Vega Y, et al.: Different baseline characteristics and different outcomes of HIV-infected patients receiving HAART through clinical trials compared with routine care in Mexico. J Acquir Immune Defic Syndr. 2012; 59(2): 155–60. PubMed Abstract | Publisher Full Text\n\nAlemayehu YK, Bushen OY, Muluneh AT: Evaluation of HIV/AIDS clinical care quality: the case of a referral hospital in North West Ethiopia. Int J Qual Health Care. 2009; 21(5): 356–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBurua A, Nuwaha F, Waiswa P: Adherence to standards of quality HIV/AIDS care and antiretroviral therapy in the West Nile Region of Uganda. BMC Health Serv Res. 2014; 14: 521. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMapunjo S, Urassa DP: Quality standards in provision of facility based HIV care and treatment: a case study from Dar es Salaam Region, Tanzania. East Afr J Public Health. 2007; 4(1): 12–8. PubMed Abstract\n\nMusa BM, Ibekwe E, Mwale S, et al.: Dataset 1 in: HIV treatment and monitoring patterns in routine practice: a multi-country retrospective chart review of patient care. F1000Research. 2018. Data Source"
}
|
[
{
"id": "37430",
"date": "03 Sep 2018",
"name": "Thomas Crellen",
"expertise": [
"Reviewer Expertise Epidemiology",
"regression modelling",
"pathogen 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 article by Musa and colleagues uses a snapshot of patient records from four HIV clinics in Nigeria, Zambia and Uganda to describe the variability in a number of patient measures. Data are not reported longitudinally and all analysis is descriptive.\nThe most useful part of this report is to show that ordinary clinics in countries with some of the highest HIV burdens globally do not confirm with WHO guidelines. For instance, only commencing ART when CD4 cells <350/ul. Further, the high amount of missing data in viral loads suggest there are logistical or cost challenges to routine testing.\n\nIt is perhaps unsurprising that clinics in different countries vary in their patient characteristics and levels of reporting. The variability in CD4 counts is not particularly informative as it is not linked to length of time under ART. No research questions are addressed.\nTable 2 should be amended as some percentage values are missing and figures are presented to different numbers of significant figures.\nDespite the descriptive nature of the study, the authors are to be commended on presenting hard to access data from routine health care settings in sub-Saharan Africa. I hope that they will be able to take longitudinal data in the future and better understand how clinical practice in these settings is linked to patient outcomes.\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?\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? Yes",
"responses": [
{
"c_id": "3995",
"date": "26 Sep 2018",
"name": "Richard F Heller",
"role": "Author Response",
"response": "Thank you for your positive comments on the paper. In the revised version submitted, we have tried to emphasise the points you make, and have amended Table 2 as requested."
},
{
"c_id": "4333",
"date": "04 Jan 2019",
"name": "Richard F Heller",
"role": "Author Response",
"response": "In a later revision of the paper, paragraph 1 in the Methods section has been revised as follows to include the research question: “Interest in participating in this study was sought among the Masters of Public Health Alumni of Peoples-uni. A protocol was developed by those who responded (Supplementary File 1) which included the Research questions as follows: 1. Among a group of patients diagnosed with HIV by a health care facility newly diagnosed between 2 and 3 years previously, what proportion have standards of care in terms of the tests and treatments they receive and follow up to 2 years after diagnosis documented in their records? 2. Among the health care facilities in which the above patients are cared, what treatment, testing and referral facilities are available? 3. What is the extent of the variation in the above measures between facilities and countries, as well as in the availability of appropriate evidence based practice guidelines? A retrospective patient cohort was created via standardized record review, and clinic characteristics determined through an online survey form.”"
}
]
},
{
"id": "37307",
"date": "17 Sep 2018",
"name": "Sehlulekile Gumede-Moyo",
"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 paper is well written, however in some instances the authors have used outdated references. In the case of Nigeria where there are 2 health facilities with different caseloads per month, the authors could also enlighten us on the impact of having a high and low case load as presented.\nThe authors should also enlighten us on whether the records indicate the dates when the CD4 count and viral load tests were taken and when they were reported back. I suppose this of great importance since, there are reports of patients who have their blood samples taken and never receive results.\n\n1) Is The study design appropriate and is the work technically sound? The study design wand the work is technically, as the authors endeavoured to analyse multi-country settings\n2. Are sufficient details of methods and analysis provided to allow replication by others? The methods and analysis were sufficient, however in the case of Nigeria where there are 2 health facilities with different case loads per month; the authors could also enlighten us on the impact of having a high and low case load as presented.\n3. If applicable, is the statistical analysis and its interpretation appropriate? The statistical analysis was appropriate, although the authors should also enlighten us on whether the records indicate the dates when the CD4 count and viral load tests were taken and when they were reported back. I suppose this of great importance since, there are reports of patients who have their blood samples taken and never receive results.\n4. Are the conclusions drawn adequately supported by the results? The authors however need to address the issues that have been raised above.\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? No source data required\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": [
{
"c_id": "3994",
"date": "26 Sep 2018",
"name": "Richard F Heller",
"role": "Author Response",
"response": "Thank you for your generally positive comments. In the revised version submitted, we have added four recent references, and given further details on the dates of CD4 counts and viral load tests (although we have no data on the time that the results were reported to the patients). We have also discussed the reasons that the two Nigerian clinics might have different caseloads and the potential impact of this."
}
]
}
] | 1
|
https://f1000research.com/articles/7-713
|
https://f1000research.com/articles/7-1850/v1
|
27 Nov 18
|
{
"type": "Research Article",
"title": "Feasibility of incorporating mindfulness based mental health promotion to the pregnancy care program in Sri Lanka: a pilot study",
"authors": [
"Thilini Agampodi",
"Subhashini Katumuluwa",
"Thulani Pattiyakumbura",
"Nilupulee Rankaduwa",
"Thushari Dissanayaka",
"Suneth Agampodi",
"Subhashini Katumuluwa",
"Thulani Pattiyakumbura",
"Nilupulee Rankaduwa",
"Thushari Dissanayaka",
"Suneth Agampodi"
],
"abstract": "Background: Though widely discussed, mindfulness-based interventions (MBI) to improve maternal mental health is limited by lack of studies with system incorporation. We evaluate the feasibility of incorporating a MBI program into routine antenatal care (ANC) in Sri Lanka. Methods: MBI included learning mindfulness concepts, practicing mindfulness sitting/reclining meditation, performing mindful movements and practicing mindfulness in daily life. Feedback from the participants were obtained through an anonymous, self-administered, semi-structured questionnaire to determine the program’s cultural appropriateness, usefulness, and feasibility. Results: Participants reported that the training reduced the stress of their daily life, brought a sense of calmness to their mind and body, and improved their anger management. Participants felt strongly that this training would be very useful and a shortened version be included in the national ANC program. Conclusions: A systematic assessment of impact of MBI is needed with system incorporation of the suggested programme.",
"keywords": [
"Maternal mental health",
"Mindfulness",
"Pregnancy",
"Antenatal care",
"Sri Lanka"
],
"content": "Introduction\n\nPregnancy and childbirth represent a time of increased vulnerability, during which a woman is exposed to many physiological and psychosocial changes. This puts pregnant and postpartum women at increased risk of mental health problems. The prevalence of mental health disorders in pregnancy and postpartum period is much greater in low-income countries (LIC) and lower-middle-income countries (LMIC) (Fisher et al., 2012) than in high income countries (HIC) (Hendrick et al., 1998). Of the mental health problems experienced by pregnant and postpartum women, antenatal depression, anxiety, and post-partum depression are the most common (O’Hara & Swain, 1996). These mental health issues may lead to adverse pregnancy and neonatal outcomes such as small for gestational age baby (Dejin-Karlsson et al., 2000), developmental delays (Bernard-Bonnin, 2004), poor mother-infant interaction (Cohn & Tronick, 1989), negative affect in the infant (Tronick & Reck, 2009), problems with cognitive development (Singer & Fagen, 1992) and affective disorders /attention-deficit/hyperactivity disorder (ADHD) in children (Lesesne et al., 2003). This makes the antenatal period a critical time point for intervention on maternal mental health.\n\nDespite having a strong, community-based public health care system, addressing maternal mental health has yet to gain the spotlight in Sri Lanka. The prevalence estimate for antenatal depression in Anuradhapura District, Sri Lanka is 16% (Agampodi & Agampodi, 2013) and the national estimate for postpartum depression is as high as 27% (Agampodi et al., 2011). Furthermore, a secondary analysis of maternal death investigation reports from 2005–2011 in the North Central Province revealed that suicide was the leading cause of maternal mortality in the province, with 17.8% of maternal deaths attributed to suicide (Agampodi et al., 2014). These findings show the urgency of addressing the mental health of pregnant women in Sri Lanka.\n\nMindfulness is an emerging concept in mental health promotion. It was the original healing method used for centuries in ancient cultures in Asia. Clinical use of mindfulness was widely discussed in late 20th century (Baer, 2003; Deatherage, 1975). Several recent studies have shown that mindfulness-based interventions for pregnant women have been effective in increasing positive affect, decreasing negative affect (Duncan & Bardacke, 2010; Vieten & Astin, 2008a), decreasing anxiety (Duncan & Bardacke, 2010; Dunn et al., 2012; Vieten & Astin, 2008a), decreasing depression (Duncan & Bardacke, 2010; Dunn et al., 2012), and decreasing stress (Dunn et al., 2012) during pregnancy. These benefits were seen to extend into the postpartum period in some cases (Dunn et al., 2012; Tomfohr-Madsen et al., 2016; Vieten & Astin, 2008a). Nevertheless, two recent systematic reviews (Dhillon et al., 2017; Roy Malis et al., 2017) shows mixed results on mindfulness based interventions in pregnancy. One of these reviews found that “anxiety, depression and perceived stress indicated no differences between the mindfulness intervention group and the control group” in pooled results of RCTs. However, pooled results of non-RCTs showed a “significant benefit for the mindfulness group”. Other review also showed similar results with probable effect on maternal anxiety. Both reviews suggested that more work is needed on this area of research to produce evidence-based guidelines. However, studies on integrating the mindfulness-based activities to routine program to improve maternal and newborn health outcomes are scared. We report here a feasibility study on developing and incorporating a mental health promotion program based on mindfulness for pregnant women of Sri Lanka.\n\n\nMethods\n\nThis preliminary study was carried out in Anuradhapura, Sri Lanka. The study was planned as a part of field healthcare delivery in the field practice area of the Department of Community Medicine, Faculty of Medicine and Allied Sciences, Rajarata University of Sri Lanka. In the first phase of this study, we translated and culturally adapted the Mindful Motherhood training curriculum and in the second phase, feasibility testing was done to see whether it is possible to incorporate this training to the public health system. Both phases of the study were carried out in from May 2015 to April 2016.\n\nPhase I of this study was to develop a culturally acceptable mindful motherhood training curriculum. We used the “Mindful Motherhood” training curriculum (Vieten, 2009), published by the Institute of Noetic Sciences (IONS) and made available for professional use for our training programme. The general structure and content of the IONS training was left the same; however, some of the reading materials and activities were modified to be more relevant and acceptable to a Sri Lankan population. In addition, supplemental readings were added from other sources. In the Sri Lankan community, “loving kind medication” is a well-established practice. We included supplementary reading and practice component of loving kind medication in to all eight sessions. The sources used were simple Buddhist teaching, without addition religious components. The changes were done after extensive discussions among three community physicians, a public health physician, maternal health service providers at divisional and grassroots level and medical officers. All supplemental readings were translated from English to Sinhalese, the main language spoken in Sri Lanka. Translations were performed by two medical graduates, who were fluent in both languages as well as mindfulness teaching. Tamil translations were not made at this time, as the majority population of the Nuweragam Palath Central (NPC) region, where the study was conducted, speaks Sinhalese. Back translation was not done, as the procedure was rather a cultural adaptation than tool translation. However, three investigators individually looked at the translations and discussed and finalized the materials.\n\nThe aim of our programme was to introduce the concept of mindfulness and develop the skills of the pregnant mothers to be mindful in day-to-day life, with the ultimate goal of promoting mental well-being. During the programme development, four authors (T.A., S.K., T.P. and N.R.) who did the training, had undergone mindful training. Three of the authors involved in the training (T.A., S.K. and N.R.) were regular practitioners of Buddhist mindfulness in daily life. Our curriculum consisted of eight, 2- to 3-hour weekly training sessions. Each session consisted of several components; introduction of a short mindfulness concept; practice of mindfulness sitting/reclining meditation (i.e. breathing meditation, meditation with focus on thoughts/feelings/sensations, and several types of body scans); performing mindful movement (i.e. prenatal yoga, mindful walking, simple mindful movement series); and sometimes a mindfulness in daily life activity or discussion (i.e. mindful eating, how to complete daily tasks mindfully, mindfulness in relationships). In addition to the weekly sessions, each week the women were provided with readings summarizing or related to the week’s mindfulness concepts. The initial 8-week curriculum, supplementary materials, and all training related documents are publically available on OSF (Agampodi et al., 2018a).\n\nIn the second phase, we field-tested the feasibility of incorporating a mindfulness-training program into routine antenatal care (ANC) in Sri Lanka. For the intervention, we recruited pregnant women at less than 32 weeks gestational age, and who were able to read and write Sinhalese, from the NPC region. Those who were not permanent residents and planning to leave the area for delivery (in Sri Lanka, some females go to their parents’ place for delivery) were excluded from the study. Participants were recruited at the antennal clinics. We also included the Public Health Midwife (PHM), the grassroots level healthcare provider for pregnant women in the program. The PHM was included because she has extensive experience working with pregnant women in rural areas and has a unique understanding of their needs and perspectives on a range of pregnancy-related issues, providing valuable information regarding the feasibility of our training. In addition, if any intervention to be included into the routine public health system in Sri Lanka, it needs to be carried out by the PHM. The PHM selected the participants for the study purposefully, based on interest and availability to participate. Pregnant women with a history of mental disorder with a psychotic component were excluded from the study. The sample size was predetermined (not calculated) as 12–15 participants, to ensure an adequate level of interaction between the participants and instructors and based on feasibility to visit the mindfulness-training center at Faculty of Medicine and Allied Sciences.\n\nIn addition to attending the weekly training sessions developed in phase 1, as a homework assignment, women were asked to practice each of the mindfulness practices learned in the session at home. The women recorded their practice on a weekly record sheet and handed it in the following week. They were also asked to keep a diary and write any thoughts or feelings about the sessions and how they felt during the week while practicing. After seven weeks of the program, participants were given an anonymous, detailed, self-administered, semi-structured questionnaire to obtain their feedback. The questionnaire sought to gain the participants’ detailed insights into how the program affected their lives (if at all), the most useful aspects of the training, their suggestions for improvement, and whether and how a similar training should be incorporated into the current ANC program. We ask about acceptability of the programme and also feasibility of practicing it at home. No maternal mental health outcome variables were evaluated in this pilot, as it was only a feasibility study.\n\nTo evaluate the feasibility and to triangulate data for future system incorporation, facilitators of the training also recorded their thoughts on the program content, structure, and the logistics of conducting. The self-reported descriptions were collected separately to analyses and thematic analysis was done for these feedbacks as well.\n\nEthical clearance was obtained from the Ethics Review Committee, Faculty of Medicine and Allied Sciences, Rajarata University of Sri Lanka (ERC/2015/15). Written informed consent was obtained from all the participants included in the study.\n\n\nResults\n\nIn total, 12 women were recruited and participated in the training. The women ranged in age from 18 to 30 years, and were between 10 and 27 weeks gestational age. After completion of the program, participants described many benefits they received through the training. A total of eight participants completed the diaries and the feedback.\n\n\nPerceived psychological benefits\n\nMost of the women (7/8) reported that the training reduced daily life stressors and brought a sense of calmness to their mind. All the women stated that participating in the training had also improved the way they responded to stressful situations.\n\n“I was bored and stressed with the daily routine. With the training I was able to face day-to-day problems in a calm mindful manner.” 22-year-old pregnant female.\n\nThe women reported improvements in interpersonal communication. The training had helped them to improve the way they interacted with their spouse and/or current child(ren), or that their view of how best to interact with their yet unborn child had changed based on what they had learned.\n\n“When my elder child was stubborn, I was able to modestly tackle it. Even when other people came to argue, I was able to face them with patience. It was successful.” 30-year-old pregnant female.\n\nSome women also reported that the training helped them to control their emotions, with several (3 of 8) stating that they were better able to control their anger.\n\n“My elder child had a fall and was injured. All the family members were panicked. I was able to remain calm and do the needful. I was not like this before. This was a new experience for me. I acquired this skill from the training.” 28-year-old pregnant female.\n\nThe women told that they were able to do their duties at home and work place in a more efficient manner than in their past.\n\n“Now I’m able to do each and every task with more understanding. This gives me clarity and calmness in mind. I feel less exhausted. I’m able to sleep well.” 18-year-old pregnant female.\n\n\nPerceived physical benefits\n\nIn addition to the mental and emotional benefits, all the women reported physical benefits, including feeling a sense of comfort and relaxation for the body when practicing the mindful movement at home. One of the women reported successfully using several of the meditation and body scan techniques help her cope with an experience of abdominal pain, stating,\n\n“when I fell ill and I started practicing the things we learned, putting them into action one by one, the body which was full of pain became normal in an unbelievable way.” 25-year-old pregnant female.\n\nEvery participant reported that they intended to continue the daily practice of various techniques to cultivate mindfulness, learned during the training.\n\n\nEvaluation of content\n\nParticipants evaluated the different components or strategies used to cultivate mindfulness during this training. The mindful movement series was the most popular strategy among women. Of the mindfulness concepts taught, the women found the topics of “introduction to mindfulness,” “staying in the present moment,” “acceptance,” and “avoiding controlling and judging” to be the most interesting and helpful in their daily lives. Of the sitting/lying down meditations, women enjoyed and perceived the most usefulness from the body scan, loving kindness meditation, and mindful breathing practice. Everyone reported that they found the “Mindful Mother Check-in” cards (cards given to them at the beginning of the training with simple steps to bring the mind to the present moment applying the practices they had learned), which they posted around their home, to be helpful to calm their mind, particularly during stressful situations. Of the mindful movements taught, the majority of the women (6 of 8) enjoyed the short mindful movement series the most.\n\nIn addition to participant feedback, the facilitators of the training also recorded their thoughts on the program content, structure, and the logistics of conducting it. We experienced that the depth of the material being allocated for each session was advance, so that it cannot be grasped with just a simple explanation for a participant who missed a session. For this same reason, we found it a challenge to complete all the material in the allotted 2-hour time, often running over. Additionally, the women seldom asked questions about the material (which may be a cultural factor) or for clarification during the sessions.\n\n\nFeasibility and timing\n\nEach participant felt time allocated for each of the training segments was appropriate, except for one who felt it would be better if more time was given for the mindful movement section. Although we took into consideration the demands on the average pregnant woman in our area when deciding a meeting time, we still found that it was very difficult to get the women to come consistently to every session. This is likely due to several cultural factors. In Sri Lanka, many women in more rural areas are stay-at-home wives/mothers, taking care of the household and the children. So unless there is someone at home, usually a relative, who can watch the children or pick them up from school while the mother is out, she may not be able to come, may come late or need to leave early, which we experienced on multiple occasions. All of the women also had to take public transportation to get to the trainings, which added to the challenge.\n\n\nFeasibility of incorporation into ANC system\n\nWomen were asked their thoughts on the potential benefit of incorporating a mindfulness training into routine ANC in Sri Lanka. All the participants felt that this training would be very useful and important for other pregnant women. While agreeing with this, one woman raised the concern of whether one two-hour session every three months would provide sufficient time and guidance for women to fully understand the meaning and feel the value of the concepts taught. It was also suggested that it would be beneficial to have the husbands of the pregnant women involved in the program as well. Overall, all the participants, including the PHM, felt that it would be beneficial and feasible to incorporate a mindfulness program into the national ANC system if the current curriculum was shortened and included in the once per trimester antenatal sessions that are already part of the ANC system in Sri Lanka.\n\nBased on Phase I and Phase II of this feasibility assessment, we prepared the final version of intervention. The final structure of the mindfulness based intervention to improve mental health of pregnant mothers is illustrated in Table 1.\n\n\nDiscussion\n\nThis preliminary study shows that a incorporating a mindfulness based mental health program for pregnant mothers to the pregnancy care program in Sri Lanka is feasible and culturally acceptable. It also suggests that learning mindfulness concepts and the techniques to cultivate mindfulness may improve a woman’s ability to cope with stress during her pregnancy. Important future steps should include efforts to capture the latter outcome quantitatively, as has been done in prior pilot studies (Duncan & Bardacke, 2010; Dunn et al., 2012; Vieten & Astin, 2008a).\n\nThe program developed in this pilot study was based on the standard 8-week mindfulness intervention used in many other previous studies (Astin et al., 2003)(Byrne et al., 2014)(Dunn et al., 2012)(Vieten & Astin, 2008b). However, a brief mindfulness interventions even for a 3-week duration has proven the potential usefulness of mindfulness interventions (Beattie et al., 2017; Matvienko-Sikar & Dockray, 2017). Although there are number of studies on mindfulness interventions, there is not enough evidence found about the most effective duration for a mindfulness program. Thus, further studies has to be conducted in the future to assess the effective time duration for mindfulness interventions. Taking into account the challenges the facilitators experienced during the intervention, trying to cover all the material in a comprehensible manner within the allotted 16-hour time period of the training may not be feasible. Careful consideration will be needed to select some of these concepts that could practically be covered during the three ANC sessions, and would likely be of greatest benefit and/or interest to the pregnant women. The feedback from our pilot study participants will be valuable when making these decisions.\n\nGiven the positive response from our small feasibility pilot, the next steps will include using the participants’ feedback to create a modified, shortened mindfulness training curriculum for pregnant women, and running a larger field study using this revised curriculum. Maternal mental health outcomes should also be evaluated using pre- and post-intervention questionnaires that have been validated for use in Sri Lankan populations. This will be an important step in establishing any effect that the mindfulness training may have on maternal mental health-related outcomes, including distress, depression, and coping skills. The possibility of any benefits of the training extending into the postpartum period, as suggested by pilot studies (Roy Malis et al., 2017; Vieten & Astin, 2008a) should also be evaluated. The degree of success of launching such a program nationally, or even locally, will depend heavily on the willingness and enthusiasm of the PHMs, as eventually, if the program is to be sustainable, it will be trained PHMs who will be leading each of the mindfulness sessions. To date, the support from the local PHM community and Medical Officer of Health (MOH) has been strong for this program. When presented an overview of the program and the evidence behind the teachings, many of the PHMs were visibly enthusiastic about the benefits such a program could have for their patients and were eager to offer their support.\n\nLimitations of our study include the small sample size, the inconsistent attendance of some of the participants, and the sample of only Sinhalese-speaking women. To serve both the Sinhalese and Tamil populations in Sri Lanka, the curriculum will also need to be translated and tested in a Tamil-speaking population.\n\n\nConclusion\n\nThis pilot study suggests that a mindfulness-training program is a culturally appropriate and feasible way to incorporate promotion of maternal mental health into routine ANC in Sri Lanka. Further research must investigate the effectiveness of a shortened, modified program on impacting maternal mental health-related outcomes during pregnancy and beyond.\n\n\nData availability\n\nRaw data from the participants’ answers to the questionnaires are available on OSF. Please note that data are provided in Sinhalese. DOI: https://doi.org/10.17605/OSF.IO/QFA6X (Agampodi et al., 2018b).\n\nAll documents related to the project, including the initial proposal, ERC certificate, questionnaire (with translation) and training material are available on OSF. DOI: https://doi.org/10.17605/OSF.IO/45HNR (Agampodi et al., 2018a).\n\nAll data 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\nThis study was funded by the University grants of Rajarata University of Sri Lanka under the grant number RJT/RP and HDC/2015/FMAS/R/06.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nAgampodi SB, Agampodi TC: Antenatal depression in Anuradhapura, Sri Lanka and the factor structure of the Sinhalese version of Edinburgh post partum depression scale among pregnant women. PLoS One. 2013; 8(7): e69708. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgampodi SB, Agampodi TC, Katumuluwa S: Mindful Motherhood Training Materials [WWW Document]. 2018a. http://www.doi.org/10.17605/OSF.IO/45HNR\n\nAgampodi SB, Agampodi TC, Katumuluwa S: Mindful Motherhood Data [WWW Document]. 2018b. http://www.doi.org/10.17605/OSF.IO/QFA6X\n\nAgampodi TC, Agampodi SB, Wickramasinghe WA, et al.: Post partum depression - a problem that needs urgent attention. Ceylon Med J. 2011; 56(4): 183–4. PubMed Abstract | Publisher Full Text\n\nAgampodi S, Wickramage K, Agampodi T, et al.: Maternal mortality revisited: the application of the new ICD-MM classification system in reference to maternal deaths in Sri Lanka. Reprod Health. 2014; 11(1): 17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAstin JA, Berman BM, Bausell B, et al.: The efficacy of mindfulness meditation plus Qigong movement therapy in the treatment of fibromyalgia: a randomized controlled trial. J Rheumatol. 2003; 30(10): 2257–2262. PubMed Abstract\n\nBaer RA: Mindfulness training as a clinical intervention: A conceptual and empirical review. Clin Psychol-Sci Pr. 2003; 10: 125–143. Publisher Full Text\n\nBeattie J, Hall H, Biro MA, et al.: Effects of mindfulness on maternal stress, depressive symptoms and awareness of present moment experience: A pilot randomised trial. Midwifery. 2017; 50: 174–183. PubMed Abstract | Publisher Full Text\n\nBernard-Bonnin AC: Maternal depression and child development. Paediatr Child Health. 2004; 9(8): 575–598. PubMed Abstract | Free Full Text\n\nByrne J, Hauck Y, Fisher C, et al.: Effectiveness of a Mindfulness-Based Childbirth Education pilot study on maternal self-efficacy and fear of childbirth. J Midwifery Womens Health. 2014; 59(2): 192–197. PubMed Abstract | Publisher Full Text\n\nCohn JF, Tronick E: Specificity of infants’ response to mothers’ affective behavior. J Am Acad Child Adolesc Psychiatry. 1989; 28(2): 242–8. PubMed Abstract | Publisher Full Text\n\nDeatherage G: The Clinical Use of “Mindfulness” Meditation Techniques in Short-Term Psychotherapy. Journal of Transpersonal Psychology. 1975; 7: 133–143. Reference Source\n\nDejin-Karlsson E, Hanson BS, Ostergren PO, et al.: Association of a lack of psychosocial resources and the risk of giving birth to small for gestational age infants: a stress hypothesis. BJOG. 2000; 107(1): 89–100. PubMed Abstract | Publisher Full Text\n\nDhillon A, Sparkes E, Duarte RV: Mindfulness-Based Interventions During Pregnancy: a Systematic Review and Meta-analysis. Mindfulness (N Y). 2017; 8(6): 1421–1437. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDuncan LG, Bardacke N: Mindfulness-Based Childbirth and Parenting Education: Promoting Family Mindfulness During the Perinatal Period. J Child Fam Stud. 2010; 19(2): 190–202. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDunn C, Hanieh E, Roberts R, et al.: Mindful pregnancy and childbirth: effects of a mindfulness-based intervention on women’s psychological distress and well-being in the perinatal period. Arch Womens Ment Health. 2012; 15(2): 139–143. PubMed Abstract | Publisher Full Text\n\nFisher J, Cabral de Mello M, Patel V, et al.: WHO | Prevalence and determinants of common perinatal mental disorders in women in low- and lower-middle-income countries: a systematic review. Bull World Health Organ. 2012; 90(2): 139–149H. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHendrick V, Altshuler L, Cohen L, et al.: Evaluation of mental health and depression during pregnancy: position paper. Psychopharmacol Bull. 1998; 34(3): 297–9. PubMed Abstract\n\nLesesne CA, Visser SN, White CP: Attention-deficit/hyperactivity disorder in school-aged children: association with maternal mental health and use of health care resources. Pediatrics. 2003; 111(5 Pt 2): 1232–1237. PubMed Abstract\n\nMatvienko-Sikar K, Dockray S: Effects of a novel positive psychological intervention on prenatal stress and well-being: A pilot randomised controlled trial. Women Birth. 2017; 30(2): e111–e118. PubMed Abstract | Publisher Full Text\n\nO’hara MW, Swain AM: Rates and risk of postpartum depression—a meta-analysis. Int Rev Psychiatry. 1996; 8: 37–54. Publisher Full Text\n\nRoy Malis F, Meyer T, Gross MM: Effects of an antenatal mindfulness-based childbirth and parenting programme on the postpartum experiences of mothers: a qualitative interview study. BMC Pregnancy Childbirth. 2017; 17(1): 57. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSinger J, Fagen J: Negative affect, emotional expression, and forgetting in young infants. Dev Psychol. 1992; 28: 48–57. Publisher Full Text\n\nTomfohr-Madsen LM, Campbell TS, Giesbrecht GF, et al.: Mindfulness-based cognitive therapy for psychological distress in pregnancy: study protocol for a randomized controlled trial. Trials. 2016; 17(1): 498. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTronick E, Reck C: Infants of depressed mothers. Harv Rev Psychiatry. 2009; 17(2): 147–56. PubMed Abstract | Publisher Full Text\n\nVieten C: Mindful Motherhood Training: Guidelines for Facilitators. 2009.\n\nVieten C, Astin J: Effects of a mindfulness-based intervention during pregnancy on prenatal stress and mood: results of a pilot study. Arch Womens Ment Health. 2008a; 11(1): 67–74. PubMed Abstract | Publisher Full Text\n\nVieten C, Astin J: Effects of a mindfulness-based intervention during pregnancy on prenatal stress and mood: results of a pilot study. Arch Womens Ment Health. 2008b; 11(1): 67–74. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "41129",
"date": "17 Dec 2018",
"name": "Moraendage Wasantha Gunathunga",
"expertise": [
"Reviewer Expertise Prof. Wasantha Gunathunga: Body mind and consciousness",
"mental well being",
"happiness",
"mindfulness"
],
"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\nMaternal health is an important concern in Sri Lanka as in many other countries. Though maternal well being is addressed with vigor in Sri Lanka mental well being is not equally addressed. Hence, exploring possibilities of incorporating mindfulness into maternal care services is of much importance and has the potential to contribute significantly to scientific literature. It is commendable that authors embarked on this much needed research.\nIn phase one, what the authors mean by “Buddhist teaching, without addition religious components” is unclear. If any use can be obtained, for health benefits from any religious material we do not see any reason of not using it and acknowledging it.\nAmong those who took part in extensive discussions no mention is made about experience in actually practicing mindfulness. The fact that three researchers have been trained in mindfulness and are regular practitioners is noteworthy. This is essential as mindfulness involves a phenomenological component. Without substantial direct experience researchers will not understand certain components of behaviours related to mindfulness. It is also required to mention whether the researchers are capable of dealing with advanced practitioners’ queries or such services are made available when needed. This is important as all mothers who start mindfulness practice are not at the same baseline and progress is not uniform. They can be expected to behave very close to a normal distribution where some progress very fast and needing advanced help.\nIn phase two “Pregnant women with a history of mental disorder with a psychotic component were excluded from the study”. It is unclear why only psychotic component was excluded. It would have been more appropriate to exclude those who have a known psychiatric illness in this phase.\nIt has been mentioned that mental health variables were not assessed. However, perceived psychological benefits were assessed and positive. It is too early to count on these as the benefits cannot be directly attributed to the mindfulness practice in this design. This association, though, is noteworthy.\nThe sample size of 12 is inadequate to have a sufficient idea on the feasibility of the intervention due to the wide variation in the variable related to feasibility.\n\nAs mentioned in “Feasibility of incorporating into Antenatal Care(ANC) system” Participants opinion regarding incorporating this intervention to national antenatal care in Sri Lanka has to be taken with reservation as they do not see the operational aspects of such a program when it is scaled up. In the responses there is also an element of social desirability bias.\n“The degree of success of launching such a program nationally, or even locally, will depend heavily on the willingness and enthusiasm of the Public Health Midwive(PHM)s, as eventually, if the program is to be sustainable, it will be trained PHMs who will be leading each of the mindfulness session”. This is asking too much from already over burdened PHM in Sri Lanka. Our experience is that, out of the numbers who take up mindfulness training, only a small proportion will qualify to be trainers. Hence we consider planning to train PHMs as trainers is a weakness in the design.\nAs mentioned in the discussion inconsistent attendance can be a difficult challenge to meet as take-up of this kind of interventions by different individuals vary widely.\nIt is still premature to conclude “..a feasible way to incorporate promotion of maternal mental health into routine ANC in Sri Lanka”. Though it appears feasible to implement the intervention whether it improves mental well being as an outcome has not been assessed in this work. Hence this claim is not justified.\nIn the abstract, there is no conclusion in relation to the set objective of evaluating feasibility of intervention.\nOverall, considering the dearth of literature in this area of research and the importance in relation to promotion of maternal mental well being this work warrants indexing with modifications.\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?\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": "4327",
"date": "02 Jan 2019",
"name": "Suneth Agampodi",
"role": "Author Response",
"response": "Reply to reviewer’s comments We would like to appreciate the comments by the reviewer and it definitely helped us to enhance the quality of this manuscript. Based on the points raised by the reviewer, we have revised the manuscript. The point by point explanations for reviewers comments are below. In phase one, what the authors mean by “Buddhist teaching, without addition religious components” is unclear. If any use can be obtained, for health benefits from any religious material we do not see any reason of not using it and acknowledging it. In the phase one we had several meetings with stakeholders, before starting the project. Service providers had a concern that this intervention looks as a Buddhist intervention. To overcome this barrier, we mentioned the term “mindfulness” though the teaching and training were obtained from Buddhist practices. We acknowledge the materials in the article, but not during the training. Among those who took part in extensive discussions no mention is made about experience in actually practicing mindfulness. The fact that three researchers have been trained in mindfulness and are regular practitioners is noteworthy. This is essential as mindfulness involves a phenomenological component. Without substantial direct experience researchers will not understand certain components of behaviours related to mindfulness. It is also required to mention whether the researchers are capable of dealing with advanced practitioners’ queries or such services are made available when needed. This is important as all mothers who start mindfulness practice are not at the same baseline and progress is not uniform. They can be expected to behave very close to a normal distribution where some progress very fast and needing advanced help. Of the six authors, three were practicing mindfulness for a long period. However, it is difficult to reply to this comment about needing advanced help. The purpose of this training was to improve day to day mindful practices and to help the pregnant mothers to practice it up to a certain extent. It was neither our intention to go up to the advance mindful practices nor in assisting pregnant mothers who are already having such practices, because it was beyond any routine intervention. In phase two “Pregnant women with a history of mental disorder with a psychotic component were excluded from the study”. It is unclear why only psychotic component was excluded. It would have been more appropriate to exclude those who have a known psychiatric illness in this phase. In the research protocol, we had this as an exclusion criteria, but none of the selected pregnant mothers had psychiatric illnesses. We revised the manuscript to explain this. It has been mentioned that mental health variables were not assessed. However, perceived psychological benefits were assessed and positive. It is too early to count on these as the benefits cannot be directly attributed to the mindfulness practice in this design. This association, though, is noteworthy. We fully agree with this comment. We got a qualitative feedback and a proper trial is required to comment on the benefits. We have included this suggestion in the discussion.The sample size of 12 is inadequate to have a sufficient idea on the feasibility of the intervention due to the wide variation in the variable related to feasibility. This also we agree fully. This was a “pilot study’ to assess feasibility as mentioned in the title. The study is not even a full feasibility study. We emphasize this in the revised limitation component of the manuscript. As mentioned in “Feasibility of incorporating into Antenatal Care (ANC) system” Participants opinion regarding incorporating this intervention to national antenatal care in Sri Lanka has to be taken with reservation as they do not see the operational aspects of such a program when it is scaled up. In the responses there is also an element of social desirability bias. Again we agree with this comment fully and included in the discussion. “The degree of success of launching such a program nationally, or even locally, will depend heavily on the willingness and enthusiasm of the Public Health Midwive (PHM)s, as eventually, if the program is to be sustainable, it will be trained PHMs who will be leading each of the mindfulness session”. This is asking too much from already over burdened PHM in Sri Lanka. Our experience is that, out of the numbers who take up mindfulness training, only a small proportion will qualify to be trainers. Hence, we consider planning to train PHMs as trainers is a weakness in the design. First, we also share the same concern about overburden PHM duties. At the same time, in Sri Lanka, national maternal health intervention through other means has never had sustainability. The second point is well taken and only a small proportion will be qualify to be mindfulness trainers; not only among PHMs but also among other groups. This limitation and challenge was added to the revised discussion. As mentioned in the discussion inconsistent attendance can be a difficult challenge to meet as take-up of this kind of interventions by different individuals vary widely.Agree and it's already in the discussion. It is still premature to conclude “..a feasible way to incorporate promotion of maternal mental health into routine ANC in Sri Lanka”. Though it appears feasible to implement the intervention whether it improves mental well being as an outcome has not been assessed in this work. Hence this claim is not justified. Agree and this conclusion was revised.In the abstract, there is no conclusion in relation to the set objective of evaluating feasibility of intervention.Conclusion of the abstract is revised."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1850
|
https://f1000research.com/articles/6-192/v1
|
27 Feb 17
|
{
"type": "Software Tool Article",
"title": "Automated Wormscan",
"authors": [
"Timothy Puckering",
"Jake Thompson",
"Sushruth Sathyamurthy",
"Sinduja Sukumar",
"Tirosh Shapira",
"Paul Ebert",
"Timothy Puckering",
"Jake Thompson",
"Sushruth Sathyamurthy",
"Sinduja Sukumar",
"Tirosh Shapira"
],
"abstract": "There has been a recent surge of interest in computer-aided rapid data acquisition to increase the potential throughput and reduce the labour costs of large scale Caenorhabditis elegans studies. We present Automated WormScan, a low-cost, high-throughput automated system using commercial photo scanners, which is extremely easy to implement and use, capable of scoring tens of thousands of organisms per hour with minimal operator input, and is scalable. The method does not rely on software training for image recognition, but uses the generation of difference images from sequential scans to identify moving objects. This approach results in robust identification of worms with little computational demand. We demonstrate the utility of the system by conducting toxicity, growth and fecundity assays, which demonstrate the consistency of our automated system, the quality of the data relative to manual scoring methods and congruity with previously published results.",
"keywords": [
"WormScan",
"Caenorhabditis elegans",
"toxicology",
"software",
"phosphine"
],
"content": "Introduction\n\nSeveral techniques exist for digitisation and computational analysis of Caenorhabditis elegans, with the highest throughput techniques being those that utilise photo-grade flatbed scanners for image acquisition. The first of these, WormScan1, lacks post-acquisition automation of data processing, whereas the second, The Lifespan Machine2, is more complicated to implement and is computationally intensive. We present a very simple to implement, low-cost system based on automating elements of WormScan, which can score C. elegans on agar plates for mortality, size, and fecundity at a rate of tens of thousands of worms per hour with very little operator input.\n\nC. elegans is the premier model organism for ageing and toxicological research. They are among the most intensively studied organisms of the last 50 years, which has resulted in the development of powerful genetic and molecular research tools. These tools include an extensive collection of readily available mutant strains3, comprehensive and commercially available RNAi libraries4, and a large collection of strains that each contain a GFP tagged gene. These strains provide multiple avenues by which genes of toxicological importance can be efficiently investigated, either individually or in combination. Additionally, the size, simplicity, and fecundity of the organism make it a popular choice for toxicological studies5–7. Furthermore, C. elegans have a very short natural lifespan, making them ideal for studies into the mechanisms of lifespan determination8–11. The ability to assay the worms in large numbers makes it possible to address the immense combinatorial challenge of testing large chemical libraries to identify bioactive compounds against a large numbers of strains.\n\nThere has been a recent surge of interest in rapid data acquisition to increase the potential throughput and reduce the labour costs of large scale C. elegans studies. In 2012, Mathew et al. developed WormScan, a low cost analysis tool that uses high resolution flatbed scanners and computer analysis to collect mortality data from C. elegans on petri dishes1. This system replaces microscopy with commercially available transmission flatbed scanners, and replaces the mechanical stimulus of a platinum wire probe with the aversive stimulus of the bright light from the scanner, which was demonstrated to induce comparable responses. However, while WormScan employed computer image processing of scanned images to replicate human counting, the system was not automated. The 'Lifespan Machine', a more sophisticated, higher throughput system was then developed, which, like WormScan, relies on machine learning to identify worm-like objects2. In its current implementation, this system requires sophisticated IT support.\n\nHere we present an automated software system, Automated WormScan. It is extremely easy to implement and use. It employs low-cost hardware that can be set up by a person with average computer skills in under an hour, and can be used by any researcher with typical computer expertise. Any number of scanners can be run simultaneously from a single standard desktop PC. Three scanners permit an operator to count over ten thousand individual organisms per hour, with only approximately ten minutes of actual operator labour. By comparison, traditional microscope-based counting methods often count fewer than 500 individuals per hour of uninterrupted labour.\n\nThe automated system does not depend on machine learning for image recognition, but rather relies on an alternative worm counting method of WormScan, the generation of a difference image from sequential scans, with moving objects scored as live worms. This approach results in robust identification of worms with little computational demand. We demonstrate the utility of the system by conducting toxicity, growth and fecundity assays that demonstrate the consistency of our automated system, the quality of the data relative to manual scoring methods and congruity with previously published results. For this analysis, we use fumigant phosphine as a toxic stress as our lab has studied the toxicology of this compound quite extensively6,7,12–16. We do, however, routinely use the system to study the toxic or protective effects of a wide range of other gases and dissolved compounds as well.\n\n\nMethods\n\nAs with both WormScan and the recent Lifespan Machine system, an Epson v700 (or v800) photo scanner is used to take transmission scans of C. elegans grown on a thin lawn of Escherichia coli on agar growth medium in 6cm petri dishes1,2. In order to minimise the possibility that worms will be in physical contact at the time of scanning and accidentally counted as a single worm, up to 100 worms are scanned per plate. Scans are initially acquired as .tiff images in 16 bit greyscale at 2400 dpi, but are then saved as condensed jpeg images. This ensures sufficient pixel count to be able to distinguish small differences between sequential images, while minimising file size. Up to 12 populated petri dishes can be scanned at a time in each scanner, and all are scanned simultaneously. This is followed immediately by a second scan, each scan taking approximately ten minutes. While all experiments in this study used three scanners simultaneously, there is no absolute limit to the number of scanners that could be utilised in parallel.\n\nThe software for automated analysis of the images exists as two components; a plugin for the FIJI image processing application17, and a macro for Microsoft Excel. The FIJI plugin automates a sequence of image manipulations that generates a reliable count of the number of light-responsive individual organisms on each agar plate. The Excel macro then automates the extraction, labelling and concatenation of these counts into a spreadsheet.\n\nWorm maintenance was carried out according to the standard procedures in Wormbook18. Briefly, worms were grown on solid nematode growth medium seeded with E. coli OP50, and maintained at 20 degrees. Populations of nematodes were synchronised by dissolving gravid adult worms in 1.5% hypochlorite bleach and 0.75M NaOH to release their eggs. Eggs were immediately rinsed in M9 buffer and maintained overnight at room temperature in M9 buffer with gentle shaking to allow the eggs to hatch. The resulting L1 larvae were then transferred to NGM agar plates seeded with OP50 to initiate growth18.\n\nPhosphine was generated as described in Cheng et al., 20067 by dissolving commercially available ammonium phosphate tablets ('Quickphos'; UPL ltd.) in 5% sulphuric acid in a glass Valmas Chamber. The resultant phosphine was then quantified using a phosphine gas monitor (Canary ‘SmarTox-O’ PLZZ) and injected through a septum into a sealed glass desiccator, to achieve the concentration required by the parameters of each experiment.\n\nWorms were grown on NGM agar plates seeded with OP50 at 20°C for 48 hours, by which time they had reached the early L4 stage of development. Prior to exposure to phosphine, the number of live worms on each plate was determined by the Automated Wormscan procedure described in this paper. These plates were then sealed inside the desiccators prior to phosphine injection and were exposed to phosphine for 24 hours. Worms were then removed from the chambers and permitted to recover for 48 hours, at which time the number of surviving worms on each plate was determined by Automated Wormscan.\n\n\nResults\n\nThe required software for Automated WormScan can be downloaded as an easy to install complete software package available at doi, 10.5256/f1000research.10767.d15269719. Installation and running of the software are fairly intuitive, but detailed instructions can be found in the comprehensive user guide (Supplementary File 1). The software is based on the FIJI image analysis software package, version 1.49m. If FIJI is already installed, the plugins and macros can be added to the current installation, as described in the user guide (Supplementary File 1). A macro designed to work with Microsoft Excel is also provided to assist with data compilation and formatting. The software has been developed to run on a computer running Windows 7 or higher. The minimum system requirements are low, with the program confirmed to run on a desktop PC with an Intel Core 2 processor and 4GB ram.\n\nA set of twelve 6 cm plates are scanned as described in Methods and in the user guide (Supplementary File 1). Computational analysis of the images begins with the flipping of the images around the vertical axis to make the plate locations correspond intuitively to their locations on the scanner. This image is then split into 12 individual images corresponding to each individual plate. The second scanned image of the same 12 plates is likewise split into 12 individual images. Each of the 12 individual images from the first scan is paired with the corresponding images from a second scan. These paired images are then precisely aligned using image stabilization (Kang Li. 2008. The image stabilizer plugin for ImageJ. Available: http://www.cs.cmu.edu/~kangli/code/Image_Stabilizer.html) so that each pixel can be compared for a difference in intensity between the two images. The outcome is a greyscale difference image, in which the whiteness intensity of each pixel corresponds to the degree to which that pixel differed in intensity between the first and second scan. This is then converted to true black-and-white with the use of an edge detection function (Boudier T, Meys J. 2015. Edge Detection. Available: http://imagejdocu.tudor.lu/doku.php?id=plugin:filter:edge_detection:start). Particle analysis of this image is then performed, which generates a count of areas on the image that differ between the two scans, as well as circumference measurements for each of these. Size parameters are pre-set to exclude identified areas that are too large or too small to be L4 nematodes (>6mm, <0.4mm in diameter), such as refractive areas of the petri dish or dust particles. It is possible to alter the pre-set parameters to identify and count worms of other developmental stages.\n\nThe worm identification, measurement and counting functions are fully automated, so will proceed until an entire folder of scanned images has been processed. A difference image with worm-like features highlighted is created for each plate and is also archived to allow visual confirmation of results and as a permanent record of the experiment (Figure 1). The processing of a large number of scans (>30) requires several hours. However, a human operator is not required during this entire period, so it is convenient to run the process overnight.\n\n(A) A high-resolution scan of a petri dish containing a population of approximately 100 early L4 stage C. elegans individuals. (B) A difference image produced from a pixel-by-pixel comparison with a second image taken of the same plate ten minutes later. White regions correspond to a worm that moved from its original location during this time. (C) Regions of interest from the difference image that have worm-like size and shape attributes are identified. Archival images are created with an outline of identified worm-like objects overlaid on the original images.\n\nThe end product of the computational analysis is a spreadsheet for each individual plate, containing a series of measures of each area of difference (and hence, of each worm). These measurements include circumference, area and circularity, which are potentially useful for comparing worm size or developmental stage. An Excel macro has also been developed, which imports the data from the FUJI plugin into an Excel spreadsheet template (Supplementary File 2). This results in a table of raw data suitable for statistical analysis and graphing (Figure 2).\n\nA count of live worms from every scanned petri plate is extracted from the output of the FIJI plugin and is associated with its scan number and plate position. Researchers then manually enter experimental parameters to automatically fill the 3x4 grids with presumed plate labels by an auto-completion function. If the design of an experiment does not fit this default model, the labels can be inserted manually. The end result is a single table of raw data from the entire collection of images that can easily be tracked back to the archived images.\n\nAccuracy and consistency of Automated WormScan relative to counting by humans. The reliance of Automated WormScan on comparing the worm count prior to an experiment to the worm count after an experiment, simplifies the computational burden relative to employing machine learning to identify and assess mortality of worms from a single pair of post-experiment scans. While our procedure makes worm identification extremely robust, errors associated with both the pre- and post-experiment counts contribute to the final error of the analysis. This differs from human counting, as well as from the machine learning procedure originally described in WormScan and utilised by the Lifespan Machine, both of which count live and dead worms from a single set of scans.\n\nTo establish the error rate of the automated system, an assay was performed to count the number of live worms on agar plates using both traditional human counting and Automated Wormscan. In total, 12 plates of worms were counted in triplicate using a microscope, each time by a different trained human. The same plates were then scanned sequentially three separate times, each time in a different scanner, with counts performed using Automated WormScan (Figure 3). We performed a Generalised Linear Mixed-Effects Model20 in R (version 3.3.2)21. Overall, we found that counts produced by human observers vs counts produced by Automated Wormscan did not differ significantly (P>0.2), indicating that Automated Wormscan closely replicates human counting.\n\nVariability in numbers of individual worms counted by Automated Wormscan on the same plate is not significantly different from the variation in counts between three independent human observers using traditional methods. Error bars indicate standard deviation over three independent counts.\n\nWormscan is able to replicate previously published toxicological data. We replicated phosphine exposure protocols used by our laboratory to demonstrate that Automated Wormscan generates results consistent with published results derived from manual counting. Depending on the dose of exposure, toxins can impair mobility or alter the shape of the worms, but this does not materially affect the outcome of the analysis. We exposed wild type N2, and a phosphine resistant strain, dld-1(wr4), to a range of phosphine concentrations and obtained survival curves indicating LC50 values of 400ppm for N2 and 1800ppm for dld-1. These results, demonstrating a phosphine survival rate for dld-1(wr4) 4.5 times greater than wild type, are consistent with a previously published study that used manual phenotype scoring, which found the survival rate of dld-1(wr4) to be 4.4 times greater than wild type (Figure 4)16.\n\nAutomated Wormscan accurately duplicates previously published phosphine toxicology results. Wild type worms (N2) display an LC50 of ~400ppm, while the phosphine resistant mutant dld-1(wr4) display an LC50 of ~1800ppm, congruous with recently published results6. Error bars indicate standard deviation over three experimental replicates.\n\nAutomated wormscan can be used to measure growth rate and fecundity. While counting living, mobile and stimulus responsive worms is the primary objective of the system, the perimeter of each identified worm-like object is also determined during analysis. We therefore performed an assay to determine how well this data could reproduce published size difference of two different worm strains, N2 and daf-2, as daf-is well known to have a slow growth rate and delayed maturation relative to the wild type N2 strain22,23. Plates were scanned 48 hours and 90 hours after seeding with synchronised L1 worms. After 48 hours, no progeny were present on any of the plates, and the average perimeter data was recorded for each worm and length was calculated as 1/2 × perimeter. Using Automated WormScan, daf-2 worms were determined to be significantly shorter than wild type worms after 48 hours growth at 20°C, by an average of ~0.25mm (Figure 5). This reflects the well-established size difference of the daf-2 strain.\n\nLength was calculated as ½ of the perimeter measurement determined by Automated WormScan. Error bars represent standard deviation over three experimental replicates.\n\nAfter 90 hours at 20°C, both strains had produced progeny that were easily recognised by their size, with adults being well over 1.5mm in perimeter and juveniles being under 0.5mm. The rate of reproduction of each strain is presented as juveniles present per adult worm (Figure 6). The developmental stage of each individual was confirmed visually on the source image. After 90 hours, wild type worms produce approximately 4 times as many progeny as daf-2 worms, which is consistent with the well-established decrease in reproduction of the daf-2 strain.\n\nFecundity is expressed as juvenile-sized objects (<0.5mm perimeter) per adult-sized object (>1.5mm perimeter) as counted by WormScan. Error bars represent standard deviation over three experimental replicates (>600 juvenile individuals scored).\n\nAccuracy of automated wormscan using compressed image formats. We find that it is essential to scan at high resolution to obtain accurately identified worms, but this produces TIFF format image files of individual petri plates of ~60 MB. For the small experiment shown in Figure 4, the 14 data points were generated from 3 experimental replicates, each of which contained 2 technical replicates. As each plate is scanned a total of 4 times, a total of 336 images were generated. Given a TIFF image size of 60 MB, the total storage requirements if using high resolution images would be 20 GB. High resolution images also increase the computational processing demands.\n\nTherefore, we carried out an experiment to see if the performance of our system was compromised by use of compressed image formats. A series of plates of N2 worms in the early L4 stage of development were analysed by Automated WormScan in two different ways. In both cases, the worms were initially scanned as 16 bit greyscale images. In the first case, the images were saved in the lossless TIFF format (~60 MB) by the scanning software. In the second case, the images were saved as compressed high quality jpg images (~1.5 MB) by the scanning software. These images were then processed by Automated WormScan to compare the number of worms counted per plate in each image format. We found that analysing compressed jpg images had no effect on the number of worms that could be identified by the program (Supplementary File 3).\n\n\nDiscussion\n\nAutomated Wormscan is an extremely low cost system (scanner USD$500, 3TB internal hard drive $125, hard drive dock $25, scanner plate holder $50, standard PC with Windows operating system) capable of replacing manual counting of worm populations and scoring of survival. We have demonstrated that the system is capable of replicating previously published toxicological results, while also yielding greatly improved speed and throughput. The system is remarkably easy to set up, requiring less than 60 minutes. Once set up, very little operator time is required to carry out the image acquisition and analysis. An archival record is automatically stored as a compressed image file on which identified worms are highlighted. Accuracy of record keeping is enhanced by the ability to collect all image files for an experiment in a single folder and process them all at once, with the output being a single Excel spreadsheet of data with filenames and plate numbers transferred directly from the image files.\n\nC. elegans toxicological assays are often limited to only a handful of strains, treatments, time points and/or chemical concentrations, due to the inability of researchers to quickly score very large numbers of individual worms. The use of an automated mortality scoring system removes this significant bottleneck in the throughput of most toxicological assay designs. This greatly improves the speed at which such experiments can be performed and enables complicated experiments to be contemplated, such as the simultaneous determination of the combinatorial effects of many substances at a variety of concentrations. Furthermore, the increased throughput enables the use of larger numbers of technical replicates and extra experimental replicates, which provides greatly improved statistical power that previously would have been impractical due to the labour required.\n\nAutomated WormScan is one of several systems recently developed that utilise image capture and computer-based phenotype assessment to improve the throughput or accuracy of such C. elegans assays1,2. Automated WormScan uses very inexpensive and simple hardware, requiring only a basic desktop PC and at least one Epson v700 (or v800) scanner and a plate holder to align petri plates on the scanner (LabPro Scientific; part LPST12-1). Automated WormScan requires no modifications or alterations to the scanners and the software is also especially easy to install and simple to use. Setup can be accomplished in under an hour by anyone with basic computer skills. In addition, scans are taken while the worms are on an NGM agar plate, which is the typical format for most C. elegans assays. This means the scanning can be incorporated into existing experimental methods without requiring alterations to the experimental procedure.\n\nSince the system focuses on the ability of the animals to move under stimulus, the assay is very robust, with a demonstrated ability to consistently count live worms with the same accuracy as a human. The robustness of the system is achieved largely by avoiding the machine learning steps integrated into the primary component of WormScan, in favour of the simpler difference image analysis method that was originally developed for monitoring survival in lifespan experiments1. This design decision increases the amount of handling required compared to the machine learning strategies of the original WormScan and Lifespan Machine systems1,2 (i.e. two pairs of scans must be taken instead of one). In practice, this requirement is not onerous. Automated Wormscan is not capable of resolving multiple worms in close contact. In practice, this is not usually a problem, as clearly demonstrated in Figure 1. The solution is simply to use ≤100 worms per plate, in which case the simplicity of the Automated WormScan software does not significantly impinge on the ability of the system to replicate human counting. Automated WormScan will not work with mutants or experimental conditions that induce aggregation of worms, but the same is true of the other counting systems and probably human counting as well.\n\nAutomated WormScan is time and labour efficient, requiring ~10 minutes of operator time per 60 plates (5 scanners) and 20 minutes of scanning time (during which an operator need not be present). Once all scans have been completed and saved to a single folder, another 5 minutes of operator time is required to start the analysis software. The software requires about 30 seconds to analyse each pair of images (but this requires no operator involvement). Thus 6,000 individual C. elegans (at a density of 100 per plate) can be analysed with 15 minutes of operator time and about 60 minutes of elapsed time. The total capacity of the system is limited only by the number of scanners that are used in parallel to generate the source images, so the system can be scaled to suit the requirements of the user.\n\nUser bias in methods that involve manual counting is a known source of data variation24, especially between institutions25. The use of this automated method eliminates this possibility while simultaneously improving consistency. Furthermore, the use of the light stimulus obviates the need to open the plates and physically touch the animals, which minimises the potential for contamination and removes the possibility of physically damaging the animals.\n\nWhile the lifespan machine utilises modified scanners to reduce temperature fluctuations, the focus of Automated WormScan for mortality based toxicology assays means that worms do not spend much time in the environment of the scanner and are able to be housed in temperature controlled conditions for the great majority of time, obviating the need to modify the scanners to compensate for temperature. Also, while Lifespan Machine requires modifications to the scanner to adjust the instruments focal plane, the robustness of difference imaging produces accurate counts of worm populations using entirely unmodified scanners.\n\nSince compressed image formats do not hamper the ability of Automated WormScan to accurately quantify C. elegans populations, we utilise compressed .jpg files, rather than the much larger TIFF images utilised by other systems1,2. This obviates the need for greater than terabyte data storage for most applications, and brings the system specifications required to run previously computer resource intensive image processing phases into the range of a standard modern desktop PC.\n\n\nSoftware and data availability\n\nAutomated Wormscan source code: doi, 10.5256/f1000research.10767.d15269719\n\nLicense: GNU General Public License\n\nDataset 1: Raw worm counts for Figure 3 - comparison of worm counts produced by three different human observers vs three independent scans. doi, 10.5256/f1000research.10767.d15269826\n\nDataset 2: Raw worm counts for Figure 4 - phosphine mortality curves of wild type and resistant strains of Caenorhabditis elegans. doi, 10.5256/f1000research.10767.d15269927\n\nDataset 3: Raw worm counts for Figure 5 - growth rate of daf-2 worms compared with wild type. doi, 10.5256/f1000research.10767.d15270028\n\nDataset 4: Raw worm counts for Figure 6 - reduced fecundity at 90 hours of daf-2 worms compared to wild type. doi, 10.5256/f1000research.10767.d15270129",
"appendix": "Author contributions\n\n\n\nTP: Conceived and designed the automation project, performed all assays, drafted the manuscript, aided in programming, troubleshooting, and testing. JT: Performed the majority of the programming, assisted with some assays, assisted with troubleshooting, contributed substantially to the conception and design of the work. SS, SS and TS: Assisted with some assays, assisted with troubleshooting, performed extensive beta-testing, contributed substantially to the conception and design of the work. PRE: Contributed substantially to the conception and design of the work, substantially revised and edited the manuscript.\n\n\nCompeting interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nAustralian Postgraduate Award scholarship to T. Puckering. Plant Biosecurity Cooperative Research Centre Grant (PBCRC63119) to T. Puckering.\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\nDrafting of this manuscript was supported by Plant Biosecurity Cooperative Research Centre Grant, PBCRC63119 to T. Puckering. The research for the project was carried out under an Australian Postgraduate Award scholarship also to T. Puckering. The C. elegance and bacterial strains were provided by the Caenorhabditis elegans Genetic Stock Center, which is funded by the NIH Office of Research Infrastructure Programs (P40 OD010440).\n\n\nSupplementary material\n\nSupplementary File 1: Automated Wormscan hardware and software installation and setup guide. A step-by-step guide to installing and setting up all hardware and software necessary to run Automated Wormscan. Includes troubleshooting information.\n\nClick here to access the data.\n\nSupplementary File 2: Automated Wormscan Excel Mastersheet.\n\nClick here to access the data.\n\nSupplementary File 3: TIF scans vs JPG scans. Data is live worm counts produced by Automated Wormscan from 12 plates scanned first in JPG and then in TIF file format.\n\nClick here to access the data.\n\n\nReferences\n\nMathew MD, Mathew ND, Ebert PR: WormScan: a technique for high-throughput phenotypic analysis of Caenorhabditis elegans. Lehner B, editor. PLoS One. Public Library of Science; 2012; 7(3): e33483. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStroustrup N, Ulmschneider BE, Nash ZM, et al.: The Caenorhabditis elegans Lifespan Machine. Nat Methods. Nature Publishing Group, a division of Macmillan Publishers Limited. All Rights Reserved.; 2013; 10(7): 665–70. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThompson O, Edgley M, Strasbourger P, et al.: The million mutation project: a new approach to genetics in Caenorhabditis elegans. Genome Res. 2013; 23(10): 1749–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKamath RS, Fraser AG, Dong Y, et al.: Systematic functional analysis of the Caenorhabditis elegans genome using RNAi. Nature. 2003; 421(6920): 231–7. PubMed Abstract | Publisher Full Text\n\nZhou KI, Pincus Z, Slack FJ: Longevity and stress in Caenorhabditis elegans. Aging (Albany NY). 2011; 3(8): 733–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchlipalius DI, Valmas N, Tuck AG, et al.: A Core Metabolic Enzyme Mediates Resistance to Phosphine Gas. Science. 2012; 338(6108): 807–810. PubMed Abstract | Publisher Full Text\n\nCheng Q, Valmas N, Reilly PE, et al.: Caenorhabditis elegans mutants resistant to phosphine toxicity show increased longevity and cross-resistance to the synergistic action of oxygen. Toxicol Sci. 2003; 73(1): 60–5. PubMed Abstract | Publisher Full Text\n\nKaletsky R, Murphy CT: The role of insulin/IGF-like signaling in C. elegans longevity and aging. Dis Model Mech. 2010; 3(7–8): 415–9. PubMed Abstract | Publisher Full Text\n\nArtal-Sanz M, Tavernarakis N: Mechanisms of aging and energy metabolism in Caenorhabditis elegans. IUBMB Life. 2008; 60(5): 315–22. PubMed Abstract | Publisher Full Text\n\nAntebi A: Genetics of aging in Caenorhabditis elegans. PLoS Genet. 2007; 3(9): 1565–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGallo M, Riddle DL: Regulation of metabolism in Caenorhabditis elegans longevity. J Biol. 2010; 9(1): 7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nValmas N, Ebert PR: Comparative toxicity of fumigants and a phosphine synergist using a novel containment chamber for the safe generation of concentrated phosphine gas. Fox D, editor. PLoS One. Public Library of Science; 2006; 1(1): e130. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCha'on U, Valmas N, Collins PJ, et al.: Disruption of iron homeostasis increases phosphine toxicity in Caenorhabditis elegans. Toxicol Sci. 2007; 96(1): 194–201. PubMed Abstract | Publisher Full Text\n\nSchlipalius DI, Chen W, Collins PJ, et al.: Gene interactions constrain the course of evolution of phosphine resistance in the lesser grain borer, Rhyzopertha dominica. Heredity (Edinb). 2008; 100(5): 506–516. PubMed Abstract | Publisher Full Text\n\nValmas N, Zuryn S, Ebert PR: Mitochondrial uncouplers act synergistically with the fumigant phosphine to disrupt mitochondrial membrane potential and cause cell death. Toxicology. 2008; 252(1–3): 33–9. PubMed Abstract | Publisher Full Text\n\nZuryn S, Kuang J, Ebert P: Mitochondrial modulation of phosphine toxicity and resistance in Caenorhabditis elegans. Toxicol Sci. 2008; 102(1): 179–86. PubMed Abstract | 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\nStiernagle T: Maintenance of C. elegans. WormBook. 2006; 1–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPuckering T, Thompson J, Sathyamurthy S, et al.: Dataset 1 in: Automated Wormscan. F1000Research. 2017. Data Source\n\nBates D, Mächler M, Bolker B, et al.: Fitting Linear Mixed-Effects Models Using lme4. J Stat Softw. 2015; 67(1): 1–48. Publisher Full Text\n\nR Core Team: A Language and Environment for Statistical Computing [Internet]. Vienna Austria R Foundation for Statistical Computing. R Foundation for Statistical Computing, p. ISBN 3-900051-07-0. 2016. Reference Source\n\nKimura KD, Tissenbaum HA, Liu Y, et al.: daf-2, an insulin receptor-like gene that regulates longevity and diapause in Caenorhabditis elegans. Science. 1997; 277(5328): 942–6. PubMed Abstract | Publisher Full Text\n\nHonda Y, Honda S: The daf-2 gene network for longevity regulates oxidative stress resistance and Mn-superoxide dismutase gene expression in Caenorhabditis elegans. FASEB J. 1999; 13(11): 1385–93. PubMed Abstract\n\nBiston MC, Corde S, Camus E, et al.: An objective method to measure cell survival by computer-assisted image processing of numeric images of Petri dishes. Phys Med Biol. 2003; 48(11): 1551–1563. PubMed Abstract | Publisher Full Text\n\nLumley MA, Burgess R, Billingham LJ, et al.: Colony counting is a major source of variation in CFU-GM results between centres. Br J Haematol. 1997; 97(2): 481–484. PubMed Abstract | Publisher Full Text\n\nPuckering T, Thompson J, Sathyamurthy S, et al.: Dataset 2 in: Automated Wormscan. F1000Research. 2017. Data Source\n\nPuckering T, Thompson J, Sathyamurthy S, et al.: Dataset 3 in: Automated Wormscan. F1000Research. 2017. Data Source\n\nPuckering T, Thompson J, Sathyamurthy S, et al.: Dataset 4 in: Automated Wormscan. F1000Research. 2017. Data Source\n\nPuckering T, Thompson J, Sathyamurthy S, et al.: Dataset 5 in: Automated Wormscan. F1000Research. 2017. Data Source"
}
|
[
{
"id": "20569",
"date": "06 Mar 2017",
"name": "Elena M. Vayndorf",
"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 present a system, including easy-to-acquire equipment, software and instructions, for counting live, moving and stimulus responsive C. elegans. The authors also demonstrate that human counting vs. Wormscan system counting of animal survival produces similar results that are not significantly different. In addition, the Wormscan system is able to count worm size and live progeny number and these numbers are similar to human scoring.\n\nThe Wormscan system can be a useful tool for toxicological studies and teaching laboratory experiments. Overall, the authors have outlined methods for setting up the system, addressed its limitations and performed a set of proof of concept experiments that demonstrate the system’s novelty and utility. The Wormscan will make a useful teaching tool for high school and college laboratories or any laboratory looking to setup a simple, high-throughput survival in vivo assay.\n\nI have the following questions/suggestions for the authors to consider for revision (in no particular order):\n\nIn addition to the provided documentation, the authors should consider making a short video tutorial that guides the user step by step in data acquisition and analysis. What may seem «easy» or obvious, can sometimes be less so, especially when setting up the system for the first time.\n\nFig 3 - please clarify the age of the animals and add to figure legend.\n\nFig 4 - over what time period were these data collected over? Add to figure legend.\n\nFigure 6 - how many adults were used for the progeny assay?\n\nWhat happens at the very end of life/survival when the animals do not move at all but are still alive? How can the Wormscan system distinguish between living vs. dead animals? Please address this tail end of the curve. Or is this point moot because you are always comparing to the moving control group? If so, please clearly state this in your write-up. Also, in this regard, it would be interesting to determine if the Wormscan can pick up whether an animal is dead even when it is not moving on a petri dish by scanning over a longer period of time.\n\nIn regard to point in #5 above, beyond toxicological studies, do you envision this system for use in aging lifespan studies? Please address in the discussion.\n\nPerhaps I missed this - if the scan takes 10 minutes, are the animals exposed to a bright light for the entire duration of the time? Why does it take so long to make a single scan? Is it due to the high resolution necessary for image acquisition? In this regard, I am glad to see that overheating was not a problem, but if the time of the scan and overall stress level could be decreased, perhaps this system could be expanded to other types of experiments e.g. motility scoring? This could be an avenue to explore in the future, and perhaps this point could be addressed in the discussion?\n\nIs it possible to make this work with a Mac? Or does that depend on scanner compatibility? As it is clear that you have done a lot of testing of the scanner component of the Wormscan system, perhaps in the discussion, you could address the minimum scanner and PC requirements that are needed to set this up with other scanners beyond those that work on the Windows system? $500 may not seem like a lot for a scanner, but it may still be prohibitive for some researchers and teachers, so a thorough understanding of the requirements of the most expensive piece of the system, the scanner, would be most useful.\n\n«The required software for Automated WormScan can be downloaded as an easy to install complete software package available at doi, 10.5256/f1000research.10767.d15269719 I was not able to access this link (it was circular for me and took me back the main page, I tried it from multiple sources).\n\nPlease clarify in figure legends and/or methods how many times experiments were repeated and how many biological replicates were completed.\n\nConsider expanding the first paragraph of or supplying addition citation to your «Introduction» by doing a more thorough review of the technologies currently available for high-throughput image acquisition and phenotyping of C. elegans. For instance, see this review by Kinser and Pincus 2016: http://www.sciencedirect.com/science/article/pii/S104474311630046X",
"responses": []
},
{
"id": "20573",
"date": "06 Mar 2017",
"name": "Nathaniel Szewczyk",
"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\nThere is little doubt that some C. elegans researchers will find the software provided by the authors to be a valuable addition in automating analysis of C. elegans grown on standard C. elegans laboratory medium. The authors provide good validation data of the principal elements of the analysis for counting and assessing non-moving worms following toxic insult in Figures 3 and 4 which should give others confidence in the ability to employ this software.\n\nHowever, I found the narrative which I viewed as largely a “sales pitch” for the software was both unneeded for this journal and distracting. Similarly, I was quite disappointed at the lack details of the software, what data is captured, how one might modify the software in obvious ways of interest to C. elegans researchers, and what values returned from the software might represent with regards to C. elegans biology. While my specific comments below may seem extensive, they are largely and broadly summarized in the preceding two sentences (e.g. make the introduction and discussion more balanced and make the methods/results/case use more descriptive/detailed).\n\nSpecific points that should be addressed in a revision (listed more or less in the order that they came up reading the manuscript in the presented order):\nThis manuscript does not appear to comply with the software tool article template. Specifically, the Methods section is lacking both the implementation and operation sections. Additionally, the authors have opted for a Results section but mostly describe non-novel data and therefore, perhaps should have used a Use Cases section.\n\nThe Introduction should make some mention of other “digitisation and computational analysis of Caenorhabditis elegans.” It is certainly not the case that only the WormScan and The Lifespan Machine exist. Early efforts developed a “wormsorter”, I believe in the 1970s, which has now been commercialized as the COPAS1 and is fairly standard in many C. elegans labs. A quick google search returned loads of other efforts at automation a few examples: drug discovery2, gene expression3, and locomotion4. There are also an entire host of micro-fluidics devices that are continually being published and some reference to this is also probably warranted.\n\n“C. elegans is the premier model organism for ageing and toxicological research.” Needs to be changed, while I might agree with “is a premier model organism for…” it is certainly not the case that the toxicology field agrees with this statement as explained in a recent review on the worms’ role in toxicity testing5. NB. This review also explains, briefly, the current state of play for large scale toxicity screening and explains why axenic medium is preferable for toxicology screening.\n\nRelated to point (3), If you want to make argument about the utility of this analysis for toxicology, better references for the use of C. elegans as a model system for toxicology should be provided, especially as your paper doesn’t really go through toxicology endpoints in much detail. The above review is one example but there are other older excellent reviews as well.\n\nRelated to point (2), if you want to talk about the “recent surge” of interest in automated and high throughput experimentation and analysis on C. elegans you really need to cite other examples, particularly microfluidics which, frankly, is where the surge is.\n\nGiven that your analysis relies on worm movement you really need to be explicit about the limitation of using this device for ageing studies. Specifically, worms are immobile prior to death and this can last several days. Thus, your device will generate falsely truncated survival curves in lifespan assays.\n\nGiven that I had a new post-doc count roughly 150,000 worms in 75 min last week I don’t find the number of animals that can be processed with this system vs. a manual system at all convincing. If you want to argue about throughput I would stick to the number of conditions that can be analysed as being where the saving is (we only looked at 5 conditions in that 75 min).\n\n“Cheng et al., 20067” should be “2003”\n\nA better description of what the plugins and macros “do” and “how they do it” should be included, probably in an implementation section in methods.\n\nRelated to point (9) there needs to be enough detail such that someone who wants to modify the software can easily do so. For example, what if someone wants to image different size petri dishes or even 24 well tissues culture?\n\nRelated to point (9) there needs to be more explanation of what the size parameters are likely to represent for example what likely corresponds to an L1 (can you image an L1?), and L2 (can you image an L2?), an L3, and L4, and adult, etc.\n\nRelated to point (9) and (11), what are circumference, area, and circularity indicative of? What do different values likely mean, what are the strengths and limitation of each measurement with regards to individual worms?\n\nCheck your worm count in Figure 1, the legend says approximately 100 early L4 stage worms but a quick manual count (prompted by a quick visual estimate of more than 100) suggests closer to 200. If this is closer to 200 and you are making statements about 100 worms being optimal for maximal reliability either provide a better example image or revise your statements to reflect 200 (in which case additional details of how you get the approximate number that you put down should be provided, for example are you counting worms, visually estimating, automatically putting down counted worms (as the COPAS could do)).\n\nPlease provide details of what parameters were used to generate the worm count in Figure 3. Are these likely to be L1, L2, L3, L4, Adult, some combination? And did visual conformation match the expectation for animal stage?\n\nSimilar to (14) please provide details of what parameters were used to generate the survival curve and what stage these are likely to represent and if visual conformation matched expectations. (As written your paper suggests you dosed L4 animals and they remained L4 animals but this does not match your lab’s past work so I suspect this is a lack of clarity in this paper).\n\nProvide details for what strain of daf-2 was used and temperature of cultivation.\n\nSimilar to (14) please provide details of what parameters were used to generate this length data and how this compares with life stage and visual confirmation. In both instances the estimated lengths seem too short for the populations to be entirely adults, this should be reflected in the title (perhaps Size after 48 hr growth from L1 is a better title?). NB. You refer to perimeter in the figure legend and results section but perimeter is not a parameter that is otherwise mentioned in the description of the software.\n\nFigure 5 does not provide a growth rate as a single data point is provided. I suggest simply removing “rate” from the title may fix this problem.\n\nThe well-established size difference of daf-2 strains is for width (for example Depuydt et al.6) yet you are claiming your length data reflects this well established difference. Please provide a reference for how your length data match past published literature for daf-2 length at a comparable growth point.\n\nIn the results you state “adults being well over 1.5mm in perimeter and juveniles being under 0.5mm.” This does not seem consistent with past published literature where L4 and young adult animals are similar in length with juveniles being in a somewhat linear distribution of smaller sizes down to the size of a newly hatched L1 and adults being in a somewhat linear distribution of larger sizes up until mid-late adulthood. Please clarify/correct this statement. N.B. this relates to point (14) and the related points all of which request more clarity of what is being measured and how it precisely relates to visual inspection.\n\nFigure 6, Fecundity is typically number of eggs produced. I suggest changing the title to progeny produced or viable progeny produced in two(?) days of egg laying. The figures seem low as your wild-type data are suggesting 2 viable progeny for each adult where >200 are expected. To be honest, from the description you provide it sounds like you have measured a decline in juveniles that results from both a developmental delay in daf-2 and decreased fecundity. Thus, you really need to be clearer about what you are really detecting and if you want to make clear claims about fecundity you should run a more appropriately controlled example.\n\nIf you are going to make claims about the Wormscan improving workflow etc (for example “greatly improves the speed at which such experiments can be performed”, you should discuss what other bottlenecks will occur. For example, in your set up how many plates can you put in a desicator at once? There is a reason, afterall, that high throughput systems go, largely, to liquid culturing. Similarly, do plates need to be poured at a specific depth for optimal use in the scanner, does this add time?\n\nI would disagree that most C. elegans assays are done on NGM plates, these days molecular endpoint and/or sub-cellular endpoint are much more common than movement (perhaps excluding toxicology and/or ageing). You might revise the sentence.\n\nIt is certainly not true that human counting is precluded in conditions that induce aggregation. One simple disaggregates the worms to count them. This statement should be removed from the discussion.\n\nSince it is clear that mobility defects impair use of the system some guidance as to how little mobility is required would be quite helpful in the discussion as would be discussion of if this can be overcome by changing the scan parameters.",
"responses": []
},
{
"id": "20572",
"date": "21 Mar 2017",
"name": "Courtney Scerbak",
"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 ability to automatically count live C. elegans on agar plates accurately, in a time-efficient manner, and with relatively low cost is of great utility to many C. elegans researchers. This manuscript and its supplementary materials provide an open access tool to allow researchers to do just that. I especially appreciate the discussion of the benefits of keeping an archival record (of images and data) in the reproducibility of scientific research. Overall, the authors describe the use and demonstrate the accuracy of this tool well enough to allow other groups to troubleshoot its application to their particular needs.\n\nWith some more clarification in the manuscript, I believe this manuscript will have an even greater practical impact on many C. elegans research groups. Below are a list of suggestions and questions:\n\nIn the Introduction, you should mention or at least refer to more techniques that are currently available for automated C. elegans handling and data collection. For example, see the WormBook chapter on Microfluidics: http://www.wormbook.org/chapters/www_microfluidics/microfluidics.html\n\nWhen referring to Matthews et al. (2012) in the Introduction, it would be worthwhile to describe in more detail their findings that the light of the scanner (a) works to detect mortality and (b) habituates the worms to the same degree as using a platinum pick. This would address any concerns about potential detrimental health effects of the scanner light right in the introduction.\n\nIs there a reason you use this particular scanner (Epson v700 or v800)?\n\nGeneral Methods questions:\nCan this program be adapted to analyze different sized plates? Is there a particular thickness the agar needs to be or will any depth work? How does your group actually get an estimated 100 animals onto each plate?\n\nWorm Strains and Maintenance should mention which exact strains/genotypes were used and acknowledge the CGC if applicable. The “daf-2” strain examined later in the manuscript has no genotype mentioned that I can find.\n\nYou mention in the results section (pdf page 3) the different measurements that are available in the end product after analysis: circumference, area and circularity. None of these measurements are listed in Figure 2 – do these data appear in a different output spreadsheet? Also, in Figures 5 and 6, you mention using perimeter to calculate the length of worms. Is perimeter another measurement that is included in the automated end product after analysis? If so, please add that item to the list in the results (given that you use it in the data presented in the Figures). If not, please provide more details on how perimeter was calculated from the automated end product dataset.\n\nPotential typo – page 4, under “Automated wormscan can be used to measure growth rate and fecundity” – second mention of the daf-2 strain is missing the “-2.” Also, related to #5 above, it would be best to name the full strain name and genotype again in this section.\n\nThe protocol used to determine fecundity in Figure 6 should be described as a section in the methods. Doing this would likely address many of my concerns/questions below in #9.\n\nI have many questions and concerns related to Figure 6:\nI am not surprised to see that daf-2 mutants have fewer viable progeny than N2, but how are there only 2 juveniles per adult in the wildtype strain? How was this data collected? Does it align with data collected manually by a trained observer with a microscope? If so, it would be useful to add that comparison to this Figure. Did you start this experiment with ~100 adults per plate and then end up with ~300 on the N2 plates? Did you remove the parents at some point or are they on these plates as well? Please clarify the number of adults and progeny per plate analyzed by the scanner as limiting the number of animals per plate was an important limitation to this tool. Did you set the size exclusion to include all animals on these plates at once? Or did you run the Automated Wormscan plug-in two times (once for the objects with >1.5mm perimeter and then again for the objects <0.5mm perimeter)?\n\nAre there limitations to using this tool in lifespan studies? Can the scanners pick up the slight movement of the head or tail of older individuals that are no longer mobile?\n\nAre there any other mutations or scenarios when you would not recommend using this tool? For example, are there specific mutations prevent the animals from responding to the scanner light?",
"responses": []
}
] | 1
|
https://f1000research.com/articles/6-192
|
https://f1000research.com/articles/8-13/v1
|
03 Jan 19
|
{
"type": "Systematic Review",
"title": "Cutaneous neonatal Langerhans cell histiocytosis: a systematic review of case reports",
"authors": [
"Victoria Venning",
"Evelyn Yhao",
"Elizabeth Huynh",
"John W. Frew",
"Victoria Venning",
"Evelyn Yhao",
"Elizabeth Huynh"
],
"abstract": "Background: Cutaneous langerhans cell histiocytosis (LCH) is a rare disorder characterized by proliferation of cells with phenotypical characteristics of Langerhans cells. Although some cases spontaneously resolve, no consistent variables have been identified that predict which cases will manifest with systemic disease later in childhood. Methods: A systematic review (Pubmed, Embase, Cochrane database and all published abstracts from 1946-2018) was undertaken to collate all reported cases of cutaneous LCH in the international literature. This study was registered with PROSPERO (CRD42016051952). Descriptive statistics and correlation analyses were undertaken. Bias was analyzed according to GRADE criteria. Results: A total of 83 articles encompassing 128 cases of cutaneous LCH were identified. Multiple lesions were weakly associated with an increased length of survival (R=0.304 (p<0.05)), Worse prognosis was associated with internal organ involvement with a statistically significant chi squared statistic (χ2=14.96, 2DF p<0.001) and an elevated odds ratio ((OR)= 12.30 95% CI=2.67-56.74). Vesicular lesions (OR=10.8 95% CI=2.83-41.26), but not ulceration (OR=0.53 95% CI 0.12-2.05) were associated with greater risk of mortality. Conclusions: Congenital and neonatal LCH most commonly presents as multiple lesions in multiple anatomical sites at birth. Significant differences, including the associations of mortality with lesion morphology and number were seen in this neonatal cohort compared to overall pediatric LCH. These findings require validation in a large prospective cohort.",
"keywords": [
"Histiocytic Disorders",
"Lumps/Bumps",
"malignant Neoplasms",
"benign Neoplasms",
"Skin signs of systemic disease"
],
"content": "Background\n\nCutaneous Langerhans cell histiocytosis (LCH) is a rare disorder manifest in the proliferation of cells with phenotypical characteristics of Langerhans cells which involves the cutaneous structures1. We have used the term ‘cutaneous’ in this review to differentiate from ‘skin-limited’ which implies the absence of systemic disease involvement1,2. The incidence of cutaneous LCH varies from two to nine cases per million children per year1,2. Rarely, the disease is present at birth or in the neonatal period. A proportion of these cases spontaneously resolve however no consistent variables have been identified which provide predictive value as to which cases will resolve or remain skin-limited, and which will manifest with multisystem LCH later in life1,2. The rate of progression of cutaneous LCH to other organs has varied widely in previous studies, from 0 to 60%1. This lack of accurate and reliable data makes it difficult to provide information to patients regarding the risk of progression of disease and limits the development of evidence-based screening measures to identify the presence of systemic disease. Currently, consensus guidelines1 state that most cases of cutaneous LCH spontaneously regress but some cases do progress to multisystem disease1. It is unclear whether cutaneous LCH is merely clinically more easily identified and hence often precedes diagnosis of internal disease. This would also suggest that widespread screening of cases of cutaneous LCH may produce lead-time bias in the survival rates of individuals with multisystem LCH with cutaneous involvement, an issue which to date has not been explored. Currently, in cases of cutaneous LCH screening is considered mandatory1,2.\n\nRegarding identified risk factors for disease progression and mortality, overwhelmingly the data is sourced from cases of systemic LCH3,4, which may or may not include cutaneous disease. Data from older children also far exceeds data from neonatal cohorts, limiting or knowledge of differences between presentations in the neonatal population and older pediatric age groups. Only one retrospective case series of 19 patients5 examined survival outcomes in infants diagnosed with cutaneous LCH within the first 4 weeks of life, with long-term follow-up beyond 10 years being limited to small case series of less than 10 patients6. In the setting of systemic LCH, inadequate response to initial therapy and risk organ involvement, (defined as bone marrow, liver, spleen and/or lung), are the currently associated with adverse clinical outcomes and mortality in LCH1,2.\n\nIsolated bone involvement portends significantly prolonged survival compared with other organ involvement3,4. As expected, patients with multiple organ involvement have been found to have the highest risk of progression and mortality7. Detection of the BRAF-V600E mutation (often seen in systemic LCH but rarely in skin-limited LCH), has also been associated with increased risk of disease recurrence4.\n\nOverall, given the reports (albeit uncommon) of progression of cutaneous LCH to multisystem disease, the identification of clinical or histological predictive variables may reduce rates of unnecessary invasive screening in neonates with skin-limited LCH.\n\nTo collate all published cases of cutaneous congenital/neonatal LCH.\n\nTo perform a descriptive analysis of cases and reviews to evaluate mortality risk and risk of progression to systemic disease.\n\nTo identify risk factors which may contribute to mortality risk and risk of progression to systematic disease.\n\n\nMethods\n\nThis systematic review was registered with PROSPERO (Registration number CRD42016041425) and was conducted in line with the PRISMA statement8\n\nInformation Sources for this review encompassed Medline (1946-March 1 2018), Embase (1980- March 1 2018) as well as “Epub ahead of print, and non-indexed citations” as shown in Figure 1. The search strategy is presented in Table 1. The databases searched were PubMed (National Library of Medicine), EMBASE, Cochrane Database of Systematic Reviews and published abstracts on Ovid Medicine (date limits for all: January 1 1980 to March 1 2018). The search terms used were: (Langerhans Cell Histiocytosis OR Hashimoto-Pritzker) AND (Congenital OR Birth OR Neonate) AND (Skin OR Cutaneous)\n\nEligibility criteria for this review included published case reports, case series and reviews with no restrictions of patient sex or ethnicity and language of publication. Eligible cases included:\n\n1) Cases of histologically diagnosed LCH at birth (congenital) or within the first 4 weeks of life involving the skin.\n\n2) Cases which report data pertaining to evidence of systemic involvement (clinical examination, skeletal survey etc.) and/or histological data (CD1a, eosinophil density etc.)\n\n3) Cases with follow up data of any period.\n\nData collection was performed independently by two independent authors (EH and EY), with any disagreements regarding inclusion of citations being referred to a third author (VV) for mediation. Information was collected using a standardized data collection form (available as extended data on OSF9) with the principal outcomes of interest being mortality, age at demise and length of follow up. Data not available from the published article was requested via email contact with the relevant corresponding authors.\n\nPotential sources of bias in collating cases were acknowledged including publications bias and reporting bias regarding the overall incidence of congenital and neonatal LCH, therefore only cases with a diagnosis of cutaneous LCH at birth (congenital) or within the first 4 weeks of life (neonatal) were included, and no attempt to quantify the number of cases of systemic LCH with a “missed” diagnosis of self-resolving cutaneous congenital LCH was undertaken. Particular effort was made to include unpublished cases and cases presented as posters and abstracts in order to reduce the impact of publication bias in our analyses.\n\nAn exploratory univariate analysis (using Pearson correlation coefficients for categorical variables and chi-squared tests for binary variables) was undertaken to correlate mortality and the progression to systemic disease with the clinical and histological variables collated.\n\n\nResults\n\nA total of 211 articles were identified in the literature review; 82 of these articles were removed upon review of titles and abstracts against eligibility criteria. Full-text review of 129 articles excluded 12 review articles, 1 duplicated case report and 33 articles (containing 42 cases) due to lack of follow up data. The remaining 83 articles5,10–91 containing 128 individual cases were used as the basis of this review.\n\nThe summarized demographic data of the included cases is presented in Table 2.\n\nThe results of univariate correlation analysis are summarized in Table 3. The presence of multiple lesions was associated with an increased length of survival (r=0.304 p<0.05), whilst the presence of systemic disease portends a worse prognosis, with a statistically significant chi squared statistic (χ2 =14.96, 2DF p<0.001). Having lesions at birth had an odds ratio (OR) of mortality of 1.38, which did not reach statistical significance (95%CI=0.417-4.56). Individuals presenting with either weight loss, hepatosplenomegaly and internal organ involvement also had a worse prognosis and decreased overall survival (OR= 8.01 95% CI=2.07-30.86)\n\nThe presence of ulcerated lesions did not change risk of survival (OR=0.53 95% CI 0.11-2.05) Having less than 10 lesions increased the risk of mortality but not to a statistically significant degree (OR=1.77 95% CI= 0.76-17.30). Vesicular lesions were significantly more likely to be associated with mortality (OR=10.8 95% CI=2.83-41.26). Of 128 cases, 112 were screened for systemic involvement (87.5%). Of the screened cases, 66 were found to have cutaneous involvement only (51.6%). The mortality rate for those with identified systemic involvement was 27.4% (n= 17/62). The calculated OR for mortality based upon the presence of systemic involvement was 12.3 (95% CI). No statistically significant associations or OR were seen between histological markers and clinical outcomes including mortality or length of survival in the data examined. Given the heterogeneity of the sample, no multivariate analysis was performed on the collated data.\n\n\nDiscussion\n\nThe results of this systematic review of case reports of cutaneous neonatal LCH differ from the pre- existing literature in several areas. This may be because existing data includes all cases of pediatric LCH, as opposed to the congenital and neonatal cases focused on in this review. This highlights the need for recognition that congenital and neonatal LCH have inherently different clinical characteristics compared to other pediatric cases of LCH. Minkov et al.3 have reported that the trunk was the most common overall site of disease. However, our data suggest that a large proportion of congenital and neonatal cases involve multiple anatomical sites (n=65). No significant gender predominance was identified (males=63; females=55). A weak association was seen between a later onset of disease and a worse prognosis (r=0.263, p<0.05). This is in line with the literature with earlier onset disease significantly associated with spontaneous resolution92,93.\n\nSystemic disease was identified in 48.4% of cases (n=62) lower than the rates for the overall pediatric group at 59%3, and those reported by Stein et al. (63.1%)5. In line with previous research and recommendations5, systemic disease was significantly correlated with mortality (r=0.453, p<0.05), with persistent cutaneous lesions associated with poorer outcomes3,75,92. The overall mortality rate for all cases in the population of this review was 14.05%.\n\nPrevious studies have suggested high rates of spontaneous clinical remission (from 60%3 to 100%5) in skin-limited LCH, and 8% in multisystem disease94. The accuracy of such figures is disputed due to the absence of systemic screening and long term follow up in these published reports. We attempted to identify cases of spontaneous remission (both clinical and biological) in the literature. Clinical remission was documented in 41/128 cases (32.1%); however, due to the high variability in length of follow-up and low rates of systemic screening post clinical remission, rates of biological remission could not be accurately established. We would suggest that long term prospective follow-up studies with systemic screening (both at diagnosis and post clinical remission) are required to accurately quantify rates of spontaneous biological remission in future studies.\n\nRegarding lesion morphology, Battistella et al.92 suggest that single, necrotic, hypopigmented macules and distal topography (lesions present at a distal site) suggest a self-regressive form of disease92. This is still an area of contention with no reliable data from cohorts larger than 20 patients5,93,95–98. We identified a weak correlation between skin lesion descriptors and overall mortality as well as length of survival in the neonatal and congenital LCH population. The presence of multiple lesions was associated with increased length of survival, although the presence of lead-time bias was likely given the non-significant differences in mortality between the two groups. Vesicular lesions were associated with increased mortality whereas no impact of survival was seen in the presence of ulcerated lesions. We anticipated that reporting bias would result in confirmation of an association between ulcerated lesions and mortality if one existed, although this has not been confirmed by our review. One explanation is that vesicular lesions commonly progress to ulceration during the stages of healing, thus emphasizing the need for consistent descriptors in case reports of LCH. Alternatively, ulcerated lesions might have an association with mortality but not in the congenital and neonatal LCH cohort.\n\n\nLimitations\n\nMost cases identified were congenital (67.9%; n=87), although some controversy exists regarding whether congenital cases exist at all93. Morren states that LCH presents prior to 3 months of age but does not occur congenitally93. Given the retrospective nature of our study, we were unable to shed further light on this debate as we were reliant upon multiple authors’ observations and recordings.\n\nGiven the variability in patient follow-up in this review, the current estimates of mortality risk are only valid until 18 months of age (the mean length of follow-up). The lack of long-term follow-up is the major reason why data is lacking regarding long term recurrence rates in neonatal LCH and thus our review is limited to conclusions regarding short- and medium-term outcomes.\n\nFuture research should expand upon this by analyzing longer-term outcomes. Haupt1 has recommended a long-term follow-up of 5 years for patients, mirroring that of childhood cancer survivors. This is applicable even to both skin-limited LCH and systemic disease. Progression of skin-limited LCH to multisystem involvement is documented in the literature1,5.\n\nWe had a limited ability to identify statistically significant variables that contribute to LCH mortality due to limited follow-up in documented cases. The GRADE approach99 to assessing the quality of evidence and strength of recommendations (available as extended data on OSF9) shows the absence of control groups, and incomplete follow up. Long-term, prospective, multicenter collaborative studies needed to confirm the findings of this review and are important steps in characterizing the progression of neonatal LCH.\n\n\nConclusions\n\nWe present a systematic review of case reports of cutaneous congenital and neonatal LCH. The descriptive characteristics in this review significantly differ from descriptions of overall pediatric LCH, highlighting the clinical differences between these entities. Congenital and neonatal LCH most commonly presents in multiple anatomical sites at or shortly after birth, with the presence or absence of systemic involvement significantly impacting mortality. Further prospective, long-term multicenter collaborative studies are required to corroborate the results of this review.\n\n\nData availability\n\nThe Data Collection Proforma and GRADE Bias Assessment are available on OSF. DOI: https://doi.org/10.17605/OSF.IO/TRX429.\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\nReporting Guidelines\n\nA completed PRISMA Checklist for this study is available on OSF. DOI: https://doi.org/10.17605/OSF.IO/TRX429.",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nHaupt R, Minkov M, Astigarraga I, et al.: Langerhans cell histiocytosis (LCH): guidelines for diagnosis, clinical work-up, and treatment for patients till the age of 18 years. Pediatr Blood Cancer. 2013; 60(2): 175–184. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSimko SJ, Garmezy B, Abhyankar H, et al.: Differentiating skin-limited and multisystem Langerhans cell histiocytosis. J Pediatr. 2014; 165(5): 990–996. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMinkov M, Grois N, Heitger A, et al.: Response to initial treatment of multisystem Langerhans cell histiocytosis: an important prognostic indicator. Med Pediatr Oncol. 2002; 39(6): 581–585. PubMed Abstract | Publisher Full Text\n\nAllen CE, Ladisch S, McClain KL: How I treat Langerhans cell histiocytosis. Blood. 2015; 126(1): 26–35. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStein SL, Paller AS, Haut PR, et al.: Langerhans cell histiocytosis presenting in the neonatal period: a retrospective case series. Arch Pediatr Adolesc Med. 2001; 155(7): 778–783. PubMed Abstract | Publisher Full Text\n\nZunino-Goutorbe C, Eschard C, Durlach A, et al.: Congenital solitary histiocytoma: a variant of Hashimoto-Pritzker histiocytosis. A retrospective study of 8 cases. Dermatology. 2008; 216(2): 118–124. PubMed Abstract | Publisher Full Text\n\nBerres ML, Lim KP, Peters T, et al.: BRAF-V600E expression in precursor versus differentiated dendritic cells defines clinically distinct LCH risk groups. J Exp Med. 2014; 211(4): 669–683. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiberati A, Altman DG, Tetzlaff J, et al.: The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med. 2009; 6(7): e1000100. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFrew J: Cutaneous Neonatal Langerhan Cell Histiocytosis: A Systematic Review of Case Reports. 2018. http://www.doi.org/10.17605/OSF.IO/TRX42\n\nChen AJ, Jarrett P, Macfarlane S: Congenital self-healing reticulohistiocytosis: the need for investigation. Australas J Dermatol. 2016; 57(1): 76–77. PubMed Abstract | Publisher Full Text\n\nErsoy-Evans S, Gursoy T, Yigit S, et al.: Solitary congenital self-healing reticulohistiocytosis in monozygotic twins. Pediatr Dermatol. 2006; 23(3): 273–275. PubMed Abstract | Publisher Full Text\n\nFlann S, Samman P, Garioch J, et al.: Two cases of congenital self-healing histiocytosis (Hashimoto–Pritzker syndrome). J Am Acad Dermatol. 2012; 66(4 Supplement 1): AB63. Publisher Full Text\n\nGee SN, Huang JT, Schmidt BA, et al.: Rapidly fatal multiorgan Langerhans cell histiocytosis in a neonate. Pediatr Dermatol. 2013; 30(5): e85–6. PubMed Abstract | Publisher Full Text\n\nGökmen Z, Özkiraz S, Kıyıcı H, et al.: Kendiliğinden düzelen doğumsal retikülohistiyositoz Hashimoto Pritzker Hastalığı: olgu sunumu. Türk Pediatri Arşivi. 2011; 46: 84–6. Publisher Full Text\n\nHashimoto K, Bale GF, Hawkins HK, et al.: Congenital self-healing reticulohistiocytosis (Hashimoto-Pritzker type). Int J Dermatol. 1986; 25(8): 516–523. PubMed Abstract | Publisher Full Text\n\nHashimoto K, Pritzker MS: Electron microscopic study of reticulohistiocytoma. An unusual case of congenital, self-healing reticulohistiocytosis. Arch Dermatol. 1973; 107(2): 263–270. PubMed Abstract | Publisher Full Text\n\nHashimoto K, Takahashi S, Lee RG, et al.: Congenital self-healing reticulohistiocytosis. Report of the seventh case with histochemical and ultrastructural studies. J Am Acad Dermatol. 1984; 11(3): 447–454. PubMed Abstract | Publisher Full Text\n\nHerman LE, Rothman KF, Harawi S, et al.: Congenital self-healing reticulohistiocytosis. A new entity in the differential diagnosis of neonatal papulovesicular eruptions. Arch Dermatol. 1990; 126(2): 210–212. PubMed Abstract | Publisher Full Text\n\nHuang CY, Chao SC, Ho SF, et al.: Congenital self-healing reticulohistiocytosis mimicking diffuse neonatal hemangiomatosis. Dermatology. 2004; 208(2): 138–141. PubMed Abstract | Publisher Full Text\n\nInuzuka M, Tomita K, Tokura Y, et al.: Congenital self-healing reticulohistiocytosis presenting with hemorrhagic bullae. J Am Acad Dermatol. 2003; 48(5 Suppl): S75–7. PubMed Abstract | Publisher Full Text\n\nJahnke I, Stieler KM, Bartels NG: Cutaneous manifestation in multi-system Langerhans cell histiocytosis (lch): a clue for early diagnosis: P90. Journal der. 2014.\n\nJensen ML, Bygum A, Clemmensen O, et al.: Congenital self-healing reticulohistiocytosis - an important diagnostic challenge. Acta Paediatr. 2011; 100(5): 784–786. PubMed Abstract | Publisher Full Text\n\nJordaan HF, Drusinsky SF: Congenital self-healing reticulohistiocytosis: report of a case. Pediatr Dermatol. 1986; 3(6): 473–475. PubMed Abstract | Publisher Full Text\n\nKapur P, Erickson C, Rakheja D, et al.: Congenital self-healing reticulohistiocytosis (Hashimoto-Pritzker disease): ten-year experience at Dallas Children's Medical Center. J Am Acad Dermatol. 2007; 56(2): 290–294. PubMed Abstract | Publisher Full Text\n\nKassardjian M, Patel M, Shitabata P, et al.: Congenital self-healing reticulohistiocytosis: an underreported entity. Cutis. 2016; 97(4): 296–300. PubMed Abstract\n\nKim JE, Kim BJ, Kang H: Solitary congenital erosion in a newborn: report of a solitary congenital self-healing reticulohistiocytosis. Ann Dermatol. 2014; 26(2): 250–253. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim KJ, Jee MS, Choi JH, et al.: Congenital self-healing reticulohistiocytosis presenting as vesicular eruption. J Dermatol. 2002; 29(1): 48–49. PubMed Abstract | Publisher Full Text\n\nKüpeli S, Varan A, Ayvaz M, et al.: Congenital solid thigh mass: a report of nonself healing form of langerhans cell histiocytosis. J Pediatr Hematol Oncol. 2010; 32(4): 312–314. PubMed Abstract | Publisher Full Text\n\nLarralde M, Rositto A, Giardelli M, et al.: Congenital self-healing Langerhans cell histiocytosis: the need for a long term follow up. Int J Dermatol. 2003; 42(3): 245–246. PubMed Abstract | Publisher Full Text\n\nLarsen L, Merin MR, Konia T, et al.: Congenital self-healing reticulohistiocytosis: concern for a poor prognosis. Dermatol Online J. 2012; 18(10): 2. PubMed Abstract\n\nLee YH, Talekar MK, Chung CG, et al.: Congenital self-healing reticulohistiocytosis. J Clin Aesthet Dermatol. 2014; 7(2): 49–53. PubMed Abstract | Free Full Text\n\nLongaker MA, Frieden IJ, LeBoit PE, et al.: Congenital “self-healing” Langerhans cell histiocytosis: the need for long-term follow-up. J Am Acad Dermatol. 1994; 31(5 Pt 2): 910–916. PubMed Abstract | Publisher Full Text\n\nMandel VD, Ferrari C, Cesinaro AM, et al.: Congenital “self-healing” Langerhans cell histiocytosis (Hashimoto-Pritzker disease): a report of two cases with the same cutaneous manifestations but different clinical course. J Dermatol. 2014; 41(12): 1098–1101. PubMed Abstract | Publisher Full Text\n\nMaurice PD, Mahon C: Blueberry muffin babies: three recent cases and a brief review. Australas J Dermatol. 2012; 53(4): A6. Reference Source\n\nMorgan KW, Callen JP: Self-healing congenital Langerhans cell histiocytosis presenting as neonatal papulovesicular eruption. J Cutan Med Surg. 2001; 5(6): 486–489. PubMed Abstract\n\nOfuji S, Tachibana S, Kanato M, et al.: Congenital self-healing reticulohistiocytosis (Hashimoto-Pritzker): a case report with a solitary lesion. J Dermatol. 1987; 14(2): 182–184. PubMed Abstract | Publisher Full Text\n\nOranje AP, Vuzevski VD, de Groot R: Congenital self-healing non-Langerhans cell histiocytosis. Eur J Pediatr. 1988; 148(1): 29–31. PubMed Abstract | Publisher Full Text\n\nOrle J, Mósca AM, Sousa MA, et al.: Congenital self healing reticulohistiocytosis in a newborn (Hashimoto Pritzker). An Bras Dermatol. 2011; 86(4): 785–788. PubMed Abstract | Publisher Full Text\n\nPaquet P, Hermanns-Lê T, Piérard GE: [Hashimoto-Pritzker self-healing reticulohistiocytosis and congenital histiocytosis]. Arch Pediatr. 1994; 1(6): 578–581. PubMed Abstract\n\nParentin F, Ventura G, Pastore S, et al.: A singular case of congenital self-healing histiocytosis with skin, liver and atypical eye involvement. Ocul Immunol Inflamm. 2011; 19(5): 337–339. PubMed Abstract | Publisher Full Text\n\nPastore S: Congenital self-healing histiocytosis with eye and liver involvement. Clin Exp Rheumatol. 2011; 29(2): 424. Reference Source\n\nRicart J, Jimenez A, Marquina A, et al.: Congenital self-healing reticulohistiocytosis: report of a case and review of the literature. Acta Paediatr. 2004; 93(3): 426–429. PubMed Abstract | Publisher Full Text\n\nRufli T, Fricker HS: [Congenital, self-healing reticulohistiocytosis]. Z Hautkr. 1979; 54(12): 554–558. PubMed Abstract\n\nSchaumburg-Lever G, Rechowicz E, Fehrenbacher B, et al.: Congenital self-healing reticulohistiocytosis--a benign Langerhans cell disease. J Cutan Pathol. 1994; 21(1): 59–66. PubMed Abstract | Publisher Full Text\n\nSertznig P, Megahed M: [Congenital self-healing Langerhans cell histiocytosis]. Hautarzt. 2011; 62(11): 804–807. PubMed Abstract | Publisher Full Text\n\nTan Q, Gan LQ, Wang H: Congenital self-healing Langerhans cell histiocytosis in a male neonate. Indian J Dermatol Venereol Leprol. 2015; 81(1): 75–77. PubMed Abstract | Publisher Full Text\n\nTantiwongkosi B, Goske MJ, Steele M: Congenital solid neck mass: a unique presentation of Langerhans cell histiocytosis. Pediatr Radiol. 2008; 38(5): 575–578. PubMed Abstract | Publisher Full Text\n\nTay YK, Friednash MM, Weston WL, et al.: Solitary congenital nodule in an infant. Solitary congenital self-healing reticulohistiocytosis (CSHR). Arch Dermatol. 1998; 134(5): 627, 630. PubMed Abstract\n\nTerry J, Pluchinotta FR, Sanders SP, et al.: Congenital langerhans cell histiocytosis with placental involvement. Pediatr Dev Pathol. 2013; 16(3): 224–228. PubMed Abstract | Publisher Full Text\n\nThong HY, Dai YS, Chiu HC: An unusual presentation of congenital self-healing reticulohistiocytosis. Br J Dermatol. 2003; 149(1): 191–192. PubMed Abstract | Publisher Full Text\n\nTimpatanapong P, Rochanawutanon M, Siripoonya P, et al.: Congenital self-healing reticulohistiocytosis: report of a patient with a strikingly large tumor mass. Pediatr Dermatol. 1989; 6(1): 28–32. PubMed Abstract | Publisher Full Text\n\nUaratanawong R, Kootiratrakarn T, Sudtikoonaseth P, et al.: Congenital self-healing reticulohistiocytosis presented with multiple hypopigmented flat-topped papules: a case report and review of literatures. J Med Assoc Thai. 2014; 97(9): 993–997. PubMed Abstract\n\nTuregano M, Chirinos R, Epps R: A case of congenital multisystem Langerhans cell histiocytosis. Am J Dermatopathol. 2011; 33: 424.\n\nTucker C: Langerhans cell histiocytosis presenting as a blueberry muffin baby. J Am Acad Dermatol. 2013; 68(4): AB175. Publisher Full Text\n\nTabassum F, Debelenko L, Savasan S: 19 month old male with clinical diagnosis of congenital lamellar Ichthyosis and multiorgan involvement by langerhans cell histiocytosis. 2011.\n\nMitsiadi V, Papadakis V, Stefanaki K: Congenital Langerhans Cell Histiocytosis with systemic involvement. Eur J Pediatr Dermatol. 2010; 20: 69.\n\nJenning C, Benjamin L, Kim J: An unusual mixed population of langerhans cells and intermediate cells in cutaneous histiocytosis of a newborn. 2012.\n\nUber M, Robl R, Carvalho VO, et al.: Congenital ulcer: what do you think? Arch Dis Child Fetal Neonatal Ed. 2014; 99(5): F437. PubMed Abstract | Publisher Full Text\n\nUdayasankar UK, Alazraki AL, Simoneaux SF: Role of PET/CT in congenital histiocytosis. Pediatr Radiol. 2010; 40 Suppl 1: S57–61. PubMed Abstract | Publisher Full Text\n\nVetter-Laracy S, Salinas JA, Martin-Santiago A, et al.: Digestive tract symptoms in congenital langerhans cell histiocytosis: a fatal condition in an illness usually considered benign. J Pediatr Hematol Oncol. 2014; 36(6): 426–429. PubMed Abstract | Publisher Full Text\n\nWalia M, Paul P, Mishra S, et al.: Congenital Langerhans cell histiocytosis: the self-healing variety. J Pediatr Hematol Oncol. 2004; 26(6): 398–402. PubMed Abstract\n\nWheller L, Carman N, Butler G: Unilesional self-limited Langerhans cell histiocytosis: a case report and review of the literature. J Cutan Pathol. 2013; 40(6): 595–599. PubMed Abstract | Publisher Full Text\n\nWhitehead B, Michaels M, Sahni RI, et al.: Congenital self-healing Langerhans cell histiocytosis with persistent cellular immunological abnormalities. Br J Dermatol. 1990; 122(4): 563–568. PubMed Abstract | Publisher Full Text\n\nWilliams V, Yang D, Julapalli M, et al.: Congenital self-healing reticulo- histiocytosis: Review of a “Wait and See” Hypothesis. Pediatr Dermatol. 2010; 27: 583. Reference Source\n\nYang CC, Chen YA, Tsai YL, et al.: Neoplastic skin lesions of the scalp in children: a retrospective study of 265 cases in Taiwan. Eur J Dermatol. 2014; 24(1): 70–75. PubMed Abstract | Publisher Full Text\n\nYurkovich M, Wong A, Lam JM: Solitary congenital self-healing Langerhans cell histiocytosis: a benign variant of Langerhans cell histiocytosis? Dermatol Online J. 2013; 19(1): 3. PubMed Abstract\n\nZaenglein AL, Steele MA, Kamino H, et al.: Congenital self-healing reticulohistiocytosis with eye involvement. Pediatr Dermatol. 2001; 18(2): 135–137. PubMed Abstract | Publisher Full Text\n\nParimi LR, You J, Hong L, et al.: Congenital self-healing reticulohistiocytosis with spontaneous regression. An Bras Dermatol. 2017; 92(4): 553–555. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSuresh SG, Jagadeesan S, Srinivasan A, et al.: Rare diagnosis in a neonate with isolated skin lesions. Int J Comtemp Pediatr. 2017; 4(4): 1534–1536. Publisher Full Text\n\nChun SI, Song MS: Congenital self-healing reticulohistiocytosis--report of a case of the solitary type and review of the literature. Yonsei Med J. 1992; 33(2): 194–198. PubMed Abstract | Publisher Full Text\n\nSingh A, Mandal A, Singh L, et al.: Delayed Treatment Response in a Neonate with Multisystem Langerhans Cell Histiocytosis Case report and review of literature. Sultan Qaboos Univ Med J. 2017; 17(2): e225–e228. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWalsh M, Sharpe G, Parslew R: Two cases of congenital self-healing histiocytosis and a review of the literature. Br J Dermatol. 2012; 168: e1–e10. Reference Source\n\nde Cerqueira A, Lima O, de Azevedo JOC: Congenital self-healing reticulohistiocytosis in a newborn: A case report. J Am Acad Dermatol. 2009; 60(3 Supplement 1): AB144. Publisher Full Text\n\nOh Y, Morimoto A, Itoh T, et al.: A case of langerhans cell histiocytosis with a recurrent CNS mass lesion successfully treated with 2-chlorodeoxyadenosine. Pediatr Blood Cancer. 2011; 56(4): 697. Reference Source\n\nWang H, Chen C, Chen Z: Another case of new-born baby with blueberry-muffin rashes: congenital viral infection or immune related hematologic disorders. Biomed Res. 2017; 28(7): 3204–3208. Reference Source\n\nHashimoto K, Schachner LA, Huneiti A, et al.: Pagetoid self-healing Langerhans cell histiocytosis in an infant. Pediatr Dermatol. 1999; 16(2): 121–127. PubMed Abstract | Publisher Full Text\n\nXiao-Bing P, Xiao-Xia W: Congenital self-healing Langerhans cell histiocytosis. Am J Clin Dermatol. 2011; 40: 84–86.\n\nYu J, Rubin AI, Castelo-Soccio L, et al.: Congenital Self-Healing Langerhans Cell Histiocytosis. J Pediatr. 2017; 184: 232–232.e1. Publisher Full Text\n\nAlexis JB, Poppiti RJ, Turbat-Herrera E, et al.: Congenital self-healing reticulohistiocytosis. Report of a case with 7-year follow-up and a review of the literature. Am J Dermatopathol. 1991; 13(2): 189–194. PubMed Abstract | Publisher Full Text\n\nBaillie L, Carman N, Butler G: Solitary lesion congenital self-healing reticulohistiocytosis: A case report and review of the literature. Australas J Dermatol. 2011. Reference Source\n\nBains A, Parham DM: Langerhans cell histiocytosis preceding the development of juvenile xanthogranuloma: a case and review of recent developments. Pediatr Dev Pathol. 2011; 14(6): 480–484. PubMed Abstract | Publisher Full Text\n\nBerger TG, Lane AT, Headington JT, et al.: A solitary variant of congenital self-healing reticulohistiocytosis: solitary Hashimoto-Pritzker disease. Pediatr Dermatol. 1986; 3(3): 230–236. PubMed Abstract | Publisher Full Text\n\nBernstein EF, Resnik KS, Loose JH, et al.: Solitary congenital self-healing reticulohistiocytosis. Br J Dermatol. 1993; 129(4): 449–454. PubMed Abstract | Publisher Full Text\n\nBhaskaran S, Egler R, Sandhaus L: unique Case Of Cutaneous Lch Progressing To Disseminated Jxg: poster# 3043. Pediatric Blood & …. 2014.\n\nBrazzola P, Schiller P, Kühne T: Congenital self-healing langerhans cell histiocytosis with atrophic recovery of the skin: clinical correlation of an immunologic phenomenon. J Pediatr Hematol Oncol. 2003; 25(3): 270–273. PubMed Abstract | Publisher Full Text\n\nChunharas A, Pabunruang W, Hongeng S: Congenital self-healing Langerhans cell histiocytosis with pulmonary involvement: spontaneous regression. J Med Assoc Thai. 2002; 85 Suppl 4: S1309–13. PubMed Abstract\n\nDeurloo E, van den Bos C, Ahout I, et al.: calcified Splenic Lesions In A Child With Congenital Multi-system Langerhans Cell Histiocytosis. Pediatr Blood Cancer. 2009. Reference Source\n\nDorjsuren G, Kim HJ, Jung JY, et al.: Solitary Type of Congenital Self-healing Reticulohistiocytosis. Ann Dermatol. 2011; 23 Suppl 1: S4–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDvir R, Sayar D, Levin D, et al.: Congenital Langerhans Cell Histiocytosis: From A Diffuse Self Limiting Rash To Extensive Intracranial Disease—a Case Report. Pediatr Blood Cancer. 2011. Reference Source\n\nEdwards AN, Altman D, Altman J, et al.: Molluscumlike papules in a 4-month-old boy--quiz case. Langerhans cell histiocytosis (LCH)–congenital self-healing reticulohistiocytosis. Arch Dermatol. 2011; 147(3): 345–350. PubMed Abstract | Publisher Full Text\n\nZunino-Goutorbe C, Eschard C, Durlach A, et al.: Congenital solitary histiocytoma: a variant of Hashimoto-Pritzker histiocytosis. A retrospective study of 8 cases. Dermatology. 2008; 216(2): 118–124. PubMed Abstract | Publisher Full Text\n\nBattistella M, Fraitag S, Teillac DH, et al.: Neonatal and early infantile cutaneous langerhans cell histiocytosis: comparison of self-regressive and non-self-regressive forms. Arch Dermatol. 2010; 146(2): 149–156. PubMed Abstract | Publisher Full Text\n\nMorren MA, Vanden Broecke K, Vangeebergen L, et al.: Diverse Cutaneous Presentations of Langerhans Cell Histiocytosis in Children: A Retrospective Cohort Study. Pediatr Blood Cancer. 2016; 63(3): 486–92. PubMed Abstract | Publisher Full Text\n\nJubran RF, Marachelian A, Dorey F, et al.: Predictors of outcome in children with Langerhans cell histiocytosis. Pediatr Blood Cancer. 2005; 45(1): 37–42. PubMed Abstract | Publisher Full Text\n\nEsterly NB, Maurer HS, Gonzalez-Crussi F: Histiocytosis X: a seven-year experience at a children’s hospital. J Am Acad Dermatol. 1985; 13(3): 481–496. PubMed Abstract | Publisher Full Text\n\nEnjolras O, Leibowitch M, Bonacini F, et al.: [Congenital cutaneous Langerhans histiocytosis. Apropos of 7 cases]. Ann Dermatol Venereol. 1992; 119(2): 111–117. PubMed Abstract\n\nHoward JE, Dwivedi RC, Masterson L, et al.: Langerhans cell sarcoma: a systematic review. Cancer Treat Rev. 2015; 41(4): 320–331. PubMed Abstract | Publisher Full Text\n\nMandel VD, Ferrari C, Cesinaro AM, et al.: Congenital “self-healing” Langerhans cell histiocytosis (Hashimoto-Pritzker disease): a report of two cases with the same cutaneous manifestations but different clinical course. J Dermatol. 2014; 41(12): 1098–1101. PubMed Abstract | Publisher Full Text\n\nGuyatt GH, Oxman AD, Vist G, et al.: GRADE guidelines: 4. Rating the quality of evidence--study limitations (risk of bias). J Clin Epidemiol. 2011; 64(4): 407–415. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "47280",
"date": "02 May 2019",
"name": "Jolie Krooks",
"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 clearly state their rationale for the review: to identify particular cutaneous presentations that are more frequently associated with progression to multisystem disease. In fact, the authors undermine their study’s potential implications in only mentioning reduced invasive screening. Perhaps equally, if not more important, it can help guide management, decreasing the use of systemic therapy in cases that will likely regress while prompting the use of systemic therapy in patients with high-risk lesions. Nevertheless, while the objective stated by the authors was to identify risk factors which may contribute to mortality risk and risk of progression to systemic disease, this aspect was not adequately addressed in the paper (we do not think that collating published cases is the appropriate methodology for such a scientific question) and is not part of the conclusions.\nMore details of the methods and analysis are needed to allow for replication by others. A standardized data collection form (not received by reviewers) should accompany the article for readers to understand some of the statements that are not based on findings described in the section “Results”. Furthermore, comparing single versus multiple lesions implies clear definitions for a skin lesion (at least for size and type), and the authors do not provide such definitions. For instance, multiple lesions may describe widespread rashes involving most of the body, two or three distinct nodules, or a rash covering a significant portion of the diaper area. Are these all truly equivalent?\nCertain components of the authors’ interpretation need to be addressed. For instance, it is unclear why the authors preferred to study the influence of multiple lesions on length of survival, rather than on progression to systemic disease or mortality, especially considering the limitation of short follow-up period. Also, the authors conclude that the descriptive characteristics in this review significantly differ from descriptions of overall pediatric LCH. However, they do not provide a statistical comparison or clarification of the age range that they consider pediatric. It is also questionable whether meaningful analysis of the impact of lesions type can be performed considering that 31/128 (24%) of lesions were listed as “others” and another 9 were missing.\n\nIn the last paragraph of the results, the authors state that no statistically significant associations or OR were seen between histological markers and clinical outcomes. How could such a conclusion be drawn without central pathology review? In the view of the reviewers, such a correlation cannot be made based on a description of the pathology findings, regardless of how well and detailed they are described, but needs a review of the biopsy specimens by experienced pathologists in a structured way. Additionally, the authors report that they were unable to perform a multivariate analysis due to data heterogeneity. If by heterogeneous the authors were implying that the data was inconclusive and/or insufficient, then this poses questions to the findings of the univariate analysis either.\nThe following components should also be addressed: In the 1st paragraph of the background, the authors write: “Cutaneous Langerhans cell histiocytosis (LCH) is a rare disorder manifest in the proliferation of cells with phenotypical characteristics of Langerhans cells which involves the cutaneous structures1.” However, it should be noted that Langerhans cells are dendritic cells of the skin and mucosa, and that despite their phenotypic resemblance to Langerhans cells (and shared immunohistochemical markers), the pathologic cells of LCH derive from immature myeloid precursor cells1. Additionally, reference one used by the authors is a set of guidelines. Accordingly, the original studies should be cited when statistics are noted (i.e. the incidence of cutaneous LCH and rate of progression). The authors write: “Currently, consensus guidelines1 state that most cases of cutaneous LCH spontaneously regress but some cases do progress to multisystem disease1”. Do they mean \"skin-limited\" rather than \"cutaneous\" according to the distinction made by the authors that “skin-limited” implies the lack of systemic involvement? Though spontaneous remission is the norm for \"skin-limited\" LCH, skin-limited disease is rare (2% of cases)2. The authors write: “It is unclear whether cutaneous LCH is merely clinically more easily identified and hence often precedes diagnosis of internal disease\". Actually, cutaneous disease is often misdiagnosed due to its resemblance to other more common conditions (i.e diaper dermatitis and cradle-cap)3. In addition to noting in the background section that screening for multisystem disease at the time of initial presentation is mandatory, the authors should also mention here the need for long-term follow-up due to the potential for disease reactivation following resolution or future progression to multisystem disease. The authors note this in the conclusion, but it should also be stated here. The authors only note one source in the conclusion giving this recommendation, which undermines its importance4,5,6.\nIn the 3rd paragraph of the background, the authors write “Detection of the BRAF-V600E mutation (often seen in systemic LCH but rarely in skin-limited LCH), has also been associated with increased risk of disease recurrence4”. Of note, BRAF-V600E mutations are not only associated with recurrence, but also with treatment-refractory disease and permanent sequelae. These associations are observed in both isolated and disseminated LCH (and in disseminated disease are also associated with risk organ involvement)7,8,9. Furthermore, Héritier’s study in a cohort of 315 patients with determined BRAF status conflict with the statement that BRAF-V600E is rare in skin-limited disease. They report the presence of BRAF-V600E mutations in 87.5% of patients with multifocal single system cutaneous disease and in 80.2% of patients with multifocal cutaneous multisystem disease. What might also be of interest to the authors is that the mutation was absent in the 6 infants with solitary cutaneous lesions and single system disease10.\nIn the 4th paragraph of the background, the authors write: “Overall, given the reports (albeit uncommon) of progression of cutaneous LCH to multisystem disease, the identification of clinical or histological predictive variables may reduce rates of unnecessary invasive screening in neonates with skin-limited LCH\". However, as we note above, multisystem disease in patients presenting with cutaneous involvement is the norm.\nIn the 1st paragraph of the discussion, the authors write: “This is in line with the literature with earlier onset disease significantly associated with spontaneous resolution92,93”. In contrast, because high-risk multisystem disease has been negatively correlated with age, younger age is associated with a worse prognosis11. Subsequent to Minkov’s findings, Gadner et al. reported no difference in treatment response between different age groups when correcting for the involvement of risk organ systems. Thus, the difference in prognosis between age groups is likely due to the higher prevalence of multisystem disease in younger patients12. Similarly, in the 2nd paragraph, the authors write: “Systemic disease was identified in 48.4% of cases (n=62) lower than the rates for the overall pediatric group at 59%3, and those reported by Stein et al. (63.1%)5”. However, in a retrospective analysis of 61 neonates with LCH, Minkov et al. note a higher prevalence of multisystem disease in neonates (ironically, 59%)11.\nThe authors write in the 4th paragraph of the discussion: “The presence of multiple lesions was associated with increased length of survival, although the presence of lead-time bias was likely given the non-significant differences in mortality between the two groups”. Despite the evolution in the classification system with increased recognition that the previously distinct categories have some overlapping features, it might still be worth mentioning. Specifically, Hashimoto-Pritzker disease (a.k.a congenital self-healing reticulohistiocytosis) was described as a widespread eruption of red-brown nodules presenting within the first few weeks from birth that resolve spontaneously with isolated cutaneous involvement13.\nThe authors acknowledge a number of limitations, but this does not compensate for them. As they note, limited follow-up prevented them from identifying statistically significant variables that contribute to LCH mortality. Though their stated objective of identifying particular cutaneous presentations that are more frequently associated with progression to multisystem disease and overall mortality is a good one, they were not able to identify any new prognostic factors that would contribute to the literature.\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? No\n\nIs the statistical analysis and its interpretation appropriate? No\n\nAre the conclusions drawn adequately supported by the results presented in the review? No",
"responses": []
},
{
"id": "49618",
"date": "20 Jun 2019",
"name": "Joseph M. Lam",
"expertise": [
"Reviewer Expertise Pediatric dermatology"
],
"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\nDr. Venning and colleagues perform a review on neonatal and congenital cases of Langerhans cell histiocytosis (LCH) to examine predictors for progression to systemic disease. While this endeavour will help to risk-stratify patients, there are several issues that need to be addressed:\nABSTRACT\nIn the abstract, the text should make reference to congenital and neonatal onset LCH, as this is the focus of their research.\n\nThe title should be more specific as to the purpose of the review (i.e. that this is to look for variables the predict systemic involvement).\n\nBACKGROUND\nIn the background, the authors state that \"Currently, consensus guidelines state that most cases of cutaneous LCH spontaneously regress but some cases do progress to multisystem disease\". However, the reference states that “progression to MS-LCH is common.\"\n\nThroughout the manuscript, the discussion intermingles description of cutaneous LCH, skin-limited LCH and neonatal/congenital LCH. Since the focus appears to be on the progression of neonatal/congenital cutaneous LCH, the discussion and data collection should be focused on this. As well, it would have been helpful to have more background information as to the previously documented frequency of skin-limited LCH, as the paper doesn't clearly provide this (i.e. How common is skin-limited LCH that doesn't progress (or regresses)).\n\nThe authors mention the lung may be a high-risk organ. However, pulmonary involvement is no longer considered a risk organ, as Ronceray et al.1 found it to have no significant impact on survival in a multivariate analysis.\n\nThe authors mention that \"Overall, given the reports (albeit uncommon) of progression of cutaneous LCH to multisystem disease...\". However, the study from Lau et al.2 in 2006 demonstrated that 4/12 (33%) of patients with isolated skin LCH progressed to multi-system LCH.\n\nMETHODS\nIt is unclear if the the study eligibility criteria requires all 3 criteria to qualify. Please elaborate. As well, please document reasons for exclusion. I note that the study by Lau et al.2 has data on patients but was not included.\n\nIt is unclear what the statement \"no attempt to quantify the number of cases of systemic LCH with a “missed” diagnosis of self-resolving cutaneous congenital LCH was undertaken\" means. Does this mean these cases were included or excluded?\n\nRESULTS\nPlease define \"multiple\". Is this more than 1 lesion or 10 and why was this number chosen? As before, the distinction between skin limited LCH and cutaneous LCH should be clear.\n\nAmong the paper's objectives is: \"to identify risk factors which may contribute to mortality risk and risk of progression to systemic disease.\", however, when I finished reading the paper, I still did not have a clear idea of what risk factors contribute to the mortality and progression risk in the congenital/neonatal subset of LCH. In the conclusion they state that the presentation of congenital/neonatal LCH more often involves multiple sites compared to paediatric LCH, and that systemic involvement significantly impacts mortality. To me, the latter point holds in both paediatric and congenital/neontal LCH and isn't new information, nor helpful in distinguishing the two.\n\nLIMITATION\nThe statement \"Morren states that LCH presents prior to 3 months of age but does not occur congenitally\" is odd, as Morren's report lists 2 congenial cases3.\n\nOverall, this is an interesting undertaking. However, methodologically, there a number of issue that should be examined.\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? No\n\nIs the statistical analysis and its interpretation appropriate? No\n\nAre the conclusions drawn adequately supported by the results presented in the review? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-13
|
https://f1000research.com/articles/7-1780/v1
|
12 Nov 18
|
{
"type": "Research Article",
"title": "Studies on Biological Test of Formulated Diets Supplementation of Vitamin E for the Broodstock of Females Blue Swimming Crab, Portunus pelagicus (Linnaeus, 1758)",
"authors": [
"Efrizal Efrizal",
"Indra Junaidi Zakaria",
"Rusnam Rusnam",
"Suryati Suryati",
"Nofa Yolanda",
"Indra Junaidi Zakaria",
"Rusnam Rusnam",
"Suryati Suryati",
"Nofa Yolanda"
],
"abstract": "Background: Currently, great progress in the artificial propagation of commercially important\n\nportunid crabs of the genus Portunus has been achieved, and various methods have been adopted in mass-scale hatchery activities. This study analyzed the biological testing of formulated diets with different dose supplementations of vitamin E for the broodstock of female blue swimming crabs, P. pelagicus (Linnaeus, 1758) Methods: Female crab samples were collected from the coastal region of Padang, West Sumatra. The method used in this study was completely randomized design, with four treatment regimens (n=5 crabs each) of dietary vitamin E (0, 300, 600, and 900 IU/kg formulated diets). Results: The results show that the supplementation of vitamin E in the formulated diet had a significant effect (P <0.05) on the absolute weight growth, carapace length and carapace width. Conclusions: Supplementation of vitamin E on in formulated diet causes broodstock blue swimming crab molting, with a percentage value of 40–80% on day 20 and 20% on day 30, with a 100% survival rate.",
"keywords": [
"Blue swimming crab",
"female broodstock diet",
"vitamin E"
],
"content": "Introduction\n\nThe market share of the crab, Portunus pelagicus, at the international level from year to year tends to increase, with the main markets in Singapore, Japan, the Netherlands, France, USA, Taiwan, Hong Kong, Korea, etc1–4. The high market value of, and high demand for, crabs in the global market provide a strong stimulus for their intensive culture. Although the mass production of blue swimming crab seeds was first reported by the Research and Development Center for Oceanology (RDCO)-LIPI in Indonesia5, its culture industry has not developed as expected. A shortage of seeds, lack of suitable feed, lack of an appropriate culture system and poor management are recognized as the major bottlenecks in the development of the crab culture industry in this country. P. pelagicus seeds were previously caught in the wild, causing an insufficient seed supply for commercial firms and recreational fisheries. Currently, great progress in the artificial propagation of commercially important portunid crabs in the genus Portunus has been achieved and various methods have been adopted for mass-scale hatchery activities. However, farming this species has always been limited by the seasonal availability of spawning crabs from the wild and their inability to mature and spawn in culture ponds. The second major problem associated with the blue swimming crab culture is the high larval mortality during the zoeal and megalopa stages. Thus, it has become necessary to biologically test the supplementation of vitamin E for the formulated diets of broodstocks of females blue swimming crabs, Portunus pelagicus (Linnaeus, 1758).\n\n\nMethods\n\nThe study was conducted at the Fish Seed Center Beaches (BBIP) Gulf Buo, Fish Seed Center (BBI) Bungus, of the city of Padang, and the Laboratory of Animal Physiology, Department of Biology, Faculty of Math and Science, Andalas University, Padang, West Sumatra. A total of 70 female crabs were collected on January 2018; 20 were females at stage II of ovarian maturity, which were selected for study. The female crab samples were collected from the coastal region of Padang, West Sumatra, and randomly placed in four concrete tanks (200 × 100 × 100 cm). A total of five units from each tank were placed in a plastic box (45.5 × 32.5 × 16.5 cm) at a maximum density of one crab per box. The tanks were provided with a sand substrate layer approximately 15 cm thick with adequate aeration2,6,7. The crabs were maintained a monitored water depth of 25–30 cm with salinity of 29–32 parts per thousand (ppt), pH of 7.26 to 8.00, temperature of 26–28°C, and DO of 6.15–7.45 ppm. Each crab was provided with a shelter made of PVC pipe, 13 cm in diameter and 40 cm in length, to serve as a refuge during molting.\n\nDietary treatments, with supplementation of vitamin E, were fed daily at 3% of the biomass (1700–1800 hours), and uneaten food was removed every morning. The method used in this study was completely randomized design with four treatments and (n=5 crabs per treatment) replications of dietary vitamin E. The different feed groups were P0 (diet 1, 0 IU/kg formulated diet), P1 (diet 2, 300 IU/kg formulated diet), P2 (diet 3, 600 IU/kg formulated diet), and P3 (diet 4, 900 IU/kg formulated diet). Formulated diet8 is a modified formulation for the broodstock of the mud crab, Scylla serrata9. Blue swimming crab broodstock fed diets P0, P1, P2 and P3 were initially fed a natural diet (fresh bivalve molluska + sardinella fish; 1:1) and were gradually acclimatized to the formulated diet until the 10th day of culture. Daily feeding rates were 10% of broodstock biomass for natural food, and 3% for formulated diet (diet 1). Diets were fed twice daily at 1700 and 1800 h, with 40% of the ration given in the afternoon and the remaining 60% in the evening. Excess diet was monitored and feeding rates were adjusted accordingly. Molting and mortality were recorded daily.\n\nThe Absolute weight growth was calculated as follows: AWG = WGf – WGo, where AWG is weight gain (g), WGf is the final size/weight (g), and WGo is the weight of crab at the start of experiment (g). The Absolute carapace length was calculated as follows: ACL = CLf – CLo, where ACL is carapace length gain (mm), CLf is the carapace length of the crab at the end of experiment (mm), and CLo is the carapace length of crab at the start of experiment (mm). The Absolute carapace width was calculated as follows: ACW = CWf – CWo, where ACW is carapace width gain (mm), CWf is the carapace width of the crab at the end of experiment (mm), and CWo is the carapace width of crab at the start of experiment (mm). The carapace length, carapace width, percentage of molting and survival rates were measured as described previously7,8,10. The water quality parameters that were monitored daily were temperature (°C), salinity (ppt), pH, and water depth (cm) while dissolved oxygen (ppm) and CO2 (ppm) thrice weekly using a maximum-minimum thermometer, hand-held Atago refractometer model 8808, Thermo Orion Benctop pH meter models 410 A plus, weighted line, YSI oxygen meter model 57, and APHA11, respectively.\n\nResults were given as the means ± SE. The biological test data (absolute weight growth, absolute carapace length, absolute carapace width and survival rate) were tested using one-way ANOVA and Duncan’s multiple range test (p<0.05) to compare the mean differences among the treatments12 were performed using SPSS software (version 19.0 for Windows; SPSS Inc., Chicago, IL). The standard error of each parameter was determined. The data on percentage of molting female broodstock are shown in tables and graphs and then analyzed descriptively.\n\n\nResults\n\nGrowth in the mother crab is a good measure of the weight gain, carapace length and carapace width within a certain time after the process of molting occurs. This study shows that each of the different dietary administrations elicited a positive response in terms of the percentage of female broodstock undergoing the molting process. In total, 40–80% of females fed diet 2 and 3 had molted by day 20, but only 20% of females receiving diet 4 had molted by day 30. For those receiving diet 1, the female broodstocks did not molt until maintenance on days 40 (Table 1).\n\n*Molting broodstock female. P0 (diet 1), 0 IU/kg formulated diet (control); P1 (diet 2), 300 IU/kg formulated diet; P2 (diet 3), 600 IU/kg formulated diet; P3 (diet 4), 900 IU/kg formulated diet. Gonad maturity stages (GMS) based on Efrizal2 (II: Ovaries are light yellow/orange, III: Ovaries are yellow/orange, large and nodulated, IV: Ovaries are dark yellow/orange and V: Ovaries are light yellow, tans or yellow-orange but not bunched up).\n\nGrowth in absolute weight is a measure of the weight difference of a female reached within a certain time period compared to her weight at the beginning of the period. The average weight and absolute weight gain of female parent crabs, P. pelagicus (Linnaeus, 1758), with different dietary treatments are presented in Table 2. Growth in absolute weight is a measure of the weight difference a female reached within a certain time period compared with a weight at the beginning of the period. The average weight and absolute weight gain of female parent crabs, P. pelagicus (Linnaeus, 1758), with different dietary treatments are presented in Table 2. Females fed all formulated diets (1, 2, 3 and 4) were likely to increase their absolute weight during the maintenance period of 0 to 40 days. The results (Table 2) show the highest absolute value of the average weight (45.38 g) was obtained in those receiving diet 2, compared to diet 1 (12.69 g), diet 4 (28.52 g), or diet 3 (32.96 g); with ANOVA showing significant differences (P<0.05). Duncan’s test revealed further significant differences (P <0.05) between the treatments of diet 1 and diet 2; and diet 3 and diet 4, while the treatments of diet 2 and diet 3; and diet 3 and diet 4 did not differ significantly(P>0.05).\n\nMean values within a given column with different superscripts were significantly different (P<0.05), Values are means ± standard errors (SE). AWG, absolute weight growth; P0 (diet 1), 0 IU/kg formulated diet (control); P1 (diet 2), 300 IU/kg formulated diet; P2 (diet 3), 600 IU/kg formulated diet; P3 (diet 4), 900 IU/kg formulated diet.\n\nAbsolute carapace length is calculated from the difference between the parent crab carapace length at a certain period of time and the carapace length at the beginning of the study. The use of different diets caused relatively large changes in the growth of the absolute carapace length during the maintenance period of 40 days, which range between 1.00 and 6.23 mm (Table 3), with ANOVA showing significant differences (P<0.05). Table 3 shows accretions of the highest absolute carapace length obtained in the treatment of diet 2 (6.23 mm) compared to treatments of diet 1 (1.00 mm), diet 3 (3.76 mm) and diet 4 (3.76 mm). Duncan’s post hoc test reveals significant differences (P <0.05) between the treatments of diet 1 and diet 2, and diet 3 and diet 4, whereas diet 3 and diet 4 treatments are not significantly different (P> 0.05).\n\nMean values within a given column with different superscripts were significantly different (P<0.05). Values are means ± standard errors (SE). ACL, absolute carapace length; P0 (diet 1), 0 IU/kg formulated diet (control); P1 (diet 2), 300 IU/kg formulated diet; P2 (diet 3), 600 IU/kg formulated diet; P3 (diet 4), 900 IU/kg formulated diet.\n\nBased on the measurements (Table 4), artificial feeding with a supplementary dose of 0 IU of vitamin E (Control) provides an added value mean carapace width that is relatively low (1.00 mm) compared to artificial feeding at a dose of 900 IU of vitamin E (7.40 mm), 600 IU (7.41 mm) and vitamin E 300 IU (13.06 mm); with an ANOVA showing significant differences (P<0.05). Similarly, post hoc Duncan’s test shows significant differences (P<0.05) between the treatments of diet 1 and diet 2, while the treatments of Diet 2 with Diet 3 and diet 4 showed no significant differences (P>0.05).\n\nMean values within a given column with different superscripts were significantly different (P<0.05). Values are means ± standard errors (SE). ACW, absolute carapace width; P0 (diet 1), 0 IU/kg formulated diet (control); P1 (diet 2), 300 IU/kg formulated diet; P2 (diet 3), 600 IU/kg formulated diet; P3 (diet 4), 900 IU/kg formulated diet.\n\nThe results show that feeding different diets to the female parent crabs during the maintenance period of 40 days in controlled cultivation containers, had a high survival rate (100%) for all treatments (Table 5). A high survival rate is due to maintenance in a controlled container, where there were no deaths in the female parent crabs. This likely occurred because the water quality (physical and chemical factors) during the study was in the viable range for living crabs (Table 6).\n\nMean values within a given column with different superscripts were significantly different (P<0.05). Values are means ± standard errors (SE). H, survival rate; P0 (diet 1), 0 IU/kg formulated diet (control); P1 (diet 2), 300 IU/kg formulated diet; P2 (diet 3), 600 IU/kg formulated diet; P3 (diet 4), 900 IU/kg formulated diet.\n\nppt, parts per thousand.\n\n\nDiscussion\n\nMolting is the process of the replacement of old shell with new shell, and is a cycle that occurs in all types of arthropods, ranging from insects to crustaceans. It is important for growth, reproduction and metamorphosis13. Fujaya et al.14 explained that the activities of seeding, growth and soft shell crab production become more efficient when the mechanisms of reproduction and growth of the animals in culture are understood. In this experiment, it is clear that an artificial diet at a dose of vitamin E 300 IU/kg (diet 2), 600 IU/kg (diet 3) and 900 IU/kg affects the physiological process of the test animal known as molting (ecdysis) (Table 1). Fujaya et al.14 and Kuballa et al.13 mention that several factors stimulate the process of molting in crustaceans, namely, external information from the environment such as light, temperature and food availability. Studies using vitamin E as a nutritional source for crustaceans have been performed by several previous investigators, including Winestri et al.15 on Scylla paramamosain and Nasution et al.16 on Macrobrachium rosenbergii. These researchers mentioned that vitamin E-supplemented feed gives the best results for the growth of mangrove crabs and the fecundity of prawns.\n\nThe high absolute weight gained in the treatment with diet 2 compared to diet 1 and diet 4 is due to the improvement in nutritional quality by supplementation with tocopherols (such as vitamin E) being balanced with the composition of other nutrients in the composition of the parent artificial diet. Cahu et al.17 found that vitamin E plays a role in the improvement of the reproductive performance of crustaceans. The addition of 300 mg/kg vitamin E in the diet of crustacean broodstocks is considered to increase the potential of reproduction that affects the growth of the absolute weight of aquatic animals18. Furthermore, Izqiuerdo et al.19 explain that a lack of vitamin E results in the compromised development of reproductive organs toward mature gonads.\n\nThe difference in carapace length in those receiving diet 2 from those receiving diets 1, 3 and 4 is significant due to the higher percentage of females that undergo the molting process with diet 2 (80% at day 20 and 20% at day 30) compared to those with diet 1 (0% at day 40), diet 3 (60% at day 20, 20% at day 30), and diet 4 (40% at day 20 and 20% at day 30). According to Fujaya et al.14 and Kuballa et al.3, the growth and reproduction of crabs and other crustaceans are closely related to the molting cycle phase and control stimulation (ecdysteroids) in its hemolymph. Furthermore, four phases are described in the molting cycle; intermolt, premolt (preparation for molting), ecdysis and postmolt (recovery from molting). During intermolt, exoskeletons formed perfectly, and the crustaceans accumulated stored calcium and energy. Premolt begins when the old exoskeleton begins to separate itself and a new epidermis begins to form. The newly formed exoskeleton is larger and is still pale and soft. Postmolt is the new exoskeleton’s hardening process. The low accretion of mean carapace length of diet 1 (1.00 mm) compared to diets 2, 3 and 4 is due to the crab not experiencing replacement of the skin (molting). The added carapace length is allegedly due to changes in the shape of the carapace. Sulaeman and Hanafi20 mention that carapace length changes in the female broodstocks that do not undergo molting are caused by changes in the curvature of the back shell margin, where the eggs mature making the rear carapace increasingly convex. According to the results reported for observations of the mud crab, S. serrata, accretion is obtained when the carapace length ranges between 0 and 3 mm during a maintenance period of 35 days.\n\nThe low-growth carapace width of the female parent in those receiving diet 1 compared with the other diets happened because the treatment did not induce molting in the maintenance period from 0 to 40 days. The increasing carapace width is supposed to have occurred because of the shape changes in the carapace, which can be seen in the change in curvature of the back of the carapace. The greater the weight of the parent crab, the more convex the carapace became. The absolute changes in carapace width are relatively large in the artificial diet treatment with a dose of 300 IU vitamin E (diet 1). The diet treatment causes the broodstock females to undergo molting. Maheswarudu et al.20 and Kuballa et al.13 state that the differences in growth in the cultivation of crabs are caused by several factors, such as feeding, age, baseline weight, space and other factors. Furthermore, they explains that the more feed consumed, the larger the crab will become and the more frequent the replacement of the skin. Crab shell replacement last occurred between 17 and 26 days, and every crab molted, and increased by one-third of its original size. According to Fujaya et al.14 and Kuballa et al.13, the molting process that occurs in crustaceans in principle is caused by two factors: internal and external (such as feed, light and other factors). Thus, both factors will affect the brain and stimulate the Y organs to produce hormone molting. Molting is controlled by a steroid circulating in the exoskeleton hemolymph that stimulates the synthesis of new and the regeneration of integument lost before molting13. Ecdysteroid is a crab molting hormone, which is secreted by the organs in the form of ecdysone-Y21. Inside the hemolymph, the hormone is converted into an active hormone, 20-OH-ecdysone hydroxylase, which is present in the epidermis organs and other body tissues13,21.\n\nWater quality is one of the limiting factors in the crab culture system. The crab is active entirely in water, where it carries out functions such as respiration, excretion of wastes, feeding habits, growth and reproduction. According to Habashy and Hassan22, the required water quality parameters for the maintenance of crustaceans are salinity, temperature and pH23. Salinity affects the osmotic pressure of the water. The higher the salinity, the greater the osmotic pressure23. For the cells of all animal organs to function properly, the cells must receive a liquid medium with the appropriate composition and concentration of ions. It appears that like other marine invertebrates, a marine crustacean living in water that is isotonic with the blood has a water medium and generally has a similar ionic composition, but this differs slightly; a large difference can even occur between the ionic composition of the fluid in the cell (haemolymph) with the sea water medium. Therefore, osmoregulation is required to ensure that the intracellular and extracellular composition and concentration of the ionic liquid remains normal or balanced21,24–26. Additionally, Stiarto et al.27 and Silvia et al.28 reported that hyperosmoregulation in crustaceans requires energy in the form of protein or lipids29,30. Table 6 shows that the water salinity range of the maintenance media during the observation period was between 29.0 and 32.0 ppt. Salinity is still within the range that is highly favorable to the survival and reproductive activities of the crabs17,31,32. Environmental temperature also affects the growth and survival of aquatic organisms. Nearly all aquatic organisms are very sensitive to abrupt environmental temperature changes. At 5°C ambient temperature abrupt changes can cause stress or even death in some types of cultured organisms12. In this study, the average water temperature during the observation ranged between 26.0 and 28.0°C. The temperature is in the support range for the activities of life, growth and reproduction of the crab33,34. The pH of the water media during the study ranged from pH 7.26 to 8.00. This shows that the pH of the water was maintained within the range of neutral and slightly alkaline, and was thus suitable for living crabs. According to the Ministry of Environment in Anonymous35, water with a pH range of 6.5–8.5 has a considerable potential for the development of aquaculture in terms of productivity. Taslihan et al.36 stated that alkaline waters would be more productive than acidic waters.\n\nDissolved oxygen plays an important role in the life of all living organisms. However, there is a difference between the oxygen required by living organisms in terrestrial and aquatic organisms. Terrestrial organisms consume oxygen contained in the air, while aquatic organisms take in dissolved or bound oxygen. The oxygen requirements for every type of aquatic biota differ depending on the species toleration of the rise and fall of oxygen. In general, all types of cultured organisms (fish, shrimp, crabs, clams, and sea cucumbers) are unable to tolerate fluctuations in oxygen that are too extreme14. Dissolved oxygen (O2) was relatively high in this experiment, ranging between 6,15 to 7,56 ppm. The amount of oxygen that must be maintained to ensure a good life for aquatic organisms is not less than 3 ppm. If the oxygen content drops to less than 2 ppm, some kinds of crustaceans will be under pressure and will even die37,38. According to Millamena and Quinitio9. Crab cultivation requires dissolved oxygen at a concentration greater than 4 ppm. Although the role of carbon dioxide (CO2) is considerable for living aquatic organisms, very excessive levels would interfere, and are even toxic to cultivated biotas. The tolerance of each biota varies depending on its type and body size, and is generally no more than 15 ppm. A concentration of over 25 ppm of carbon dioxide is dangerous for cultured organisms39,40 because its presence in the blood can inhibit the binding of oxygen by hemoglobin. Furthermore, Dodd et al.,33 Long et al.,41 and Landes and Zimmer42 reported that acidification caused by CO2 has been found to strongly affect calcification rates, animal behavior, predator foraging behavior and the avoidance of predators by prey. Carbon dioxide measurements (Table 6) in the study ranged between 3.78 and 6.10 ppm. Boyd and Tucker43 explain that free carbon dioxide is good for the crustacean no higher than 12 ppm and must not be less than 2 ppm.\n\n\nConclusions\n\nThe conclusions that can be drawn from the results of this experiment are as follows: (1) Diet 2, with supplementation of 300 IU/kg vitamin E formulated diet, provided for the highest absolute weight gain (45.38 g), absolute carapace length (6.23 mm), and absolute carapace width (13.06 mm); (2) supplementation of the formulated diet with 300 IU/kg vitamin E in the formulated diet also causes broodstock blue swimming crab molting, with a percentage of 40–80% on day 20 and 20% on day 30; and (3) a survival value of 100% was obtained for all treatments during the 40 days maintenance period.\n\nFurther studies are needed to observe and analyze the effects of formulated diet supplementation of vitamin E on the incubation period and reproductive performance of female broodstock blue swimming crabs, P. pelagicus (Linnaeus, 1758) in mass production.\n\n\nData availability\n\nDataset 1. Spreadsheet containing data associated with this study. Data include time of molting, carapace width and carapace length. DOI: https://doi.org/10.5256/f1000research.15885.d22186844.",
"appendix": "Grant information\n\nThis study was supported by a research grant (Penelitian Strategis Nasional Institusi and Percepatan Guru Besar) from the Directorate of Research and Community Service, Ministry of Research Technology and Higher Education, Republic of Indonesia (No. 050/SP2H/LT/DRPM/2018), and the Research and Community Service Institution, Andalas University (No. 38/UN.16.17/PP.PGB/LPPM/2018).\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 thank the entire staff at the Fish Seed Center Beaches (BBIP) Gulf Buo and Fish Seed Center (BBI) Bungus, City of Padang, Department of Maritime and Fisheries Affairs, and the Laboratory of Animal Physiology, Department of Biology, Faculty of Math and Science, Andalas University, Padang, West Sumatra, for the technical assistance rendered.\n\n\nReferences\n\nAnonymous: FAO yearbook. Fisheries statistics. Captured production. FAO Rome, Italy. 2000; 86/1: 713. Reference Source\n\nEfrizal, Arshad A, Kamarudin MS, et al.: Some Aspects of Reproductive Biology of Blue Swimming Crab Portunus pelagicus (Linnaeus, 1758) Under Laboratory Conditions. J Fish Aquat Sci. 2015; 10(2): 77–91. Publisher Full Text\n\nHassan A, Hai TN, Chatterji A, et al.: Preliminary Study on the Feeding Regime of Laboratory Reared Mud Crab Larva, Scylla serrata (Forsskal, 1775). World Appl Sci J. 2011; 14(11): 1651–1654. Reference Source\n\nRomano N, Safee MA, Ebrahimi M, et al.: Fatty acid compositional changes during the embryonic development of the swimming crab, Portunus pelagicus (Portunidae: Decapoda). Invertebr Reprod Dev. 2016; 60(2): 112–117. Publisher Full Text\n\nJuwana S: Crab culture technique at RDCO-LIPI, Jakarta, Indonesia 1994 to 2001. Proceeding Workshop on Mariculture in Indonesia. Mataram, Lombok Island. Research Center for Oceanography-LIPI, Institute of Marine Research Norwegian Bergen-Norway. 2002; 144.\n\nEfrizal: Effects of Salinity on Survival Rate and Larval Development of Blue Swimming Crab Portunus pelagicus (Linnaeus, 1758) Under Laboratory Conditions. Aust J Basic Appl Sci. 2015; 9(7): 557–563. Reference Source\n\nEfrizal: Effects of stocking density on survival rate and larval development of blue swimming crab, Portunus pelagicus (Linnaeus, 1758) under laboratory conditions. AACL Bioflux. 2017a; 10(2): 217–226. Reference Source\n\nEfrizal, Rusnam: The Influence of Dietary Vitamin E on the Gonad Development of Blue Swimming Crab, Portunus pelagicus (Linnaeus, 1758). Australian Journal of Basic and Applied Sciences. 2017b; 11(1): 7–15. Reference Source\n\nMillamena OM, Quinitio E: The effects of diets on reproductive performance of eyestalk ablated and intact mud crab Scylla serrata. Aquaculture. 2000; 181: 81–90. Publisher Full Text\n\nEfrizal: Weight-Length and Width, and Length-Width Relationships of Juvenile Blue Swimming Crab, Portunus pelagicus (Linnaeus, 1758), from Sungai Pulai Seagrass Beds, Johor, Malaysia. Aust J Basic Appl Sci. 2014; 8(7): 277–286. Reference Source\n\nAPHA: Standard Methods for examination of water and wastewater. Washington, American Public Health Association; 2012; 1360. Reference Source\n\nSteel RDG, Torrie JH: Prinsip dan prosedur statistic. PT. Gramedia Jakarta. 748 halaman. 1990.\n\nKuballa AV, Holton TA, Paterson B, et al.: Moult cycle specific differential gene expression profiling of the crab Portunus pelagicus. BMC Genomic. 2011; 12: 147. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFujaya S, Suryati E, Nurcahyono E, et al.: Titer ekdisteroid hemolimph dan ciri morfologi rajungan (Porunus pelagicus) selama fase molting dan reproduksi. Torani. 2008; 18(3) hal. 266–274.\n\nWinestri J, Rachmawati D, dan Samidjan I: Pengaruh penambahan vitamin E pada pakan buatan terhadap pertumbuhan dan kelulushidupan kepiting bakau (scylla paramamosain). Journal of Aquaculture Management and Technology. 2014; 3(4): 40–48. Reference Source\n\nNasution R, Sastranegara MH, dan Sistina Y: Fekunditas udang galah (Macrobranchium rosenbergii) betina paasca ablasi unilateral dan suplementasi Vitamin E pada pakan. Journal bio unsoed ac id. 2013. Reference Source\n\nCahu C, Fakhfakh M, Quazuguel P: Effect of Dietary a-tocopherol level on Reproduction of Penaeus indicus. Larvi '91: Fish & Crustasean Larviculture Symposium. European Aquaculture Society, 1991; 15: 242–244.\n\nMitra G, Mukhopadhyay PK, Chattopadhyay DN: Nutritions and feeding in freshwater prawn (Macrobrachium rosenbergii) farming. Aqua Feeds Formulation & Beyond. 2005; 2(1): 17–19. Reference Source\n\nIzquierdo MS, Fernandes-Palacios H, Talcon AGJ: Effect of broodstock nutrition on reproductive performance of fish. Aquaculture. 2001; 197(1–4): 25–42. Publisher Full Text\n\nMaheswarudu G, Josileen J, Manmadhan KRN, et al.: Evaluation of the seed production and grow out culture of blue swimming crab Portunus pelagicus (Linnaeus, 1758) in India. Indian J Mar Sci. 2008; 37(3): 313–321. Reference Source\n\nLockwood APM: Aspect of the physiology of crustacean. The Aberdeen University Press. 1967. Reference Source\n\nHabashy MM, Hassan MMS: Effects of temperature and salinity on growth and reproduction of the freshwater prawn, Macrobrachium rosenbergii (Crustacea- Decapoda) in Egypt. International Journal of Environmental Science and Engineering (IJESE). 2011; 1: 83–90. Reference Source\n\nVijayan KK, Diwan AD: Influence of temperature, salinity, pH and light on molting and growth in the Indian white prawn Penaeus indicus (Crustacean: Decapoda: Penaeidae) under laboratory conditions. Asian Fish Sci. 1995; 8: 63–72. Reference Source\n\nFerraris RP, Catacutan MR, Mabelin RL, et al.: Digestibility in milkfish, Chanos chanos (Forsskal): Effects of protein source, fish size and salinity. Aquaculture. 1986; 59(2): 93–105. Publisher Full Text\n\nGilles R, Pequeux A: Interactions of chemical and osmotic regulation with environmental. In FJ. Venberg, WB.Venberg, editors. The biology of crustacea, Environmental adaptations. Academic Press, New York.1983; 8: 109–177. Reference Source\n\nRomano N, Zeng C: The effects of salinity on the survival, growth and haemolymph osmolality of early juvenile blue swimmer crabs, Portunus pelagicus. Aquaculture. 2016; 260(1–4): 151–162. Publisher Full Text\n\nSetiarto A, Strussmann CA, Takashima F, et al.: Short-term responses of adult kuruma shrimp Marsupenaeus japonicus to environmental salinity: osmotic regulation, oxygen consumption and ammonia excretion. Aquacult Res. 2004; 35(1): 669–677. Publisher Full Text\n\nSilvia GJ, Antonio URA, Francisco VO, et al.: Ammonia efflux rates and free amino acid levels in Litopenaeus vannamei postlarvae during sudden salinity changes. Aquaculture. 2004; 233(1–4): 573–581. Publisher Full Text\n\nPalacios E, Bonilla A, Luna D, et al.: Survival Na+/ K+-ATPase and lipid responses to salinity challenge in fed and starved white pacific shrimp (Litopenaeus vannamei) postlarvae. Aquaculture. 2004; 234(1–4): 497–511. Publisher Full Text\n\nSang HM, Fotedar R: Growth, survival, haemolymph osmolality and organosomatic indices of the western king prawn (Peneaus latisulcatus Kishinouye, 1896) reared at different salinities. Aquaculture. 2004; 234(1–4): 601–614. Publisher Full Text\n\nDhawan RM, DWIVEDI SN, Rajamanickam GV: Ecology of the blue crab Portunus pelagicus (Linnaeus) and its potential fishery in Zuari Estuary. Indian J Fish. 1976; 23(1 & 2): 57–64. Reference Source\n\nMeagher TD: Ecology of the crab Portunus pelagicus in South Western Australia. University of Western Australia, Australia. Ph. D. Thesis. 1971; 232. Reference Source\n\nDodd LF, Grabowski JH, Piehler MF, et al.: Ocean acidification impairs crab foraging behaviour. Proc Biol Sci. 2015; 282(1810): pii: 20150333. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWayne DS, Michael AP, Glen SS: Reproduction and Growth of the Commercial Sand Crab, Portunus pelagicus (L.) in Moreton, Bay, Queensland. Asian Fisheries Science. 1994; 7: 103–113.\n\nAnonymous: Petunjuk teknis budidaya rumput laut. Departemen Pertanian. Badan Penelitian dan Pengembangan Pertanian. Pusat Penelitian dan Pengembangan Perikanan bekerjasama dengan International Development Research Centre melalui proyek Indonesian Fisheries Information System (INFIS) Direktorat Jenderal Perikanan. 1990. Reference Source\n\nTaslihan A, Supito SE, Callinan RB: [Tiger prawn culture technique]. Direktorat Jenderal Perikanan Budidaya, Balai Besar Pengembangan Budidaya Air Payau, Jepara, [in Indonesian]. 2003; 59.\n\nRe AD, Diaz F: Effect of different oxygen concentrations on physiological energetics of blue shrimp, Litopenaeus stylirostris (Stimpson). The Open Zoology Journal. 2011; 4(1): 1–8. Publisher Full Text\n\nSEAFDEC: Mudcrab culture. Asian Aquaculture; 1997; 19: 10–25.\n\nBoyd CE: Water Quality in Warmwater Fish Ponds. Auburn University Agricultural Experiment Station. 1979. Reference Source\n\nTucker CS: Carbon dioxide. in T.L. Wellborn, Jr. and J.R. MacMillan (eds.) For Fish Farmers 84–2. Mississippi Cooperative Extension Service. 1984.\n\nLong WC, Swiney KM, Harris C, et al.: Effects of ocean acidification on juvenile red king crab (Paralithodes camtschaticus) and Tanner crab (Chionoecetes bairdi) growth, condition, calcification, and survival. PLoS One. 2013 ; 8(4): e60959. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLandes A, Zimmer M: Acidification and warming affect both a calcifying predator and prey, but not their interaction. Mar Ecol Prog Ser. 2012; 450: 1–10. Publisher Full Text\n\nBoyd CE, Tucker CS: Pond aquaculture water quality management. Kluwer Academic Publisher, Massachusetts, 1998; 700. Publisher Full Text\n\nEfrizal E, Zakaria IJ, Rusnam R, et al.: Dataset 1 in: Studies on Biological Test of Formulated Diets Supplementation of Vitamin E for the Broodstock of Females Blue Swimming Crab, Portunus pelagicus (Linnaeus, 1758). F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15885.d221868"
}
|
[
{
"id": "40545",
"date": "16 Nov 2018",
"name": "Hafrijal Syandri",
"expertise": [
"Reviewer Expertise aquaculture"
],
"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\nAbstract: In method: The author did not mention the type of vitamin E used for feed supplements. Please add the type of vitamin E used including the producer of vitamin E or name of the company. In results: The author did not mention the best dose of vitamin E as the diet supplement. Which one the best dose of Vitamin E for absolute growth, carapace length and carapace width. Keywords: please choose the unique words, not similar with the title.\n\nMethods:\nLocation: The study was conducted at the Fish Seed Center Beaches (BBIP) Gulf Buo……name of location no need translate in English, example Gulf Buo, must be written \"Teluk Buo” Supplementation: Dietary treatments, with supplementation of vitamin E, were fed daily at 3%......please add initial weight, length and width of carapace from the broadstock. Vitamin E, please state the name of the company…for example Ovaprim, (manufactured for Syndel Laboratories Ltd, 2595 McCullough Rd. Nanaimo, B.C.V9S 4M9 Canada). What type of vitamin E used in the experiment? Liquid or powder? The author did not mention the name of balance scale used during the experiment….please state the model of balance, for example fish were weighed using balance scale (OHAUS model CT 6000-USA).\nResult:\nMolting female broadstock: Growth in the mother crab is a good measure…..the word “mother” is not suitable…please change to broadstock. Absolute weight growth: The average weight and absolute weight gain of female parent crabs… the word “parent” is not suitable…please change to broadstock.\nConclusions:\nThe conclusions that can be drawn from the results of this experiment are as follows: (1) Diet 2, …….please rewrite and state only the significant finding of your research.\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": "4340",
"date": "03 Jan 2019",
"name": "Efrizal Efrizal",
"role": "Author Response",
"response": "Comment 1: Abstract has been modified. Have been revised, Answer :Nutrimax ™Vitamin E-Water Soluble; The highest absolute weight gain (45.38 g), absolute carapace length (6.23 mm), and absolute carapace width (13.06 mm) were promoted by supplementation of 300 IU/kg vitamin E in formulated diet; Keywords: Molting, ovary, gonad maturity stages, Comment 2: Methods have been modified. Have been revised; Answer :The study was conducted at Balai Benih Ikan Pantai (BBIP) Teluk Buo, Balai Benih Ikan (BBI) Bungus, Padang; 20 were females at stage II of ovarian maturity, which were selected for study. The female crab samples with mean body weigth (BW) of 158.15 g, carapace length of 57.27 mm and carapace width of 123.54 mm were collected from the coastal region of Padang, West Sumatera; Weights were measured to 0.01 g on the electronic balances (BL3200H-SHIMADZU). Comment 3: Result not revised Comment 4: Conclusion : have been revised; Answer: (1) Fdiet 2 with supplementation of 300 IU/kg vitamin E formulated diet, provided for the highest absolute weight gain (45.38 g), absolute carapace length (6.23 mm), and absolute carapace width (13.06 mm); 2) supplementation of the formulated diet with 300 IU/kg vitamin E in the formulated diet also causes broodstock blue swimming crab molting, with a percentage of 40–80% on day 20 and 20% on day 30; and; (3) a survival value of 100% was obtained for all treatments during the 40 days maintenance period."
}
]
},
{
"id": "40546",
"date": "04 Dec 2018",
"name": "Ambok Bolong Abol-Munafi",
"expertise": [
"Reviewer Expertise Aquaculture"
],
"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 English revision needed trough out the manuscript.\nTitle The title did not reflect the contents of the manuscript. The term of 'Biological Test' did not appear in the text.\nIntroduction Should highlight the common parameters for biological test. Authors should also discuss the functions of vit E in fishes and the type of vit E that commonly used.\n\nMethods The methodology should provide:\nInformation on the quality of base diet, especially on the protein and fat contents. Brief explanation on how to identify the ovarian maturity stages of live samples Explanation on type of vitamin E used and how it apply to the feed. Why IU unit was used instead of mg/g diet. Only 20 stage II broodstocks were used. One crab per replicate (5 replicate per treatment). Not enough for statistical analysis\nResults\nMolting\nWhy molting occurred exactly on 10, 20, 30 and 40 days. Actual data on molting day should be presented. Table 1 showing more on maturity stages compare to molting.\nDiscussions Should discuss in detail the role of vit E on molting\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? 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": "4339",
"date": "03 Jan 2019",
"name": "Efrizal Efrizal",
"role": "Author Response",
"response": "Comment 1: Answer :English grammar have been revised. We have send to American Journal Expert (AJE), certificate attached. Comment 2: Answer: The biological test in the title have been explained in the method analysis. The biological analysis include absolute weight, absolute carapace length, absolute carapace width, survival rate and molting female broodstock. Comment 3: Answer: we have discuss it. Comment 4: Methodology. The quality of base diet not discuss in this manuscript because we would like to publish in the next article. The ovary condition can be seen by the pressing of the gap between the back and the abdomen of the crab. Explanation on type of vitamin E used and how it apply to the feed. Why IU unit was used instead of mg/g diet. We have answer the question from reviewer Hafrijal Syandri (Nutrimax ™Vitamin E-Water Soluble) The IU is an International Unit, usually used to measure fat soluble vitamins including Vitamin E. I think 5 replicate per treatment enough. Because the canibal behaviour from the crab. During the experiment we use plastic box. Each plastic box have the maximum density only one crab. 5. Comment 5. Molting. Answer: my observation every 10 days. Data molting.....see Table 1. Have been presented. 6. Comment 6: Answer: Discussion: Have been discussed detail in the manuscript."
},
{
"c_id": "4375",
"date": "15 Feb 2019",
"name": "Ambok Bolong Abol-Munafi",
"role": "Reviewer Response",
"response": "1. The quality of based diet should be discussed in this manuscript because it is one of the important affecting the results too.2. The ovary condition can be seen by the pressing of the gap between the back and the abdomen of the crab? - Are the authors is very sure about it? Because it is very impossible to say that the crabs were in stage 2 ovarian maturation/ conditions only by pressing the gap between the back of the crabs' abdomen because the crabs might be either at stage one or stage three. However, if the authors confirm about the methods to identify the stage, PLEASE insert a related reference about it."
}
]
},
{
"id": "40547",
"date": "17 Dec 2018",
"name": "Zainal Abidin Muchlisin",
"expertise": [
"Reviewer Expertise Fish Biology and Aquaculture"
],
"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\nTitle: I suggest to modify \"Growth performance, survival rate and moulting frequency of blue swimming crab (Portunus pelagic, Linnaeus, 1978) fed on varying level of vitamin E\nIntroduction: The introduction is lack of state of the arts of the study on blue swimming crab and vitamin E. Therefore, the author must emphasize and cite the current studies or previous findings by other researchers on blue swimming crab and the role of vitamin E on crab or aquatic organism.\nThere was no clear objective of the study. So please state the objective clearly at the end of introduction section.\nMethods: Fish Seed Centre Beach is not appropriate term of BBIP., need to change. Where the crab is come from? The author should mention whether the feed used is a commercial diet or experimental diet (formulated by researcher for this study). And also the proximate composition of the diets are unclear. So the author must present the composition of the nutrition content (proximate composition) of the tested diet, maybe presented in the table would be better. How the procedure to mix the vitamin E with the diet?\n\nExisting: The method used in this study was completely randomised design methods with ......\" Suggestion: The completely randomised design method with ...was used in this study\"\nResults: Table 1 is unclear. The author stated table 1 presented the data of percentage of moulting female brood stock. Actually, The percentage value should be ranged 0-100%. But the author provided the value between II - IV. In my understanding this is not percentage data, but this is the maturity stage of ovary. The authors have to clarify.\nThe author should mention the results of the ANOVA test whether the experiment gave the significant effect on measured parameters or not. And the author also have to mention briefly the differences values (data) among the treatments (based on Duncan's multi rage test). Please see superscript after the value.\nDiscussion: In general the discussion is already acceptable, but maybe the author has to cite more references to support the discussion, especially from reputable related journals.\nConclusions: The conclusion should be clear and strong if possible. Your conclusion should be focused on the objective.\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": "4338",
"date": "03 Jan 2019",
"name": "Efrizal Efrizal",
"role": "Author Response",
"response": "Comment 1: Answer: Title: I have read carefully, the suitable title from my manuscript is“Studies on Biological Test of Formulated Diets Supplementations of Vitamin E for the Broodstock of Females Blue Swimming Crabs, Portunus pelagicus (Linnaeus, 1758)” Comment 2: Answer: Introduction: We have been searching other reference, we could not find the latest research of blue swimming crab and vit E. Comment 3: Answer: Methods: have been modified. Comment 4: Answer: Results: see Table 1. The red supercript shows the percentage of molting broodstock female. ANOVA: We have stated the significant differences in the treatment. Different supercript letter means significant differences."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1780
|
https://f1000research.com/articles/8-11/v1
|
03 Jan 19
|
{
"type": "Study Protocol",
"title": "Identifying stroke therapeutics from preclinical models: A protocol for a novel application of network meta-analysis",
"authors": [
"Manoj M. Lalu",
"Dean A. Fergusson",
"Wei Cheng",
"Marc T. Avey",
"Dale Corbett",
"Dar Dowlatshahi",
"Malcolm R. Macleod",
"Emily S. Sena",
"David Moher",
"Risa Shorr",
"Sarah K. McCann",
"Laura J. Gray",
"Michael D. Hill",
"Annette O'Connor",
"Kristina Thayer",
"Fatima Haggar",
"Aditi Dobriyal",
"Hee Sahng Chung",
"Nicky J. Welton",
"Brian Hutton",
"Dean A. Fergusson",
"Wei Cheng",
"Marc T. Avey",
"Dale Corbett",
"Dar Dowlatshahi",
"Malcolm R. Macleod",
"Emily S. Sena",
"David Moher",
"Risa Shorr",
"Sarah K. McCann",
"Laura J. Gray",
"Michael D. Hill",
"Annette O'Connor",
"Kristina Thayer",
"Fatima Haggar",
"Aditi Dobriyal",
"Hee Sahng Chung",
"Nicky J. Welton",
"Brian Hutton"
],
"abstract": "Introduction: Globally, stroke is the second leading cause of death. Despite the burden of illness and death, few acute interventions are available to patients with ischemic stroke. Over 1,000 potential neuroprotective therapeutics have been evaluated in preclinical models. It is important to use robust evidence synthesis methods to appropriately assess which therapies should be translated to the clinical setting for evaluation in human studies. This protocol details planned methods to conduct a systematic review to identify and appraise eligible studies and to use a network meta-analysis to synthesize available evidence to answer the following questions: in preclinical in vivo models of focal ischemic stroke, what are the relative benefits of competing therapies tested in combination with the gold standard treatment alteplase in (i) reducing cerebral infarction size, and (ii) improving neurobehavioural outcomes? Methods: We will search Ovid Medline and Embase for articles on the effects of combination therapies with alteplase. Controlled comparison studies of preclinical in vivo models of experimentally induced focal ischemia testing the efficacy of therapies with alteplase versus alteplase alone will be identified. Outcomes to be extracted include infarct size (primary outcome) and neurobehavioural measures. Risk of bias and construct validity will be assessed using tools appropriate for preclinical studies. Here we describe steps undertaken to perform preclinical network meta-analysis to synthesise all evidence for each outcome and obtain a comprehensive ranking of all treatments. This will be a novel use of this evidence synthesis approach in stroke medicine to assess pre-clinical therapeutics. Combining all evidence to simultaneously compare mutliple therapuetics tested preclinically may provide a rationale for the clinical translation of therapeutics for patients with ischemic stroke. Dissemination: Review findings will be submitted to a peer-reviewed journal and presented at relevant scientific meetings to promote knowledge transfer. Registration: PROSPERO number to be submitted following peer review.",
"keywords": [
"stroke",
"preclinical",
"systematic review",
"network metaanalysis",
"network meta-analysis"
],
"content": "List of abbreviations\n\nNMA, network meta-analysis; DIC, deviance information criteria; NMD, normalized mean difference\n\n\nIntroduction\n\nGlobally, an estimated 15 million people suffer a stroke; stroke is the second leading cause of death, with six million people dying and an additional five million becoming permanently disabled each year1,2. The costs of stroke are high due to a combination of immediate high costs from acute care and long-term costs from resulting disability. Worldwide cost estimates range from $266 billion to $1.038 trillion per year3. Despite the enormous human and economic burden, only four acute interventions are currently used clinically: patient care in a dedicated stroke unit3, reperfusion (by pharmacological thrombolysis or endovascular mechanical thrombectomy4), oral aspirin, and surgical decompression.\n\nIn the search for novel therapies for acute stroke, more than 1,000 potential neuroprotective therapeutics (e.g. anticoagulants, calcium channel blockers, free radical scavengers, GABA mimetics, etc.) have been evaluated in preclinical models5. Of these, only reperfusion with tissue plasminogen activators6, such as alteplase3, has had a preclinical basis. Despite its efficacy, alteplase has inherent limitations such as the risk of hemorrhagic transformation, which warrants exploration of novel adjunctive therapies that can maximize therapeutic benefit. Combination therapies with alteplase might limit reperfusion injury and cell death that can sometimes occur with this drug. However, given the multitude of therapies tested preclinically (and multiple mechanisms of action) it is difficult to assess which therapies should proceed to clinical testing.\n\nPreclinical systematic reviews have served as a robust form of knowledge synthesis to evaluate transparently experimental therapies for more than a decade7–9. Previous preclinical systematic reviews have compared treatments in isolation using pair-wise meta-analyses, which limits the ability to simultaneously evaluate comparative effectiveness in the presence of many treatments of interest. Use of network meta-analysis (NMA) in comparative effectiveness research to study the relative benefits and harms of multiple interventions in humans9,10 has risen dramatically during the past decade11. Such analyses allow the comparison of many interventions based on all ‘direct’ and ‘indirect’ information. In addition, this approach has the potential to establish a more rigorous framework for decisions to embark on clinical trials while reducing risks to human trial participants and the enormous costs of preclinical translation3,7,12. Comparison of preclinical stroke therapeutics represents an excellent case study for such work. Given the novelty of this approach, this systematic review will also serve as a case study to empirically explore the methodological nuances of applying NMA in a preclinical setting.\n\nThis protocol will be registered in the international prospective register of systematic reviews (PROSPERO, CRD) following peer review. Our review protocol is reported in accordance with the Preferred Reporting Items in Systematic reviews and Meta-Analysis-Protocol guidelines (a complete checklist is available as Supplementary File 1)13. Post-protocol adjustments will be included in the final report.\n\nPrimary objective. We will perform a systematic review and NMAs to address the following question: amongst in vivo models of focal ischemic stroke, what are the relative benefits of competing therapies tested in combination with the gold standard treatment alteplase14 in (i) reducing cerebral infarction size, and (ii) improving neurobehavioural outcomes?\n\nSecondary Objective. We will also (i) assess the risk of bias of the included studies, and (ii) explore what novel considerations for statistical adjustments are necessary for NMA of preclinical studies (e.g. method of ischemic induction, timing of treatment, species, sex, and comorbidities). We will also evaluate the challenges of applying NMA to preclinical studies (e.g. consistency, heterogeneity, availability of key study covariates).\n\n\nMethods\n\nAn information specialist (RS) will construct a search strategy based on a previous review of comparative stroke therapies, and limit them to include studies which compared therapies to alteplase (representative search strategy is provided in Supplementary File 2)15. Search strategies will be peer reviewed by a second information specialist using the peer review of electronic search strategy method16. Searches of Ovid MEDLINE and Embase will be carried out for articles on the effects of combination therapies with alteplase (Supplementary File 2 contains the search strategy). Of note, no language or date restrictions will be used. We will also search the CAMARADES database which contains data extracted from existing preclinical systematic reviews on stroke15,17–28. In addition to this rigorous search, we will assess bibliographies of any new studies and reviews identified. Articles in foreign languages will be translated.\n\nEligibility criteria to identify relevant studies for the current review were established in considering the Population-Intervention-Comparators-Outcomes-Study design (PICOS) framework29.\n\nPopulation. Preclinical in vivo models of experimentally induced focal ischemia will be sought. All species/strains of animals will be eligible. Both female and male animals will be included. Neonatal animals will be excluded; however, all other ages will be considered. Studies in which focal ischemic stroke was established by transient occlusion of the middle cerebral artery or anterior cerebral artery via any method (chemical, embolic, mechanical, thermal) will be eligible. Animal models of haemorrhagic stroke, global or hemispheric brain ischemia, models of permanent occlusion without reperfusion (e.g. photothrombosis, cauterization), or delayed reperfusion such that it is considered permanent will be excluded30. Human studies and tissue culture studies will be excluded.\n\nIntervention and comparator. Studies where the treatment in combination with alteplase (e.g. alteplase + hypothermia) is compared with alteplase alone in animals that have experimentally induced focal ischemia will be eligible. Studies that compare more than one active treatment such as alteplase + hypothermia versus alteplase +FK506 (i.e. head to head comparisons) will also be included. Studies must include alteplase as a ‘foundational’ therapeutic in experimental arms to be eligible. All delivery routes and doses will be considered. To increase potential clinical relevance (i.e. construct validity), only studies that deliver therapies within 6 hours of induction of focal ischemic stroke will be included.\n\n- Primary outcome. Infarct size is a measure of injury reduction at the infarct site in the brain and can be measured via a variety of quantifiable techniques through non-invasive techniques (e.g. T2-weighted magnetic resonance imaging) or post-mortem analysis (e.g. staining of brain sections using hematoxylin and eosin). This is the most widely reported outcome in preclinical stroke studies. Infarct size outcomes will be extracted at the latest time point for each study. Separate time-point specific analyses will be conducted (e.g. an early time point <30 days vs later time points >30 days).\n\n- Secondary outcome. Neurobehavioural measures represent a valuable means of assessing functional recovery after treatment. Neurobehavioural assessment are sensitive to detecting the array of impairments, including motor/sensory deficits (e.g. ladder rung walking—foot slip errors) as well as memory/learning deficits (e.g. Morris water maze)31,32. These outcomes, while labour-intensive, are typically reported with less frequency than infarct volume even though functional outcomes may have the greatest clinical relevance)33,34. Neurobehavioral outcomes will be extracted at all timepoints and separate time-point specific analyses will be conducted as described above.\n\nControlled comparison studies testing the efficacy of therapies + alteplase versus alteplase alone will be sought.\n\nTwo reviewers (A.D. and H.S.C.), will review abstracts (Stage 1 screen) and full text reports (Stage 2 screen) from search results independently and in duplicate against the eligibility criteria below using Distiller SR® software (Evidence Partners, Ottawa, ON) to identify relevant articles. Discrepancies will be resolved through discussion with a senior team member (M.L. and D.C.). Both stages of screening will begin with a calibration exercise to ensure consistent application of eligibility criteria. A PRISMA flow diagram35 will be presented to document the process of study selection.\n\nTwo independent reviewers (A.D. and H.S.C.) will review studies and extract data into standardized, piloted forms implemented in Microsoft Excel (Microsoft Corporation, Seattle, Washington, USA). Discrepancies will be resolved through discussion with a senior team member. We will collect data related to, but not limited to, animal characteristics (Table 1); stroke model (Table 1); intervention (Table 2a, b); and outcomes (Table 3), as well as study ID (authors, year), and study design characteristics. Measures of central tendency (e.g. mean) and dispersion (e.g. standard deviation) will be extracted as reported. Data in graphical format will be extracted using Engauge Digitizer36. When measures of central tendency and dispersion or sample sizes are missing (or cannot be measured digitally), authors will be contacted; if authors do not respond, the data will be excluded.\n\nTwo independent reviewers will assess the risk of bias of each included study (quality of the design, conduct and analysis for the experiment)37. We will assess the risk of bias using a modified version of the Cochrane Risk of Bias Tool for randomized trials (Table 4). Risk of bias will be summarized38 with descriptive statistics and presented graphically using standard methods and radar charts. The assessment of risk of bias will play an important role in exploring potential limitations of the evidence base and establishing the feasibility of incorporating relevant adjustments in NMA models. The construct validity of included studies (i.e. degree to which experimental model and design reflect the clinical entity of stroke and its treatment) will be assessed using elements from the CAMARADES checklist alongside criteria established by expert consensus (Table 5).\n\nWe will begin by exploring the pattern of treatment comparisons represented by the included set of studies using network diagrams (or using a tabular approach if necessary, should the number and pattern of comparisons be too broad to be summarized graphically). Effect estimates from all included studies will be summarized. We will summarize traits of included studies focusing on clinical (e.g. age, sex, species, stroke model, reperfusion vs. permanent model, comorbidities, severity of infarct pre-treatment, infarct location)39 and methodological (e.g. risk of bias, timing of outcome assessment) features27, and review these with our clinical and preclinical experts to establish the degree of homogeneity within the included studies. For NMA, given the possibility that a large proportion of the studied interventions may have been evaluated in only a single study (and many could potentially yield very large effect sizes, which may not have been substantiated by more animals in more studies), we will exclude these interventions from NMAs performed; each of these treatments removed from NMA will neither benefit from “borrowing strength” through NMA, nor end up with a summary estimate and confidence interval different from what was reported in a single study. The reported findings for the outcomes of interest from studies removed from the NMA according to these criteria will be summarized separately in descriptive tables to ensure all relevant data are summarized. This approach will also restrict the network to a more practical size and reduce the risk of computational challenges.\n\nWhere there is homogeneity of important effect modifiers, we will perform NMAs to compare interventions9,10,40, following procedures to assess the validity of the assumptions of homogeneity, similarity, and consistency41. Based upon the extracted study characteristics, we will work with our clinical and preclinical experts to establish any additional novel aspects of preclinical studies that may be important to consider in relation to judgements regarding study homogeneity beyond those anticipated in preparing this protocol. We have anticipated different species of animals (rats, mice, gerbils, dogs, sheep, non-human primates) across studies. We also anticipate that multiple reporting formats will have been used to assess both infarct volume (e.g. mm3, % of hemisphere or total brain, etc.) and neurobehavioral changes (e.g. seconds, % of baseline).\n\nFor meta-analysis of preclinical studies, the normalized mean difference (NMD) scale is useful in serving the purpose of synthesizing the complexity of data aforementioned42. Prior to performing NMAs, we will perform traditional pairwise meta-analyses on the NMD scale for each comparison in the treatment networks where two or more studies are available to explore heterogeneity based on the I2 statistic29. To perform network meta-analyses on the NMD scale, we will use an established model from the National Institute in Health and Care Excellence’s TSD series43, adapting its identity link to the log link in order to conduct the NMA on a log ratio of means (logRoM) scale. The log ratio of means of the kth treatment and the “stroke only” control, dk = logRoMC,Tk, can be estimated after model fitting, and the corresponding NMD estimate is:\n\n1 – exp{logRoMC,Tk} = 1 – exp (dk)\n\nThe NMD of the kth treatment in comparison with alteplase (k=1) is 1 – exp (dk – d1).\n\nBoth fixed- and random-effect Bayesian NMAs will be performed using a common heterogeneity parameter according to established methods10,40,43. Model fit will be assessed by comparing the model’s posterior total residual deviance with the number of unconstrained data points43. Selection between models will be based on deviance information criteria (DIC), with a difference of five points suggesting an important difference43. All pairwise comparisons between interventions will be expressed with both summary point estimates and corresponding 95% credible intervals. Vague prior distributions will be assigned for all measures of treatment effect, as well as for the between-study variance parameter in random effects analyses. NMAs will be performed using OpenBUGS software version 3.2.344 and the R Package R2OpenBUGS45. Model convergence will be assessed using established methods including assessment of Rhat (the potential scale reduction factor) and the Gelman-Rubin convergence diagnostic to see if they are near 19. Surface under the cumulative ranking (SUCRA) values, and the mean rank of each intervention (with 2.5% and 97.5% quantiles) will also be estimated for each intervention46. Forest plots of treatment comparisons versus “stroke with no treatment” control as well as versus stroke + alteplase will be prepared for each outcome. Given the anticipated high number of interventions assessed in only a single study, a tabular approach to summarizing findings will be employed for them. We will also undertake forest plots of effects wherein interventions are ordered according to mean rank estimated from NMA.\n\nTo check the validity of the consistency assumption (i.e., transitivity of the effect size through common comparators), a consistency model as well as an unrelated mean effects model will be fit to the data47. We will compare their respective DIC values to check model fitting and their posterior mean deviance contribution per study to check the consistency assumption. We will also assess the magnitude of the estimated between-study SD measure from both models, as a reduction in this parameter in the inconsistency model also provides evidence of inconsistency.\n\nThe likelihood of important clinical and methodological heterogeneity between studies is anticipated by the research team to be high and may include several nuances which are unique to the pre-clinical setting. First, several vital aspects of preclinical studies from our risk of bias assessments (described earlier) may be important adjustment factors that could have an important impact on the findings from NMAs, including randomization and blinding24,48. In this work, we will use subgroup analyses or covariate-adjusted analyses to address and explore the impact that covariates have on findings and to establish the robustness of findings from primary syntheses49,50. We will assess the possibility to adjust for the following group level factors: animal species (e.g. mouse) and strain (e.g. C57Bl6 strain of mice), model of stroke, average animal age, percentage of female subjects, average time since stroke induction, combination therapies, cerebral blood flow, temperature, infarct location and severity, use of randomization and blinding of experimenters and outcome assessments. Alternatively, when combining data from different species, we could model animal species as an extra level in the hierarchical model for treatment effect, allowing for heterogeneity across species and assuming that treatment effects are similar across species around an overall mean effect. For the network structure, primary analyses will be performed at the treatment level. As dose may have an important effect on intervention benefits, we will also explore the range of doses associated with each intervention across studies to consider additional analyses. However, as dose response characteristics of different agents may also vary between animal species and an a priori source of information to establish appropriate dose categories is not available, any analyses pursued in relation to dose will be appropriately indicated as post-hoc. Findings from all analyses will be reported. Given the anticipated complexity of this novel application of NMA, we anticipate separate publications will be required for the primary and secondary outcomes.\n\nThe results of the study will be submitted for publication to a peer-reviewed journal and presented at relevant national and international conferences and scientific meetings to promote knowledge transfer.\n\nIf amendments are required for this protocol, date of each amendment will be provided with a description for rationale for the change in this section.\n\n\nDiscussion\n\nCurrent approaches to evaluating the relative therapeutic benefit of preclinical treatments for stroke are limited. Although systematic reviews have been conducted comparing more than a thousand candidates, many have never been systematically assessed, nor have they been assessed relative to one another, or more importantly, to the best available clinical treatment (alteplase). Use of NMA to synthesize data on all relevant available therapies may help address this knowledge gap. Thinking more broadly, the proposed review, with the application and evaluation of NMA to preclinical therapeutics, will inform translational scientists’ knowledge of which preclinical stroke therapeutics have the most promise for either further preclinical research or translation to clinical trial.\n\nIn addition to addressing an important question for clinical research, we anticipate this study will inform empirical explorations of anticipated challenges of evidence synthesis that are unique to the pre-clinical setting. First, a debate among preclinical and clinical scientists is likely to exist regarding both the appropriateness and approach to synthesizing outcome data from different species as well as different models of stroke. Second, there exists an especially important need to consider a broad range of adjustments to account for between-study heterogeneity related to animal characteristics or other features; lack of availability of these key data may be sub-optimal. Our study will provide an empirical evaluation of the degree of missingness of features, such as those significant to experimental design (e.g. randomization). This will provide an indication of the changes to the available evidence when exploring adjustments of comparisons. More specifically, if the lack of reporting proves to be severe, this will provide further high-level evidence that educational efforts are needed to improve the completeness of reporting of preclinical research51.\n\nOther challenges potentially requiring consideration will include identifying optimal strategies for presenting findings (including those with many comparators rendering analysis unfeasible), analysis of studies with small sample sizes, and strategies to select the most promising therapy to translate clinically. We anticipate that this systematic review will provide insight into these and other methodologic challenges and thereby serve as an exemplar for future NMA of preclinical data to build upon.\n\nFindings from this review will be shared with several key knowledge users including (i) the Stroke Treatment Academic Industry Roundtable52 for development of future guidelines; (ii) the Heart & Stroke Foundation and the Canadian Partnership for Stroke Recovery to inform future potential trials51; (iii) the Cochrane Stroke Group to inform a future clinical systematic review and NMA; and (iv) stroke survivors, via sharing of findings with our knowledge users.\n\n\nData availability\n\nNo data are associated with this article.",
"appendix": "Author contributions\n\n\n\nM.M.L. contributed to protocol design, with specific methodological input on systematic review methods (B.H., D.M., D.A.F.) and statistical analysis and synthesis (B.H., W.C., D.M., D.F., N.J.W.). R.S. designed the search strategy. D.C., D.D., S.K.M., provided substantive topic-specific input that informed the protocol's revision and refinement. M.M.L. and B.H. drafted the manuscript. All authors read and approved the final manuscript.\n\n\nGrant information\n\nThis work was supported by Canadian Institutes of Health Research (Grant #365473). M.M.L. is supported by The Ottawa Hospital Anesthesia Alternate Funds Association and the Scholarship Protected Time Program, Department of Anesthesiology and Pain Medicine, uOttawa. D.M. is supported by a University Research Chair. N.J.W. was supported by the NIHR Biomedical Research Centre at the University Hospitals Bristol NHS Foundation Trust and the University of Bristol. The views expressed in this publication are those of the authors and not necessarily those of the funding bodies.\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\nWe thank Sarah Schlievert for administrative assistance.\n\n\nSupplementary materials\n\nSupplementary File 1. PRISMA-Protocols checklist.\n\nClick here to access the data.\n\nSupplementary File 2. Representative search strategy.\n\nClick here to access the data.\n\n\nReferences\n\nSudlow CL, Warlow CP: Comparable studies of the incidence of stroke and its pathological types: results from an international collaboration. International Stroke Incidence Collaboration. Stroke. 1997; 28(3): 491–9. PubMed Abstract\n\nWorld Heart Federation: Stroke - The Global Burden of Stroke. 2017.\n\nHowells DW, Sena ES, O'Collins V, et al.: Improving the efficiency of the development of drugs for stroke. Int J Stroke. 2012; 7(5): 371–7. PubMed Abstract | Publisher Full Text\n\nGoyal M, Demchuk AM, Menon BK, et al.: Randomized Assessment of Rapid Endovascular Treatment of Ischemic Stroke. N Engl J Med. 2015; 372(11): 1019–1030. PubMed Abstract | Publisher Full Text\n\nO'Collins VE, Macleod MR, Donnan GA, et al.: 1,026 experimental treatments in acute stroke. Ann Neurol. 2006; 59(3): 467–477. PubMed Abstract | Publisher Full Text\n\nKarimi M, Mohammadi H, Hemmatzadeh M, et al.: Role of the HTLV-1 viral factors in the induction of apoptosis. Biomed Pharmacother. 2017; 85: 334–347. PubMed Abstract | Publisher Full Text\n\nSena ES, Currie GL, McCann SK, et al.: Systematic reviews and meta-analysis of preclinical studies: why perform them and how to appraise them critically. J Cereb Blood Flow Metab. 2014; 34(5): 737–742. PubMed Abstract | Publisher Full Text | Free Full Text\n\nde Vries RB, Wever KE, Avey MT, et al.: The usefulness of systematic reviews of animal experiments for the design of preclinical and clinical studies. ILAR J. 2014; 55(3): 427–437. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCatalá-López F, Tobias A, Cameron C, et al.: Network meta-analysis for comparing treatment effects of multiple interventions: an introduction. Rheumatol Int. 2014; 34(11): 1489–96. PubMed Abstract | Publisher Full Text\n\nIoannidis JP: Integration of evidence from multiple meta-analyses: a primer on umbrella reviews, treatment networks and multiple treatments meta-analyses. CMAJ. 2009; 181(8): 488–493. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNikolakopoulou A, Chaimani A, Veroniki AA, et al.: Characteristics of networks of interventions: a description of a database of 186 published networks. PLoS One. 2014; 9(1): e86754. PubMed Abstract | Publisher Full Text | Free Full Text\n\nScannell JW, Blanckley A, Boldon H, et al.: Diagnosing the decline in pharmaceutical R&D efficiency. Nat Rev Drug Discov. 2012; 11(3): 191–200. PubMed Abstract | Publisher Full Text\n\nShamseer L, Moher D, Clarke M, et al.: Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015: elaboration and explanation. BMJ. 2015; 350: g7647. PubMed Abstract | Publisher Full Text\n\nAmerican Stroke Association: Stroke Treatments. 2013; 2017.\n\nO'Collins VE, Macleod MR, Donnan GA, et al.: Evaluation of combination therapy in animal models of cerebral ischemia. J Cereb Blood Flow Metab. 2012; 32(4): 585–97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSampson M, McGowan J, Cogo E, et al.: An evidence-based practice guideline for the peer review of electronic search strategies. J Clin Epidemiol. 2009; 62(9): 944–52. PubMed Abstract | Publisher Full Text\n\nMcCann SK, Cramond F, Macleod MR, et al.: Systematic Review and Meta-Analysis of the Efficacy of Interleukin-1 Receptor Antagonist in Animal Models of Stroke: an Update. Transl Stroke Res. 2016; 7(5): 395–406. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcCann SK, Irvine C, Mead GE, et al.: Efficacy of antidepressants in animal models of ischemic stroke: a systematic review and meta-analysis. Stroke. 2014; 45(10): 3055–63. PubMed Abstract | Publisher Full Text\n\nEgan KJ, Janssen H, Sena ES, et al.: Exercise reduces infarct volume and facilitates neurobehavioral recovery: results from a systematic review and meta-analysis of exercise in experimental models of focal ischemia. Neurorehabil Neural Repair. 2014; 28(8): 800–12. PubMed Abstract | Publisher Full Text\n\nWu S, Sena E, Egan K, et al.: Edaravone improves functional and structural outcomes in animal models of focal cerebral ischemia: a systematic review. Int J Stroke. 2014; 9(1): 101–6. PubMed Abstract | Publisher Full Text\n\nVesterinen HM, Currie GL, Carter S, et al.: Systematic review and stratified meta-analysis of the efficacy of RhoA and Rho kinase inhibitors in animal models of ischaemic stroke. Syst Rev. 2013; 2: 33. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJerndal M, Forsberg K, Sena ES, et al.: A systematic review and meta-analysis of erythropoietin in experimental stroke. J Cereb Blood Flow Metab. 2010; 30(5): 961–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBanwell V, Sena ES, Macleod MR: Systematic review and stratified meta-analysis of the efficacy of interleukin-1 receptor antagonist in animal models of stroke. J Stroke Cerebrovasc Dis. 2009; 18(4): 269–76. PubMed Abstract | Publisher Full Text\n\nMacleod MR, van der Worp HB, Sena ES, et al.: Evidence for the efficacy of NXY-059 in experimental focal cerebral ischaemia is confounded by study quality. Stroke. 2008; 39(10): 2824–9. PubMed Abstract | Publisher Full Text\n\nvan der Worp HB, Sena ES, Donnan GA, et al.: Hypothermia in animal models of acute ischaemic stroke: a systematic review and meta-analysis. Brain. 2007; 130(Pt 12): 3063–74. PubMed Abstract | Publisher Full Text\n\nMacleod MR, O'Collins T, Horky LL, et al.: Systematic review and meta-analysis of the efficacy of melatonin in experimental stroke. J Pineal Res. 2005; 38(1): 35–41. PubMed Abstract | Publisher Full Text\n\nMacleod MR, O'Collins T, Howells DW, et al.: Pooling of animal experimental data reveals influence of study design and publication bias. Stroke. 2004; 35(5): 1203–8. PubMed Abstract | Publisher Full Text\n\nSena ES, Briscoe CL, Howells DW, et al.: Factors affecting the apparent efficacy and safety of tissue plasminogen activator in thrombotic occlusion models of stroke: systematic review and meta-analysis. J Cereb Blood Flow Metab. 2010; 30(12): 1905–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHiggins J, Green S: Cochrane Handbook for Systematic Reviews of Interventions. 2017; 2017. Reference Source\n\nMacrae IM: Preclinical stroke research--advantages and disadvantages of the most common rodent models of focal ischaemia. Br J Pharmacol. 2011; 164(4): 1062–78. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBalkaya M, Kröber JM, Rex A, et al.: Assessing post-stroke behavior in mouse models of focal ischemia. J Cereb Blood Flow Metab. 2013; 33(3): 330–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSchaar KL, Brenneman MM, Savitz SI: Functional assessments in the rodent stroke model. Exp Transl Stroke Med. 2010; 2(1): 13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCorbett D, Nurse S: The problem of assessing effective neuroprotection in experimental cerebral ischemia. Prog Neurobiol. 1998; 54(5): 531–548. PubMed Abstract | Publisher Full Text\n\nCorbett D, Carmichael ST, Murphy TH, et al.: Enhancing the alignment of the preclinical and clinical stroke recovery research pipeline: Consensus-based core recommendations from the Stroke Recovery and Rehabilitation Roundtable translational working group. Int J Stroke. 2017; 12(5): 462–471. 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. BMJ. 2009; 339: b2535. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMitchell M, Muftakhidinov B, Winchen T, et al.: Engauge Digitizer Software. Reference Source\n\nMoher D, Jadad AR, Nichol G, et al.: Assessing the quality of randomized controlled trials: an annotated bibliography of scales and checklists. Control Clin Trials. 1995; 16(1): 62–73. PubMed Abstract | Publisher Full Text\n\nKilkenny C, Browne WJ, Cuthill IC, et al.: Improving bioscience research reporting: the ARRIVE guidelines for reporting animal research. PLoS Biol. 2010; 8(6): e1000412. PubMed Abstract | Publisher Full Text | Free Full Text\n\nO’Collins VE, Donnan GA, Macleod MR, et al.: Animal models of stroke versus clinical stroke: comparison of infarct size, cause, location, study design, and efficacy of experimental therapies. In: M. C. P, ed. Animal Models for the Study of Human Disease. 2013: Waltham: Academic Press; 2013; 531–568. Publisher Full Text\n\nLu G, Ades AE: Combination of direct and indirect evidence in mixed treatment comparisons. Stat Med. 2004; 23(20): 3105–24. PubMed Abstract | Publisher Full Text\n\nDonegan S, Williamson P, D'Alessandro U, et al.: Assessing key assumptions of network meta-analysis: a review of methods. Res Synth Methods. 2013; 4(4): 291–323. PubMed Abstract | Publisher Full Text\n\nVesterinen HM, Sena ES, Egan KJ, et al.: Meta-analysis of data from animal studies: a practical guide. J Neurosci Methods. 2014; 221: 92–102. PubMed Abstract | Publisher Full Text\n\nDias S, Welton NJ, Sutton AJ, et al.: NICE DSU Technical Support Document 2: A Generalised Linear Modelling Framework for Pairwise and Network Meta-Analysis of Randomised Controlled Trials. National Institute for Health and Clinical Excellence; 2011. Reference Source\n\nSpiegelhalter D, Thomas A, Best N, et al.: OpenBUGS User Manual. 2014. Reference Source\n\nSturtz S, Ligges U, Gelman A: R2WinBUGS: A package for running WinBUGS from R. J Stat Soft. 2005; 12(3): 1–16. Publisher Full Text\n\nSalanti G, Ades AE, Ioannidis JP: Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: an overview and tutorial. J Clin Epidemiol. 2011; 64(2): 163–171. PubMed Abstract | Publisher Full Text\n\nDias S, Welton NJ, Sutton AJ, et al.: Inconsistency in Networks of Evidence Based on Randomised Controlled Trials. NICE Decision Support Unit Technical Support Documents, London: National Institute for Health and Care Excellence (NICE); 2014. PubMed Abstract\n\nHirst JA, Howick J, Aronson JK, et al.: The need for randomization in animal trials: an overview of systematic reviews. PLoS One. 2014; 9(6): e98856. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSalanti G, Marinho V, Higgins JP: A case study of multiple-treatments meta-analysis demonstrates that covariates should be considered. J Clin Epidemiol. 2009; 62(8): 857–64. PubMed Abstract | Publisher Full Text\n\nDias S, Sutton AJ, Welton NJ, et al.: Evidence synthesis for decision making 3: heterogeneity--subgroups, meta-regression, bias, and bias-adjustment. Med Decis Making. 2013; 33(5): 618–40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLandis SC, Amara SG, Asadullah K, et al.: A call for transparent reporting to optimize the predictive value of preclinical research. Nature. 2012; 490(7419): 187–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFisher M, Feuerstein G, Howells DW, et al.: Update of the stroke therapy academic industry roundtable preclinical recommendations. Stroke. 2009; 40(6): 2244–50. PubMed Abstract | Publisher Full Text | Free Full Text"
}
|
[
{
"id": "42520",
"date": "28 Jan 2019",
"name": "Peter-Paul Zwetsloot",
"expertise": [
"Reviewer Expertise preclinical meta-analysis",
"translational 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\nLalu et al. provide us with their research protocol for a network meta-analysis of novel stroke therapeutics in preclinical models. The protocol is comprehensive and complete, including all necessary items like search strategy, screening and extensive data analysis. The search seems complete, using an animal filter and searching the Embase library, including Medline through Embase.\n\nThere are some minor concerns and suggestions for this paper and approach to potentially be more optimal;\n\nThe paper describes both fixed and random effects meta-analysis. As a network meta-analysis already usually is performed with random effects and your expected variation is considerable, I would skip the fixed effect meta-analysis as a whole. Even when using an NMD, is it appropriate to combine MRI and histology based outcomes, as these are known to not generate equivalent outcomes in preclinical models (MRI > histology, see Milidonis, Stroke 2015)? Will the NMD completely correct for this or is a sensitivity analysis needed (MRI vs histology)? If you don't think a sensitivity analysis is needed, please explain why not. Will the NMD also be used for the secondary outcome? This is not completely clear to me now. Consider to not dichotomise certain potential effect modifiers (for example the analysis time < or >30d as mentioned on page 4). Sometimes a continuous variable can give you more information in your analysis (for a potential linear effect for example). You can also choose to do both. Following on the previous comment; network meta-analysis is usually performed through a form of metaregression, making it possible to correct (potentially mutivariably) for a number of potential confounders/effect modifiers in the primary analysis itself. This is already mentioned on page 12 for the 'covariate-adjusted analyses'. Please provide a list upfront of the potential factors you want to correct for (in order of importance/usage) and provide an explanation on the number of factors you want to correct for (potentially based on the number of included studies?). To my knowledge this is different form the stated 'review these with our clinical and preclinical experts to establish the degree of homogeneity' and would add to your future primary analysis. This also means that the studies does not necessarily need to be homogenous for your primary analysis, as the metaregression will appoint a certain effect to these 'covariables' (and will correct for the covariable). Please provide a minimum number of comparisons for a certain intervention/comparison to be included in the network meta-analysis. Will there also be 2 or more, as with the traditional pairwise meta-analysis mentioned? If no minimum can be mentioned upfront, please explain why.\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": []
},
{
"id": "44938",
"date": "19 Mar 2019",
"name": "Hanna M. Vesterinen",
"expertise": [
"Reviewer Expertise Systematic review and meta-analysis of pre-clinical studies"
],
"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 fascinating research question and the methods set out in this protocol seem appropriate. I am satisfied that the authors have set out the protocol in accordance with the PRISMA-P checklist.\nI have outlined a few points below:\nI think the decision to reject studies which deliver therapies outside a 6 hour window warrants some additional background information. This may well be an appropriate decision; however, I don't have expertise in stroke and so it leads me to question if there is the possibility that the combination therapies could lead to greater efficacy/less harm outside this time period. In table 2, will the authors state what \"other\" is for species and type of model? This seems like useful information. Table 2 part b, should it state \"N Initially Reported\"? Is \"potential bias due to sample size calculation\" actually related to imprecision rather than risk of bias? Dichotomous cut offs might lose valuable information e.g. infarct <40% is within reasonable limits vs >40% is not. Is 39% vs 41% really that different?\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": []
}
] | 1
|
https://f1000research.com/articles/8-11
|
https://f1000research.com/articles/8-8/v1
|
03 Jan 19
|
{
"type": "Research Article",
"title": "The prevalence of dental caries among Egyptian children and adolescences and its association with age, socioeconomic status, dietary habits and other risk factors. A cross-sectional study",
"authors": [
"Marwa M.S. Abbass",
"Sara Ahmed Mahmoud",
"Sara El Moshy",
"Dina Rady",
"Nermeen AbuBakr",
"Israa Ahmed Radwan",
"Attera Ahmed",
"Ahmed Abdou",
"Ayoub Al-Jawaldeh",
"Sara Ahmed Mahmoud",
"Sara El Moshy",
"Dina Rady",
"Nermeen AbuBakr",
"Israa Ahmed Radwan",
"Attera Ahmed",
"Ahmed Abdou",
"Ayoub Al-Jawaldeh"
],
"abstract": "Background: Dental caries is a chronic, multifactorial disease, with limited data available for the Egyptian population. The aim of this study is to assess the prevalence of dental caries among Egyptian children and adolescents in correlation with age, gender, body mass index, socioeconomic status, parental education, biological risk factors and dietary habits. Methods: A total number of 369 Egyptian children and adolescents (age ranges from 3-18 years) were examined over the period from 15th November 2017 to 13th January 2018. Socio-demographic data, oral hygiene measures and dietary habits for children were recorded. Dental status was analyzed using decayed, missing and filled tooth index (dmft) for deciduous dentition and (DMFT) index for permanent dentition. For mixed dentition (deft) index was used, d (decayed tooth indicated for filling), e (decayed tooth indicated for extraction) and f (filled tooth). Results: 74% of the children had dental caries with mean dmft: 3.23±4.07; deft: 4.21±3.21; DMFT: 1.04±1.56. In primary dentition, dmft of the children was positively correlated with age, beans, candies, crackers, chocolates and inversely correlated with gender, socio-economic status (SES), parental education, brushing frequency of the parent, brushing frequency of the parent to the child teeth, brushing frequency of the child and consumption of eggs, fruits/vegetables, milk and milk products. In mixed dentition, deft was positively correlated with candies, crackers, citric juices, while negatively correlated with age, SES, parental education, brushing frequency of the parent to the child, brushing frequency of the child, fruits/vegetables. In permanent dentition, DMFT in children was positively correlated with age and chocolates while not correlated with any of the remaining risk factors. Conclusion: The present study clarifies the significant risk factors associated with dental caries amongst Egyptian children. This will help in planning strategies to prevent and treat such disease.",
"keywords": [
"Caries",
"Prevalence",
"Age",
"Socioeconomic",
"Dietary",
"Education",
"children",
"Adolescences"
],
"content": "Introduction\n\nDental caries is a major public health issue and it is the most widespread chronic disease1. Deciduous tooth decay was ranked as the 12th most prevalent condition, affecting 560 million children in the 2015 Global Burden of Disease Study2. Dental caries is a multifactorial disease, which can affect any age. It is highly related to and influenced by the patient’s dietary habits, sugar intake, salivary flow, salivary fluoride level and preventive behaviors. These factors, together with time, promote the microbial residence in the accumulated dental plaque to initiate dental caries1.\n\nIn children, dental caries pattern depends on the timing of tooth eruption as well as harmful dietary habit. Therefore, age is considered as an important factor that affects dental caries prevalence in children3. Although caries is common, parents are infrequently concerned about oral health measures and usually poor oral health is linked with low socio-economic status4. Dental caries could only be prevented through addressing and changing the underlying etiological factors5,6.\n\nIt has been estimated that about 60% to 90% of children at school age suffer from this chronic ailment7,8. This percentage varies greatly in different population, with the incidence of dental caries in developing countries, including the Middle East, being much higher than its incidence in developed countries9,10.\n\nDespite the high prevalence of dental caries in the Egyptian population, only a few epidemiological studies of dental caries among Egyptians have been published. Moreover, most of the available data, are grey literature which are not available on common search engines11–15. The most recently published epidemiological study on the oral health status in Egypt was held by WHO in collaboration with the Egyptian Ministry of Health in 201416. Moreover, most of the epidemiological studies focused on children17,18 and only one focused on adolescents19.\n\nThe prevalence of dental caries should be assessed continuously to plan and implement an efficient children oral health agenda and awareness programs for parents and school teachers in order to improve oral health. Therefore, this study was conducted to elucidate the prevalence of dental caries among Egyptian children and adolescence in correlation with age, socioeconomic level, dietary habits, oral hygiene measures and body mass index.\n\n\nSubjects and methods\n\nThis study was carried out according to the regulations of the Research Ethics Committee of the Faculty of Dentistry, Cairo University, Egypt [Approval: 171217]. Written informed consent was obtained from children’s parents or guardians to participate in the study. Verbal consent was taken from adolescents in addition to written consent from their parent/guardian. The subjects in this study were recruited from the outpatients' clinics of Faculty of Dentistry, Cairo University and from two private nurseries in Nasr City and Maadi from 15th November 2017 to 13th January 2018. The inclusion criteria were; Age: starting from 3 years to 12 years for the children group, from 13 years to 18 years for adolescence; Gender: Males & Females; Ethnicity: Egyptians. The exclusion criteria were; Previous history/current radiotherapy and/or chemotherapy; Subjects undergoing orthodontic therapy, which might preclude normal tooth brushing; Subjects who, in the opinion of the Investigator, may be non-compliant with study procedures.\n\n\n\nn' = sample size with finite population correction, N = children and adolescents population size (50,000,000), Z = Z statistic for a level of confidence which is conventional, Z value is (1.96). P = Expected prevalence (51.4%–70%) and d = Precision (5%, d = 0.05), the sample size for caries in children and adolescents was estimated 369. The prevalence of dental caries in Egypt was estimated to be 60% according to WHO, 201416, Hamila, 201418 reported ~70%, Mubarak et al. 201119 reported 51.4% and Abou El yazeed et al., 201117 reported 60.4%. The prevalence of dental caries in India was 61.4% in adolescent21 and in Australia was 51.6% among different refugee22.\n\nThe socio-demographic data collected by the authors included; name, age, gender, address, education (governmental, experimental or private), number of family members and their guardians' occupation and level of education in addition to oral hygiene measures for the children and their guardians as well as the dietary habits through a questionnaire. (Extended data90)\n\nChildren were classified according to their age into 5 groups; group I (3–5 years old); group II (5–7 years old); group III (7– 9); group IV (9– 12); group V (12– 18). Moreover, they were categorized based on their socioeconomic status, into low, moderate and high groups based on (level of education and its type, guardians' occupation and address)25.\n\nA Beurer scale (Ulm, Germany) was used to measure weights with individuals wearing clothing but no shoes. Standing heights were measured to the nearest 0.1cm using a stadiometer according to WHO, 199523. Body mass index (BMI) thresholds were calculated from the measured height and weight24,25. The obtained BMI values were plotted on age and gender–specific percentiles given by the Centers for Disease Control and Prevention26. Children were categorized into four groups based on their BMI percentiles; underweight (<5th percentile), normal group (≥5th - <85th percentile), overweight (≥85th - <95th percentile) and obese (≥95th percentile).\n\nPrior to beginning of examination, authors (M.M.S., S.A.M, S.E., D.R., N.A., I.A.R.) were trained and caliberated over 3 sessions over 3 days. Differences in observations were discussed among the examiners for reassessment and to reach a consensus19,27. Oral examination was carried out according to WHO criteria28 on a dental chair in artificial light by using a plain mouth mirror and a dental probe. All present teeth were taken into consideration during the clinical examination27.\n\nA tooth was considered carious when, any lesion in a pit or fissure or on a smooth tooth surface had a detectably softened floor, undermined enamel or softened wall, tooth surface containing temporary filling requiring further treatment and when a tooth surface containing a permanent restoration with an area of decay (either primary or secondary caries). Caries severity was measured for permanent teeth by DMFT index, which records the number of D (decayed tooth), M (missing tooth) and F (filled tooth). For primary teeth the dmft index was used; d (decayed teeth, m (missed teeth) and f (filled tooth). For mixed dentition deft index was used; d (decayed tooth indicated for filling), e (decayed tooth indicated for extraction) and f (filled tooth)29.\n\nThe statistical analysis was performed using R statistical package, version 3.3.1 (2016-06-21). For descriptive analyses, variables were described in terms of means, standard deviations (SD), medians and ranges. Shapiro-Wilk test for normality showed that all the studied parameters were not normally distributed. For comparative analysis, the non-parametric Kruskal-Wallis test was performed. Spearman’s Correlation Coefficient was calculated for correlation analysis. Results were considered significant at a P value of ≤ 0.05.\n\n\nResults\n\nThe mean dmft, deft, DMFT, age (years) and BMI (kg/m2) for the whole sample were (3.23±4.07, 4.21±3.21, 1.04 ±1.56, 7.2±3.53, 18.19±3.9) respectively. The number and percentage of children in different categories in each studied parameter as well as comparisons between them are presented in Table 1. Different categories within each parameter were statistically significant in correlation to each other (p-value <0.05) except for the variable “Reason for the children not brushing their teeth”.\n\n*Statistical significance at p-value ≤ 0.05.\n\nAge was positively correlated with dmft (Spearman’s rho=0.32, p-value<0.0001). The highest mean dmft was for children aged between 5 to 6 years old (5.62 ±4.27) with a statistically significant difference between medians (p-value=0.0006).\n\nThe correlation coefficient, rho, ranges from -1 to +1. Where 1= perfect positive correlation, 0=no correlation, -1 = perfect negative (inverse) correlation. *Statistical significance at p-value ≤ 0.05\n\nGender, SES and parental educational were inversely correlated with dmft (Spearman’s rho=-0.16, -0.64, -0.57, p-value=0.0483, <0.0001, <0.0001) respectively. Male children, children with low SES and children with low level of parental education had the highest mean dmft (3.88 ±4.4, 5.32 ±3.63, 6.47±4.44) with statistically significant difference (p-value= 0.0486, <0.0001, <0.0001) respectively.\n\nBMI was not correlated with dmft (Spearman’s rho=0.09, p-value=0.2742) and the highest mean dmft was for underweight children (4.18 ±4.27). The difference in medians was statistically insignificant (p-value >0.05).\n\ndmft of children increased with decreased brushing frequency of parents, brushing frequency of parents to their children and children’s own brushing frequency (Spearman’s rho= - 0.5, p-value<0.0001) and the highest mean dmft were with infrequent brushing (6.44 ±4.33, 5.97 ±4.19, 4.63 ±4.39) respectively. The difference in medians was statistically significant (p-value<0.0001).\n\nPositive correlations exist between dmft and the consumption of candies, crackers, chocolate and beans (rho=0.29, 0.3, 0.34, 0.18, p-value=0.0002, 0.0001, <0.0001, 0.0206) with highest mean dmft with frequency 1–6 times per day for candies, crackers, chocolate (4.21±4.45, 4.3±4.49, 5.02 ±5.02) respectively and 3–6 times per week for beans (4.6±4.5). Negative correlations occur between dmft and the consumption of eggs, fruits/vegetables, milk and milk products (rho= -0.21, -0.22, -0.31, -0.2, p-value=0.0089, 0.0046, <0.0001, 0.0093) with highest mean dmft with frequency ≤ 2 times/week (4.55±4.44, 5.71±4.55, 4.87 ±4.35, 5.11 ±4.74) respectively. No correlation was found between dmft and the consumption of bread, other carbohydrates, jams, junk food, soda, juices, citric juices and caffeinated drinks (p-value>0.05).\n\nThe highest mean deft was for children aged between 5 to 6 years old, children with low SES and children of patients with low level of parental education (5.51 ±3.28, 4.68 ±3.24, 4.62 ±3.25). The differences in medians were statistically significant for age and SES (p-value= 0.0013, 0.0277), while borderline insignificant for parental education (p-value=0.0542). Age, SES and parental education were inversely correlated with deft in children (Spearman’s rho= -0.42, -0.19, -0.16, p-value<0.0001, 0.0123, 0.0354).\n\nThe correlation coefficient, rho, ranges from -1 to +1. Where 1= perfect positive correlation, 0=no correlation, -1 = perfect negative (inverse) correlation.\n\n*Statistical significance at p-value ≤ 0.05.\n\nThe highest mean deft was for females (4.34 ±3.41) and overweight children (4.73 ±3.65). The difference in medians was statistically insignificant (p-value>0.05) and correlation coefficients revealed no correlation between deft and either gender or BMI (Spearman’s rho=0.01, 0.03, p-value=0.8758, 0.7132).\n\nThe highest mean deft was for children whose parents don’t brush their own teeth (4.65 ±2.75), or their children teeth (5.3 ±3.34), or children who don’t brush their own teeth (4.91 ±2.62). The difference in medians were statistically insignificant for parents own brushing frequency and brushing frequency to their children’s teeth (p-value>0.05), while it was statistically significant for children’s own brushing frequency (p-value= 0.0003). The correlation coefficient revealed no correlation between deft and parents own brushing frequency (rho= -0.14, p-value=0.0815), while inverse correlation with parents brushing frequency to their children’s teeth and children own brushing frequency (rho= -0.17, -0.31, p-value=0.0456, 0.0001).\n\nThe highest mean deft was for children who consume candies, crackers, junk food, juices, citric juices and caffeinated drinks 1–6 times per day (4.86±3.27, 4.77±3.23, 4.78± 2.89, 5.02±4.06, 6.32±3.96, 4.43±3.34) and children who consume bread, beans, chocolates and soda 3–6 times per week (10±2.83, 4.5±3.06, 4.58±3.53, 4.9±3.97), as well as children who consume other carbohydrates and jams less than or equal to two times a week (4.44 ±2.8, 4.38±3.37). The lowest mean deft was for children who consume eggs 3-6 times per week (3.77 ±3.61) and children who consume fruits/vegetables, milk and milk products 1–6 times per day (3.09 ±3.12, 3.49 ±3.33, 3.89±3.48) respectively.\n\nThe differences in medians for deft with fruits/vegetables, candies, crackers and citric juices consumption were statistically significant. There were positive correlations between candies, crackers, citric juices consumption and deft of children (rho=0.29, 0.34, 0.18, p-value=0.0001, <0.0001, 0.019), while there was inverse correlation between fruits/vegetables and deft (rho= -0.36, p-value <0.0001). No correlation was detected between deft of children and the consumption of bread, other carbohydrates, eggs, milk, milk products, beans, jams, junk food, chocolates, soda, juices and caffeinated drinks (p-value>0.05).\n\nThe highest mean DMFT was for adolescents aged between 12 to 17 years (1.68 ±1.92). The difference in medians was statistically significant (p-value= <0.0001). Age was positively correlated with DMFT (Spearman’s rho=0.36, p-value<0.0001).\n\nThe correlation coefficient, rho, ranges from -1 to +1. Where 1= perfect positive correlation, 0=no correlation, -1 = perfect negative (inverse) correlation. *Statistical significance at p-value ≤ 0.05.\n\nMales and children within the normal range for BMI had the highest mean DMFT (1.28 ±1.9, 1.15 ±1.52) respectively. The difference in medians was statistically insignificant for gender (p-value>0.05) and statistically significant (p-value=0.0353) for BMI. Correlation coefficient revealed no correlation between gender and BMI with DMFT (Spearman’s rho= -0.05, -0.04, p-value=0.4363, 0.5794).\n\nThe highest mean DMFT was for children with high SES and low parental education (1.29 ±1.73, 1.45 ±1.93). The differences in medians were statistically insignificant (p-value>0.05) and correlation coefficients were also insignificant (Spearman’s rho= -0.02, -0.07, p-value=0.7829, 0.2968).\n\nThe highest mean DMFT was for children whose parents brush their own teeth three times a day (1.6 ±3.58) and their child’s teeth once a day (1.19 ±1.86) as well as for children who don’t brush their teeth (1.19±1.47). The differences in medians for DMFT were statistically insignificant (p-value>0.05) and there was no correlation between DMFT and any of the biological risk factors.\n\nThe highest mean DMFT was for children who consume chocolates 1-6 times per day (1.3±1.56) and children who consume other carbohydrates, eggs, beans, jams, candies, crackers, soda, juices, citric juices and caffeinated drinks 3-6 times per week (1.32±1.95, 1.18±1.78, 1.57±2.47, 1.3±1.72, 1.38±2.16, 1.59±2.43, 1.25±1.75, 1.86±2.23, 1.4±2.23, 1.5±1.77) respectively as well as for children who consume bread, fruits/vegetables, milk, milk products and junk food less than or equal to two times a week (1.25±1.83, 1.23 ±1.68, 1.15±1.55, 1.19 ±1.79). The differences in medians with all the investigated dietary elements were statistically insignificant (p-value>0.05) and there was no correlation between DMFT and any of them except chocolate that had a positive correlation with DMFT (rho=0.16, p-value=0.0178) despite the difference in medians being borderline statistically insignificant (p-value=0.0584).\n\n\nDiscussion\n\nThe high prevalence of dental caries is not only influenced by the biological factors that interact with the causative microorganisms, but it is also associated with socioeconomic, educational conditions and dietary habits31. Dental caries is considered a major public health problem in underdeveloped or developing countries like Egypt. Therefore, investigating the risk factors associated with it has become a major concern to researchers who seek strategies for controlling or preventing the disease.31,32.\n\nThis study investigated the prevalence of dental caries in a wide age range of children because as individuals grow, their dietary needs and habits constantly change. The prevalence of dental caries among Egyptian children was higher in primary dentition (dmft and deft) when compared to permanent dentition (DMFT), this is similar to what has been reported in India33,34. Deciduous teeth have a higher susceptibility to dental caries due to the lower calcium content and structural differences35. Moreover, caries in the primary dentition could be associated with under nutrition during early childhood. Macro and micro tooth morphology, chemical composition and eruption pattern could be affected by nutrients like vitamin A, vitamin D, calcium and phosphorus36.\n\nIn the current study, there was a significant positive correlation between dmft and age, which is in agreement with previous studies conducted in Brazil and Colombia among children aged between 3 to 5 years37,38. On the other hand, age was inversely correlated with deft in Egyptian children as in mixed dentition period, the maintenance of oral hygiene is difficult due to shedding of primary teeth and pubertal changes.\n\nInvestigating the effect of gender on dental caries in the present study revealed that the mean dmft and DMFT of males were higher than those of females while the mean deft of females was slightly higher than that of males. Spearman’s test revealed an inverse correlation between gender and dmft, while no correlation with deft and DMFT. This is similar to a study carried out in Kerala on children aged 12–15 years old, where boys and girls were almost equally affected by caries39, while differs from another cross-sectional study carried on 10-11 years old Italians, where a significant difference was found between DFT of boys (3.20) and DFT of girls (1.96)40. It has been demonstrated that dental caries prevalence switches from male to females with age, where in the 5-year-old age group 47.4% of children with caries were male, while 41.1% were female. On the other hand, in the 12-year-old age group the percentage was inversed (24.1% female versus 20.6% male)41.\n\nUp to date there is a limited evidence clarifying the association between nutritional status and oral health. According to our results, there was no statistically significant correlation between BMI and any of the caries indices. This is inagreement with the findings from previous studies in Taiwan and in Sweden42,43. Oliveira et al. in Brazil44, concluded that underweight children were more likely to have caries which is in accordance with our findings where the highest mean dmft was recorded in underweight children (4.18±4.27). The recorded non significant correlation between DMFT and BMI is consistent with the findings of a previous study among adolescents aged 12 years in public and private schools in São Paulo State45, meanwhile, it is opposite to a study in a German elementary school46 which reported an increase in the DMF with increased BMI. However, in the present investigation a positive association between BMI and deft was recorded, but this was not at a statistically significant level (p=0.9166) which is similarly reported by Elangovan et al.47.\n\nParents play a significant role in the development of their children's oral hygiene habits48. Parental education level, which is directly associated with socioeconomic status49, greatly affects the child's oral health50. In the current study, there was an inverse significant correlation between parental educational level, socioeconomic status and dmft in children which is in accordance with previous studies that reported this correlation in early years44,51,52. Low socioeconomic status is usually accompanied by poor dietary habits and unhealthy lifestylesthat contribute to the development of dental caries53–57. Meanwhile, parents with high socioeconomic and education levels start taking care of their children’s dental health before their second year of life and help them brush their teeth, as reported in a German cross-sectional study58. As the child grows, the parental impact decreases and parents may totally lose their control on the child’s dietary habits and oral hygiene measures. This explains the non-significant correlation between DMFT and parental education, as well as socioeconomic level. Moreover, it has been reported in a study carried out on Indian adolescences that area of residence appears to be a significant determinant for an adolescent to be caries/decay free. Psychosocial and behavioral factors do not mediate the same association59. Since, parents' attitude is the principal social force influencing the child’s development in the early childhood years, therefore parental oral health beliefs and practices may be helpful in the prevention of oral health diseases such as caries60. This is concomitant with our findings where dmft and deft in Egyptian children were correlated to parental brushing to the child and with the child’s own brushing frequency. Children who received oral health education from their parents started to brush their teeth at an earlier age which revealed better dental health61.\n\nStudies that aim to establish the relationship between eating habits and the development of caries preferentially use frequency of food consumption questionnaires such as that employed in our study62,63. The relationship between the development of carious lesions and dietetic factors has been investigated since the 1940s. It has been suggested that the relationship between sugar intake and the development of carious lesions is currently different from that documented in the past decades, since dental health has improved greatly in the developed countries, with no parallel decrease in the consumption of sugar and cariogenic foods64.\n\nThe prevalence of dental caries in view of our results was significantly positively correlated with chocolate, candies and crackers consumption. A direct linkage between sugar intake and caries has been reported previously as cariogenic bacteria grow with the presence of fermentable carbohydrates65,66. Higher chocolate consumption led to increased caries indices which is consistent with the results from other studies67,68. Consumption of candies more than once per week, besides insufficient oral hygiene measures have been claimed to be risk factors for caries development in primary and permanent dentition69. Candies remain on the tooth surface for hours and don't have any nutritional value70.\n\nCitric juices were also found to be positively correlated with dental caries in primary teeth in Egyptian children. Hydroxyapatite crystals start to dissolve when the pH reaches 5.5 and enamel begins to be at risk of decalcification. Subsequently, acidic drinks have been reported to play a significant role in the pathogenesis of dental erosion71–75.\n\nThe inverse significant correlation between fruits/vegetables consumption and caries agrees with a study which recorded that dental caries prevalence was higher in non-vegetarians in comparison with vegetarians76. This correlation could be attributed to a lesser tendency for sweets between meals in vegetarians compared with non-vegetarians77. In addition Egyptian children who consumed milk more frequently had lower caries experience. Milk has low cariogenic potential and contains cariostatic factors against dental caries78. Studies showed that milk contains potential caries protective factors as calcium, phosphorus and casein79,80. The variability of milk consumption manner and other factors may result in a positive association between milk and dental caries occurrence80. This was proven when the frequency of milk consumption did not show a significant association with caries (DMFT).\n\nThe insignificant correlations recorded in the current work between all dental caries indices and bread, other carbohydrates, junk food, jam, molasses, honey and juice consumption could be referred to the cross-sectional design of the study. The cross-sectional study may not accurately reflect the true dietary habits of the children before the dental caries occurred since old dietary habits may be responsible for the current development of caries81.\n\nSoda consumption frequency recorded a non-significant correlation with dental caries. This disagrees with past studies which have reported positive correlation between soft drinks and dental caries82–84. Although sugars in soft drinks lead to drop in the pH of dental plaque and saliva, salivary components can neutralize the acids within 20-30 minutes raising the pH of plaque to its resting level85. Despite the fact that no correlation was found between caffeinated drinks consumption and caries, it was reported that polyphenols in coffee and tea can reduce the cariogenic potential of foods86. Coffee is active against Streptococcus Mutans, the organism causing dental caries. Roasted coffee also has anti-adhesive properties. In this manner, it prevents adhesion of Streptococcus Mutans to the teeth87.\n\nFinally it could be concluded that in primary dentition, the caries incidence in Egyptian children was positively correlated with candies and crackers while inversely correlated with SES, parental education, brushing frequency of the parent to the child, brushing frequency of the child to him/herself and fruits/vegetables consumption. In permanent dentition DMFT was only significantly positively correlated with age and chocolates.\n\nThe World Health Organization emphasizes the need public health solutions for prevention of dental caries. Therefore, the following recommendations based on the results of the current study should be added to the WHO policy measures88 to promote the reduction of dental caries prevalence in Egypt: 1- Candies and crackers should be prohibited for children before 12 years old; 2- Children should be encouraged to eat fruits and vegetables; 3- Awareness campaigns should be carried out to encourage the parents to brush their children’s teeth and to encourage the children to brush their own teeth.\n\nThe non-significant differences and lack of correlations between some caries indices and risk factors could be attributed to the small sample size, with a larger set of samples they may have reached statistical significance. In addition, a larger population from different governorates may have allowed broader diversity for better representation for Egyptian population.\n\n\nData availability\n\nUnderlying data is available from Figshare.\n\nFigshare: Dataset 1. Raw data for caries incidence in correlation to risk factors in Egyptian children. https://doi.org/10.6084/m9.figshare.7445843.v189\n\nLicense: CC0 1.0 Universal (CC0 1.0) Public Domain Dedication\n\nThe study questionnaire is available from Figshare.\n\nFigshare: Extended data. Questionnaire for children. https://doi.org/10.6084/m9.figshare.7392170.v390\n\nLicense: CC0 1.0 Universal (CC0 1.0) Public Domain Dedication",
"appendix": "Grant information\n\nThis study was supported by the World Health Organization.\n\n\nAcknowledgments\n\nWe would like to acknowledge the support and technical guidance of nutrition unit at World Health Organization office for Eastern Mediterranean region.\n\n\nReferences\n\nSelwitz RH, Ismail AI, Pitts NB: Dental caries. Lancet. 2007; 369(9555): 51–59. PubMed Abstract | Publisher Full Text\n\nGBD 2015 Disease and Injury Incidence and Prevalence Collaborators: Global, regional, and national incidence, prevalence, and years lived with disability for 310 diseases and injuries, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet. 2016; 388(10053): 1545–1602. PubMed Abstract | Publisher Full Text | Free Full Text\n\nQue G, Hou W: [Epidemiological investigation on deciduous dental caries among children aged 2 approximately 4 years of Kaifu district in Changsha city]. Zhong Nan Da Xue Xue Bao Yi Xue Ban. 2009; 34(1): 76–80. PubMed Abstract\n\nArmfield JM, Slade GD, Spencer AJ: Socioeconomic differences in children's dental health: the Child Dental Health Survey, Australia 2001. AIHW Dental Statistics and Research Unit. 2006. Reference Source\n\nPoureslami HR, Van Amerongen WE: Early Childhood Caries (ECC): an infectious transmissible oral disease. Indian J Pediatr. 2009; 76(2): 191–194. PubMed Abstract | Publisher Full Text\n\nAlbino J, Tiwari T: Preventing Childhood Caries: A Review of Recent Behavioral Research. J Dent Res. 2016; 95(1): 35–42. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPetersen PE, Kjoller M, Christensen LB, et al.: Changing dentate status of adults, use of dental health services, and achievement of national dental health goals in Denmark by the year 2000. J Public Health Dent. 2004; 64(3): 127–135. PubMed Abstract | Publisher Full Text\n\nWidström E, Eaton KA: Oral healthcare systems in the extended European union. Oral Health Prev Dent. 2004; 2(3): 155–194. PubMed Abstract\n\nPetersen PE: Challenges to improvement of oral health in the 21st century--the approach of the WHO Global Oral Health Programme. Int Dent J. 2004; 54(6 Suppl 1): 329–343. PubMed Abstract | Publisher Full Text\n\nLeake J, Jozzy S, Uswak G: Severe dental caries, impacts and determinants among children 2-6 years of age in Inuvik Region, Northwest Territories, Canada. J Can Dent Assoc. 2008; 74(6): 519. PubMed Abstract\n\nZahran M: Dental health of primary school children in Egypt: an incremental care program. Egypt Dent J. 1973; 19(2): 265–276. PubMed Abstract\n\nZahran M: The pattern of dental caries in Egyptian school children. Egypt Dent J. 1974; 20(4): 13–26. PubMed Abstract\n\nAbdel-Azim MM: Restorative dental status and restorative dental needs among a sample of Egyptian dental students. Egypt Dent J. 1979; 25(3): 269–278. PubMed Abstract\n\nel-Beheri S: Dental caries among Egyptian school children of high and low socio-economic level. Egypt Dent J. 1986; 32(3): 255–263. PubMed Abstract\n\nWissa AA, Zahran MA: Evaluation of governmental dental health services in rural health centers in Egypt. Community Dent Oral Epidemiol. 1988; 16(1): 16–18. PubMed Abstract | Publisher Full Text\n\nhttp://www.emro.who.int/egy/egypt-events/results-of-epidemiological-study-on-oral-health-status-released.html. 2014.\n\nAbou El-Yazeed M, Rashed M, El sayed M, et al.: Dental Caries Prevalence among a group of Egyptian Nurseries Children. Life Sci J. 2011; 8(1): 412–419. Reference Source\n\nHamila NAA: Early childhood caries and certain risk factors in a sample of children 1-3.5 years in Tanta. Dentistry. 2013; 4: 180. Publisher Full Text\n\nMobarak EH, Shabayek MM, Mulder J, et al.: Caries experience of Egyptian adolescents: does the atraumatic restorative treatment approach offer a solution? Med Princ Pract. 2011; 20(6): 545–9. PubMed Abstract | Publisher Full Text\n\nDaniel WW: Biostatistics: A Foundation for Analysis in the Health Sciences. 7th edition. New York: John Wiley & Sons. 1999. Reference Source\n\nVeerasamy A, Kirk R, Gage J: Epidemiology of dental caries among adolescents in Tamil Nadu, India. Int Dent J. 2016; 66(3): 169–77. PubMed Abstract | Publisher Full Text\n\nQuach A, Laemmle-Ruff IL, Polizzi T, et al.: Gaps in smiles and services: a cross-sectional study of dental caries in refugee-background children. BMC Oral Health. 2015; 15(1): 10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorld Health Organization: Expert Committee on Physical Status: the use and interpretation of anthropometry physical status. WHO Techniques Report Series, v 854. Geneva: World Health Organization. 1995. Reference Source\n\nMartins RJ, Moimaz SA, Silva MR, et al.: Body mass index, dental caries and sugar intake in 2-5 year-old preschoolers. Braz J Oral Sci. 2014; 13(3): 209–212. Publisher Full Text\n\nEdalat A, Abbaszadeh M, Eesvandi M, et al.: The Relationship of Severe Early Childhood Caries and Body Mass Index in a Group of 3- to 6-year-old Children in Shiraz. J Dent (Shiraz). 2014; 15(2): 68–73. PubMed Abstract | Free Full Text\n\nKuczmarski RJ, Ogden CL, Grummer-Strawn LM, et al.: CDC growth charts: United States. Adv Data. 2000; 8(314): 1–27. PubMed Abstract\n\nJokić NI, Bakarčić D, Janković S, et al.: Dental caries experience in croatian school children in Primorsko-Goranska county. Cent Eur J Public Health. 2013; 21(1): 39–42. PubMed Abstract | Publisher Full Text\n\nWorld Health Organization: Oral Health Surveys: Basic Methods. World Health Organization, Geneva, Switzerland. 1997. Reference Source\n\nWorld Health Organization: Oral health surveys: Basic methods. 5th ed., Geneva, Switzerland, World Health Organization. 2013. Reference Source\n\nEl-Gilany A, El-Wehady A, El-Wasify M: Updating and validation of the socioeconomic status scale for health research in Egypt. East Mediterr Health J. 2012; 18(9): 962–968. PubMed Abstract | Publisher Full Text\n\nMcGrady MG, Ellwood RP, Goodwin M, et al.: Adolescents' perceptions of the aesthetic impact of dental fluorosis vs. other dental conditions in areas with and without water fluoridation. BMC Oral Health. 2012; 12(1): 4. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCypriano S, Hoffmann RH, de Sousa Mda L: Dental caries experience in 12-year-old schoolchildren in southeastern Brazil. J Appl Oral Sci. 2008; 16(4): 286–292. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGoyal A, Gauba K, Chawla HS, et al.: Epidemiology of dental caries in Chandigarh school children and trends over the last 25 years. J Indian Soc Pedod Prev Dent. 2007; 25(3): 115–118. PubMed Abstract | Publisher Full Text\n\nSudha P, Bhasin S, Anegundi RT: Prevalence of dental caries among 5-13-year-old children of Mangalore city. J Indian Soc Pedod Prev Dent. 2005; 23(2): 74–79. PubMed Abstract | Publisher Full Text\n\nJain A, Jain V, Suri SM, et al.: Prevalence of dental caries in male children from 3 to 14 years of age of Bundelkhand region, India. Int J Community Med Public Health. 2016; 3(4): 787–790. Publisher Full Text\n\nPsoter WJ, Reid BC, Katz RV: Malnutrition and dental caries: a review of the literature. Caries Res. 2005; 39(6): 441–447. PubMed Abstract | Publisher Full Text | Free Full Text\n\nXavier A, Bastos RDS, Arakawa AM, et al.: Correlation between dental caries and nutritional status: preschool children in a Brazilian municipality. Revista de Odontologia da UNESP. 2013; 42(5): 378–383. Publisher Full Text\n\nOjeda JC, Llanos LS: Caries prevalence of preschool age children in community homes of the Cauca Valle and related social factors. Revista Odontológica Mexicana. 2017; 21(4): e221–e226. Publisher Full Text\n\nJose A, Joseph MR: Prevalence of dental health problems among school going children in rural Kerala. J Indian Soc Pedod Prev Dent. 2003; 21(4): 147–51. PubMed Abstract\n\nMigale D, Barbato E, Bossù M, et al.: Oral health and malocclusion in 10-to-11 years-old children in southern Italy. Eur J Paediatr Dent. 2009; 10(1): 13–18. PubMed Abstract\n\nSaravanan S, Anuradha KP, Bhaskar DJ: Prevalence of dental caries and treatment needs among school going children of Pondicherry, India. J Indian Soc Pedod Prev Dent. 2003; 21(1): 1–12. PubMed Abstract\n\nAlm A, Isaksson H, Fahraeus C, et al.: BMI status in Swedish children and young adults in relation to caries prevalence. Swed Dent J. 2011; 35(1): 1–8. PubMed Abstract\n\nChen W, Chen P, Chen SC, et al.: Lack of association between obesity and dental caries in three-year-old children. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi. 1998; 39(2): 109–111. PubMed Abstract\n\nOliveira LB, Sheiham A, Bönecker M: Exploring the association of dental caries with social factors and nutritional status in Brazilian preschool children. Eur J Oral Sci. 2008; 116(1): 37–43. PubMed Abstract | Publisher Full Text\n\nSales-Peres SH, Goya S, Sant’Anna RM, et al.: [Prevalence of overweight and obesity, and associated factors in adolescents, at the central west area of the state São Paulo (SP, Brazil)]. Cien Saude Colet. 2010; 15 Suppl 2: 3175–3184. PubMed Abstract | Publisher Full Text\n\nWillerhausen B, Blettner M, Kasaj A, et al.: Association between body mass index and dental health in 1,290 children of elementary schools in a German city. Clin Oral Investig. 2007; 11(3): 195–200. PubMed Abstract | Publisher Full Text\n\nElangovan A, Mungara J, Joseph E: Exploring the relation between body mass index, diet, and dental caries among 6-12-year-old children. J Indian Soc Pedod Prev Dent. 2012; 30(4): 293–300. PubMed Abstract | Publisher Full Text\n\nSaldūnaitė K, Bendoraitienė EA, Slabšinskienė E, et al.: The role of parental education and socioeconomic status in dental caries prevention among Lithuanian children. Medicina (Kaunas). 2014; 50(3): 156–161. PubMed Abstract | Publisher Full Text\n\nPizzo G, Piscopo MR, Matranga D, et al.: Prevalence and socio-behavioral determinants of dental caries in Sicilian schoolchildren. Med Sci Monit. 2010; 16(10): PH83–9. PubMed Abstract\n\nKoposova N, Widström E, Eisemann M, et al.: Oral health and quality of life in Norwegian and Russian school children: A pilot study. Stomatologia. 2010; 12(1): 10–16. PubMed Abstract\n\nKato H, Tanaka K, Shimizu K, et al.: Parental occupations, educational levels, and income and prevalence of dental caries in 3-year-old Japanese children. Environ Health Prev Med. 2017; 22(1): 80. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHooley M, Skouteris H, Boganin C, et al.: Parental influence and the development of dental caries in children aged 0-6 years: a systematic review of the literature. J Dent. 2012; 40(11): 873–85. PubMed Abstract | Publisher Full Text\n\nMarshall TA, Eichenberger-Glilmore JM, Broffitt BA, et al.: Dental caries and childhood obesity: roles of diet and socioeconomic status. Community Dent Oral Epidemiol. 2007; 35(6): 449–458. PubMed Abstract | Publisher Full Text\n\nJürgensen N, Petersen PE: Oral health and the impact of socio-behavioural factors in a cross sectional survey of 12-year old school children in Laos. BMC Oral Health. 2009; 9(1): 29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBruno-Ambrosius K, Swanholm G, Twetman S: Eating habits, smoking and toothbrushing in relation to dental caries: a 3-year study in Swedish female teenagers. Int J Paediatr Dent. 2005; 15(3): 190–6. PubMed Abstract | Publisher Full Text\n\nAyele FA, Taye BW, Ayele TA, et al.: Predictors of dental caries among children 7-14 years old in Northwest Ethiopia: a community based cross-sectional study. BMC Oral Health. 2013; 13: 7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMafuvadze BT, Mahachi L, Mafuvadze B: Dental caries and oral health practice among 12 year old school children from low socio-economic status background in Zimbabwe. Pan Afr Med J. 2013; 14: 164. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPieper K, Dressler S, Heinzel-Gutenbrunner M, et al.: The influence of social status on pre-school children’s eating habits, caries experience and caries prevention behavior. Int J Public Health. 2012; 57(1): 207–15. PubMed Abstract | Publisher Full Text\n\nMathur MR, Tsakos G, Millett C, et al.: Socioeconomic inequalities in dental caries and their determinants in adolescents in New Delhi, India. BMJ Open. 2014; 4(12): e006391. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHooley M, Skouteris H, Boganin C, et al.: Parental influence and the development of dental caries in children aged 0-6 years: a systematic review of the literature. J Dent. 2012; 40(11): 873–885. PubMed Abstract | Publisher Full Text\n\nBoka V, Trikaliotis A, Kotsanos N, et al.: Dental caries and oral health-related factors in a sample of Greek preschool children. Eur Arch Paediatric Dent. 2013; 14(6): 363–368. PubMed Abstract | Publisher Full Text\n\nLlena C, Forner L: Dietary habits in a child population in relation to caries experience. Caries Res. 2008; 42(5): 387–393. PubMed Abstract | Publisher Full Text\n\nLlena Puy C, Forner L: Dietary habits of a school population and implications for oral health. Minerva Stomatol. 2010; 59(4): 173–180. PubMed Abstract\n\nSerra-Majem L, García Closas R, Ramón JM, et al.: Dietary habits and dental caries in a population of Spanish schoolchildren with low levels of caries experience. Caries Res. 1993; 27(6): 488–494. PubMed Abstract | Publisher Full Text\n\nÇolak H, Dülgergil ÇT, Dalli M, et al.: Early childhood caries update: A review of causes, diagnoses, and treatments. J Nat Sci Biol Med. 2013; 4(1): 29–38. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSheiham AJ, James WP: A reappraisal of the quantitative relationship between sugar intake and dental caries: the need for new criteria for developing goals for sugar intake. BMC Public Health. 2014; 14: 863. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDoichinova L, Bakardjiev P, Peneva M: Assessment of food habits in children aged 6-12 years and the risk of caries. Biotechnol Biotechnol Equip. 2015; 29(1): 200–204. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMafuvadze BT, Mahachi L, Mafuvadze B: Dental caries and oral health practice among 12 year old school children from low socio-economic status background in Zimbabwe. Pan Afr Med J. 2013; 14: 164. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMobley C, Marshall TA, Milgrom P, et al.: The contribution of dietary factors to dental caries and disparities in caries. Acad Pediatr. 2009; 9(6): 410–414. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAlmasi A, Rahimiforoushani A, Eshraghian MR, et al.: Effect of Nutritional Habits on Dental Caries in Permanent Dentition among Schoolchildren Aged 10-12 Years: A Zero-Inflated Generalized Poisson Regression Model Approach. Iran J Public Health. 2016; 45(3): 353–61. PubMed Abstract | Free Full Text\n\nMathew T, Casamassimo PS, Hayes JR: Relationship between sports drinks and dental erosion in 304 university athletes in Columbus, Ohio, USA. Caries Res. 2002; 36(4): 281–287. PubMed Abstract | Publisher Full Text\n\nAl-Majed I, Maguire A, Murray JJ: Risk factors for dental erosion in 5-6 year old and 12-14 year old boys in Saudi Arabia. Community Dent Oral Epidemiol. 2002; 30(1): 38–46. PubMed Abstract | Publisher Full Text\n\nRees J, Loyn T, McAndrew R: The acidic and erosive potential of five sports drinks. Eur J Prosthodont Restor Dent. 2005; 13(4): 186–190. PubMed Abstract\n\nTahmassebi JF, Duggal MS, Malik-Kotru G, et al.: Soft drinks and dental health: a review of the current literature. J Dent. 2006; 34(1): 2–11. PubMed Abstract | Publisher Full Text\n\nSoares PV, Tolentino AB, Machado AC, et al.: Sports dentistry: a perspective for the future. Rev Bras Educ Fís Esporte. 2014; 28(2): 351–358. Publisher Full Text\n\nShah N, Sundaram KR: Impact of socio-demographic variables, oral hygiene practices, oral habits and diet on dental caries experience of Indian elderly: a community‐based study. Gerodontology. 2004; 21(1): 43–50. PubMed Abstract | Publisher Full Text\n\nSherfudhin H, Abdullah A, Shaik H, et al.: Some aspects of dental health in young adult Indian vegetarians. A pilot study. Acta Odontol Scand. 1996; 54(1): 44–48. PubMed Abstract | Publisher Full Text\n\nJohansson I: Milk and dairy products: Possible effects on dental health. Food Nutr Res. 2002; 46(3): 119–122. Publisher Full Text\n\nMoynihan P, Petersen PE: Diet, nutrition and the prevention of dental diseases. Public Health Nutr. 2004; 7(1A): 201–226. PubMed Abstract | Publisher Full Text\n\nLim S, Sohn W, Burt BA, et al.: Cariogenicity of soft drinks, milk and fruit juice in low-income african-american children: a longitudinal study. J Am Dent Assoc. 2008; 139(7): 959–967, quiz 995. PubMed Abstract | Publisher Full Text\n\nZahara AM, Ili MN, Yahya NA: Dietary habits and dental caries occurrence among young children: Does the relationship still exist? Malaysian J Med Health Sci. 2013; 9(1): 9–20. Reference Source\n\nLevy SM, Warren JJ, Broffitt B, et al.: Fluoride, beverages and dental caries in the primary dentition. Caries Res. 2003; 37(3): 157–165. PubMed Abstract | Publisher Full Text\n\nSohn W, Burt BA, Sowers MR: Carbonated soft drinks and dental caries in the primary dentition. J Dent Res. 2006; 85(3): 262–266. PubMed Abstract | Publisher Full Text\n\nStevens A, Hamel C, Singh K, et al.: Do sugar-sweetened beverages cause adverse health outcomes in children? A systematic review protocol. Syst Rev. 2014; 3(1): 96. PubMed Abstract | Publisher Full Text | Free Full Text\n\nIsmail AI, Burt BA, Eklund SA: The cariogenicity of soft drinks in the United States. J Am Dent Assoc. 1984; 109(2): 241–245. PubMed Abstract | Publisher Full Text\n\nOoshima T, Minami T, Aono W, et al.: Oolong tea polyphenols inhibit experimental dental caries in SPF rats infected with mutans streptococci. Caries Res. 1993; 27(2): 124–129. PubMed Abstract | Publisher Full Text\n\nDaglia M, Racchi M, Papetti A, et al.: In vitro and ex vivo antihydroxyl radical activity of green and roasted coffee. J Agric Food Chem. 2004; 52(6): 1700–4. PubMed Abstract | Publisher Full Text\n\nWorld Health Organization (WHO): Sugars intake for adults and children. Geneva: WHO, 2015; accessed 17 September 2017. Reference Source\n\nAbbass M: Raw data for caries incidence in correlation to risk factors in Egyptian children. 2018. http://www.doi.org/10.6084/m9.figshare.7445843.v1\n\nAbbass M: questionnaire for children.pdf. 2018. http://www.doi.org/10.6084/m9.figshare.7392170.v3"
}
|
[
{
"id": "44056",
"date": "08 Feb 2019",
"name": "Mauro Henrique Nogueira Guimarães Abreu",
"expertise": [
"Reviewer Expertise Oral Epidemiology and Health Services 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\nThis study focused on dental caries prevalence and some associated factors among Egyptian children and adolescents. The study has a good rationale and reasonable theoretical framework. There are some issues and problems that should be corrected by authors:\nThe term “correlation” should be modified to “association” in all sections of the manuscript. The term “correlation” should be maintained only where Spearman’s Correlation Coefficient was correctly calculated. The authors should present a good justification for not using dmft index. “Subjects who, in the opinion of the Investigator, may be non-compliant with study procedures.” – This criterion is quite subjective. In the sample size calculation, the authors should clearly point out that they use the prevalence of “60%”. The most serious issue is how the individuals were selected (simple random sample? Cluster random sample? Another sampling strategy?) There are some English language errors that must be corrected. I strongly recommend that a professional native English be contacted. Statistical analysis is worrisome. Why authors use the Shapiro-Wilk test? Why did they not use the Kolmogorov-Smirnov test? The indication of correlation analysis for comparing nominal and quantitative (Eg. Gender and dmft) variables is also incorrect. The influence of age in dental caries is not a new finding in the literature. Therefore, authors should stratify the age groups with similar prevalence and severity of dental caries in order to perform a better and rigorous analysis. The authors should perform multiple regression models for all dependent variables. Without these previously corrections, it is not possible to evaluate the results, discussion, and conclusions.\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? 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? Yes\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
},
{
"id": "44369",
"date": "20 Feb 2019",
"name": "Rania Mossad Hassan",
"expertise": [
"Reviewer Expertise Dental anatomy and oral 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 study investigated the dental caries prevalence among Egyptian children and adolescents in association with different risk factors.\nAt the beginning, I would like to congratulate the authors for attempting to undertake this work, due to data shortage or even lack of data regarding this subject. I found it very interesting and of valuable additional knowledge.\nThe manuscript is well-written and well-structured. The relationships between different caries indices, SES and parental education haven't been investigated before among Egyptian children. Moreover, this is the first work that deals with the correlation between cariogenic and non cariogenic dietary elements with the caries indices among Egyptian children.\nThe discussion is well written covering all the study results. The data from this work could be used to make regression analysis to identify the causative risk factors.\nHowever, the paper is clear substantially easy to read but still there are suggested minor comments the author could deal with, or discuss for additional impact.\nThe results are given in full details, the great amount of data make it difficult for the reader to understand some relations. The age ranges differed, as they were 3- 12 and 13-18 years when mentioned in subjects and methods, meanwhile in data collection, the age groups were classified into 3-5, 5-7, 7-9, 9-12 and 12-18 years. On the other hand, in the tables, the age groups were 3-4, 5-6, 7-8, 9-11,12-17 years. So, they are to be corrected.\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": "44367",
"date": "27 Feb 2019",
"name": "Ali Arabi",
"expertise": [
"Reviewer Expertise Gastroentlogy and nutrtion"
],
"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 strongly supports the research dental health and in relation to oral health and dietary habits. But the tittle is too long Using different risk factors with no good outcomes Oral health is a key factors not well addressed in introduction especially guidelines for small children and adolescents Very detailed about dental caries but few about oral health and dietary habits Research questions not all will addressed in methodology The methodology is adequate but Lacks the methodology of dietary habits is it. Diet is a major lifestyle-related risk factor of various chronic diseases. Dietary intake can be assessed by subjective report and objective observation. Subjective assessment is possible using open-ended surveys such as dietary recalls or records, or using closed-ended surveys including food frequency questionnaires. Each method has inherent strengths and limitation not clear which method used need clarifications\n\nHowever, there are a few issues that need to be addressed in the results\n\nTable 3, 4 and 5 are confusing…. So data is reported in mean (SD) and median…with range or IQ range it’s not clear in the table caption…. Please clarify that.. A typo error in the dietary habits. please correct.\n\nAlso reporting the Mean (SD) is not suitable for all the data reported except for age.. so reporting that value add more to the confusion and increases the table sizes with no value……reporting the frequency is a must. K-W test when significant please report the pairwise comparison\n\nReferences are up-to-date\n\nQuestion to author:\n\nCould the title of paper changed to:\nThe prevalence of dental caries among Egyptian young children and adolescences and its association with oral health and dietary habits because age and socioeconomic factors and other factors has no outcomes in the study paper\n\nIn the recommendations, should be age related young children should different from from adolescence because of different dietary habits and oral health and age 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? Partly\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/8-8
|
https://f1000research.com/articles/8-7/v1
|
03 Jan 19
|
{
"type": "Research Article",
"title": "Robust and efficient identification of biomarkers from RNA-Seq data using median control chart",
"authors": [
"Md Shahjaman",
"Habiba Akter",
"Md. Mamunur Rashid",
"Md. Ibnul Asifuzzaman",
"Md. Bipul Hossen",
"Md. Rezanur Rahman",
"Md. Mamunur Rashid",
"Md. Ibnul Asifuzzaman",
"Md. Bipul Hossen",
"Md. Rezanur Rahman"
],
"abstract": "Background: One of the main goals of RNA-seq data analysis is identification of biomarkers that are differentially expressed (DE) across two or more experimental conditions. RNA-seq uses next generation sequencing technology and it has many advantages over microarrays. Numerous statistical methods have already been developed for identification the biomarkers from RNA-seq data. Most of these methods were based on either Poisson distribution or negative binomial distribution. However, efficient biomarker identification from discrete RNA-seq data is hampered by existing methods when the datasets contain outliers or extreme observations. Specially, the performance of these methods becomes more severe when the data come from a small number of samples in the presence of outliers. Therefore, in this study, an attempt is made to propose an outlier detection and modification approach for RNA-seq data to overcome the aforesaid problems of traditional methods. We make our proposed method facilitate in RNA-seq data by transforming the read count data into continuous data. Methods: We use median control chart to detect and modify the outlying observation in a log-transformed RNA-seq dataset. To investigate the performance of the proposed method in absence and presence of outliers, we employ the five popular biomarker selection methods (edgeR, edgeR_robust, DEseq, DEseq2 and limma) both in simulated and real datasets. Results: The simulation results strongly suggest that the performance of the proposed method improved in the presence of outliers. The proposed method also detected an additional 18 outlying DE genes from a real mouse RNA-seq dataset that were not detected by traditional methods. Using the KEGG pathway and gene ontology analysis results we reveal that these genes may be biomarkers, which require validation in a wet lab. Conclusions: Our proposal is to apply the proposed method for biomarker identification from other RNA-seq data.",
"keywords": [
"RNA-Seq data",
"Logarithmic transformation",
"Biomarker",
"Outliers and Robustness"
],
"content": "Introduction\n\nOne of the major objectives of researchers is to identify biomarkers from RNA-Seq data that are differentially expressed (DE) between two or more experimental conditions. Microarrays have been extensively used during the past few decades to perform this task. But after reducing the cost of sequencing, biomarker identification using RNA-seq data has emerged as an alternative choice to microarrays1,2. RNA-seq uses next generation sequencing technology to produce a vast amount of data. Curse of dimensionality is a common problem for analyzing RNA-Seq data, which means a \"large p, small n\" problem. Hence, dimension reduction of the data matrix is a primary objective for further downstream analysis. Identification of biomarkers is one of the dimensionality reduction approaches.\n\nRNA-seq count data inherently follow a Poisson or negative binomial (NB) distribution rather than normal distribution like microarray data. Numerous statistical methods have been developed to identify biomarkers from RNA-seq count data. The earliest method is DEGseq, which is based on Poisson distribution. This method suffers from overdispersion and therefore Poisson distribution-based methods are not suitable for RNA-seq data3–6. To overcome this problem, NB distribution-based methods have been proposed. Some NB based methods are: baySeq, DESeq, DESeq2, EBSeq, edgeR, edgeR (robust) and NBPSeq7–11. However, most of the methods cannot estimate properly the gene-wise dispersion parameters and they also suffer from small sample sizes12. DESeq, DESeq2, edgeR, edgeR (robust) and NBPSeq incorporate information of all genes in their algorithms.\n\nDespite the popularity of these statistical methods for identification of biomarker genes, they are sensitive to outliers and often produce lower accuracies in the presence of outliers. Outliers may arise in RNA-seq count data because there are several data generating stages from biological harvesting of RNA samples to counting of sequence read map data13. To mitigate this issue many algorithms use transformation methods. There are several transformation methods for RNA-seq data: logarithmic transformation14, variance-stabilizing transformation (vst)6, TMM transformation15, regularized logarithm8 and variance modeling at the observation level (voom)16. These methods only reduce the low level outliers into reasonable spaces during parameter estimations; however they fail to reduce the influence of high level outliers with small sample sizes in the data matrix.\n\nConsequently, most biomarker selection methods that use the aforesaid transformations, are sensitive to outliers or extreme values with small-sample sizes. Therefore, in this study, an attempt is made to propose an outlier detection and modification approach for RNA-seq data to improve the performance of the popular biomarker selection methods in the presence of outliers. To make our proposed method facilitate in RNA-seq data we transform the read count data into continuous data using regularize logarithmic transformation.\n\nThe article is organized as follows: Methods briefly describes the logarithmic transformation and formulation the proposed outlier detection and modification approach. In Results and Conclusions a broad simulation study and a real data study have been carried out.\n\n\nMethods\n\nLet ygik be the number of reads simulated from gth gene of kth replicates in the ith condition (g = 1, 2, . . ., G; i = 1,2; k =1, 2, . . . , ni); where ni is the number of replicated in condition i. G is the total number of gene. Then the negative bionomial-distribution is as follows:\n\n\n\nIn this negative binomial parameterization, E(ygik) = μgi and Var ( (ygik)=μgi+μgi2rgi; where, μgi is the mean of gth gene in ith condition and 1rgi is the dispersion parameter. Now we want to test the following null hypothesis:\n\n\n\nA gene will be declared as DE if H0 is rejected, otherwise it is equally expressed (EE).\n\nLog-transformation is very useful in RNA-Seq data. The log-transformed data usually follow the normal distribution, which depends on the degree of skewness before transformation. As the RNA-Seq count data can be equal to zero, so we shift them by one before transforming them:\n\n\n\nLog-transformed values have less extreme values (or outliers) than the untransformed data.\n\nHowever, the log-transformed values reduces the influence of low level outlying observations; however this transformation fails to reduce the influence of high level outlying observations. Therefore, we propose the following outlier detection and modification rule.\n\nSince the median and median absolute deviation is the robust measure of location and scale parameter, respectively, therefore, we used the median control chart as a measure of outlier detection. The proposed procedure is as follows:\n\n1. We declare a gene as an outlying gene if it doesn’t fall into the interval [LCL, UCL]. Where LCL and UCL are the lower and upper control limit for median and they are defined by LCL= MEDg,(i)-3×MADg,(i) and UCL=MEDg,(i)+3×MADg,(i)]. Here, MEDg,(i)=median\n\n(xgik; g =1, 2,…, G; k =1, 2,…, ni ; i=1,2) is the median of gth gene in ith condition, MADg,(i) = mediank=1,2,..,ni(|xgik − MEDg, (i)|) is the median absolute deviation.\n\n2. Check the existence of outliers for each gene from each of the conditions (i=1,2), separately using step 1. If outliers exist, replace them by their respective group medians (MEDg,(i)).\n\n3. Apply the anti-log transformation to obtain modified RNA-seq (MRS) count datasets.\n\n4. Apply traditional statistical methods in MRS data to identify biomarker genes using the p-values adjusted by Benjamini-Hochberg method.\n\n5. Obtain the functional annotations and KEGG pathways for detected biomarker genes.\n\nIn order to evaluate the performance of different biomarkers selection methods we considered the area under the receiver operating characteristic curve (ROC) curve. The ROC is created by plotting the true positive rate (TPR) against the false positive rate (FPR) for different cut-off points of a parameter. For a particular threshold each point on the ROC curve produces a TPR/FPR pair. The area under the ROC curve (AUC) is a performance measure which helps us to select an optimum method that can distinguish between two gene groups such as DE or EE well.\n\nTo investigate the performance of the proposed method in comparison with five popular methods as mentioned above, for both small-and-large-sample cases with 2 groups/conditions, we considered 100 datasets for both cases with sample sizes of n1=n2= 3 and 15, respectively. Each dataset for each case represented gene expression profiles for 1000 genes, each with n=(n1+n2) samples, where the read counts of each gene was generated using the negative-binomial distribution and this type of simulation study was also used in 11. The number of DE genes were set to 40 for each of the 100 datasets. We divided these 40 DE genes into two groups: 20, up-regulated DE genes and 20, down-regulated DE genes. To show the effect of outliers (extreme high counts) on the methods, we randomly selected 10% and 30% genes and for each of these genes, we selected a single sample randomly and multiplied the observed count of this sample with randomly selected factor between 5 and 10. This process was applied for each of the 100 datasets. We computed average values of different performance measures such as true positive rate (TPR), false positive rate (FPR) and AUC based on 40 estimated DE genes by five methods (edgeR, edgeR_robust, DESeq, DESeq2, and limma (voom)) for each of 100 original datasets and proposed MRS datasets.\n\nWe also considered a real RNA-seq mouse dataset17 to demonstrate the performance of the methods. This dataset consists of 36535 genes with 21 samples. This dataset was downloaded from ReCount website http://bowtie-bio.sourceforge.net/recount. It can also be downloaded from the GEO series accession number GSE26024. Among 21 samples, RNA-seq count expression collected from 10 C57BL/6J (B6) and 11 DBA/2J (D2) inbred mouse strains.\n\nTo demonstrate the performance of the proposed method, a comparison with five popular methods (edgeR, edgeR_robust, DESeq, DESeq2 and limma (voom)) was performed. We used both simulated and real RNA-seq count datasets. We used three R packages of other methods: edgeR, DESeq and limma. The performance measure area under the receiver operating characteristics curve (AUC) was computed for each of the methods using R package ROCR. All R packages are available in the comprehensive R archive network (cran) or Bioconductor.\n\n\nResults\n\nTable 1 summarizes the average AUCs estimated by eight methods based on 100 simulated datasets using 4% DE genes in absence and presence of single outlier in each of 10% and 30% genes for both small-sample cases (n1=n2=3) and large-sample cases (n1=n2=15), respectively. In Table 1, the results without and within the brackets (.) indicate the estimated AUCs by the five methods using the original RNA-seq datasets and proposed MRS datasets. From Table 1 we observed that in absence of outliers, five methods (edgeR, edgeR_robust, DESeq, DESeq2 and limma (voom)) produced almost similar results using the original RNA-seq datasets and proposed MRS datasets, for both small-and-large-sample cases. However, in the presence of outliers, the performance of these methods has significantly increased using the proposed MRS datasets for both cases. For example, in the presence of 10% outliers, edgeR and DESeq produce AUCs 0.829 and 0.818, respectively for small-sample case. Whereas for the same condition these two methods produce AUCs 0.842 and 0.838, respectively using our proposed MRS datasets. Figure 1a and b and Figure 1c and d show the boxplot of AUCs based on 100 simulated datasets by each of the methods in absence and presence of outliers for small-and large-sample cases, respectively. The left and right-side panels in this figure indicate the boxplot of estimated AUCs using original RNA-seq datasets and proposed MRS datasets. Similar results were found from these boxplots, as in Table 1.\n\n(a–b) for small-sample case (n1=n2= 3) and (c–d) for large-sample case (n1=n2= 15).\n\nAfter filtering we retain 11,474 genes. To investigate the performance of the proposed method, we employ three methods (edgeR, DEseq and limma) for detection of the biomarkers between the two mouse strains. Figure 2a and b represents the Venn diagram of estimated DE genes by edgeR, DESeq and Limma using the original and MRS dataset, respectively. From Figure 2a, we revealed that edgeR and Limma performed better than DESeq by sharing more genes (414). We also noticed that there are 1925 overlapping DE genes between these methods. To investigate the performance of the proposed outlier detection and modification approach in this dataset, we first detect and modify the outliers (if any) to get the MRS dataset. We detected 200 outliers in this dataset using the proposed outlier detection rule. The Venn diagram in Figure 2b represents the results of these three methods using the proposed MRS dataset. From this figure we can clearly observe that there are 1956 overlapping genes detected by these methods. Among these genes there are 18 genes that are declared as outliers by the proposed method and those were not detected as DE genes using the original mouse dataset.\n\nVenn diagram of DEGs detected by (a) the edgeR, DESeq and Limma in the original mouse dataset or by (b) the edgeR, DESeq and Limma in the modified mouse dataset using the proposed method. (c) Heatmap of 16 outlying DEGs detected by the proposed method.\n\nFurthermore, we performed the gene overexpression analyses through Database for Annotation, Visualization and Integrated Discovery (DAVID)18 to explore the biological process (BF) categories and pathway annotations of the 18 identified outlying DE genes. Out of 18 genes DAVID identified 16 genes. A heatmap is created for these 16 outlying DE genes in Figure 2c. The heatmap correctly clusters the samples between C57BL/6J (B6) and 11 DBA/2J (D2) using these genes. Among the 16 genes, 8 upregulated (Fam46b, Alx3, Dusp2, Pdyn, Agbl2, Pcdh12, Ubl5, Gpx8) and 8 downregulated (Stard5, Ptprc, Slc7a5, Slc24a1, Ehd2, Adgrg3, Tefm, Tsnaxip1) DE genes are identified. GO analysis results showed that upregulated DE genes are significantly enriched in protein side chain deglutamylation, protein deglutamylation and embryo development at BP level. Downregulated DEGs were enriched in negative regulation of CREB transcription factor activity, negative regulation of NIK/NF-kappaB signaling and sterol import at BP level (see extended data). KEGG analysis showed that the upregulated DEGs were mostly enriched in cocaine addiction, glutathione metabolism and thyroid hormone synthesis. The downregulated DEGs were enriched in phototransduction, primary immunodeficiency and central carbon metabolism in cancer (Table 2). We also constructed the protein-protein interaction (PPI) network around the proteins encoded by these 16 outlying genes using STRING database19. We considered confidence score 400 for selection these networks. Figure 3 and Figure 4 represent the PPI networks using the up-regulated and down-regulated DEGs, respectively. In addition, we explored miRNAs-target gene interactions from miRTarBase20 to identify miRNAs. The miRNAs-target gene interactions network is shown in Figure S1 in extended data.\n\n\nConclusions\n\nBiomarker identification under two or more conditions is an important task for elucidating the molecular basis of phenotypic variation. Next generation sequencing (RNA-seq) has become very popular and a competitive alternative to microarrays because of reducing the cost of sequencing and limitation of microarrays. A number of methods have been developed for detecting biomarkers from RNA-seq data. However, most of the methods are sensitive to outliers and produce misleading results in the presence of outliers. In this study, we have proposed an outlier detection and modification approach using the median control chart. From the simulation study in the presence of outliers we have observed that the performance of five biomarker selection methods are improved significantly when the datasets are modified by the proposed method, both for small-and large-sample cases. The proposed method also detected an additional 16 outlying genes from a real mouse dataset. From GO and KEGG pathway enrichment analysis, we have shown that these genes belong to some important pathways.\n\n\nData availability\n\nSimulated datasets available from: https://doi.org/10.5281/zenodo.221288121\n\nReal dataset: The mouse dataset used in this study is publicly available at the NCBI GEO website: GSE26024.\n\nZenodo: Figure S1: miRNAs-target gene interactions using the outlying genes identified by the proposed method, http://doi.org/10.5281/zenodo.227992122\n\nZenodo: Table A1. Biological process categories for 16 genes, http://doi.org/10.5281/zenodo.228001223\n\n\nSoftware availability\n\nThe R code for the proposed method is available in https://github.com/snotjanu/OutMod-RnaSeq\n\nArchived code: http://doi.org/10.5281/zenodo.227940524\n\nLicense: MIT",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nWe would like to thank the reviewers for their valuable comments on the paper, as these comments led us to an improvement of the work.\n\n\nReferences\n\nMortazavi A, Williams BA, McCue K, et al.: Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat Methods. 2008; 5(7): 621–628. PubMed Abstract | Publisher Full Text\n\nBeyer M, Mallmann MR, Xue J, et al.: High-resolution transcriptome of human macrophages. PLoS One. 2012; 7(9): e45466. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWang L, Feng Z, Wang X, et al.: DEGseq: an R package for identifying differentially expressed genes from RNA-seq data. Bioinformatics. 2010; 26(1): 136–138. PubMed Abstract | Publisher Full Text\n\nNagalakshmi U, Waern K, Snyder M: RNA-Seq: a method for comprehensive transcriptome analysis. Curr Protoc Mol Biol. 2010; Chapter 4: Unit 4.11.1–13. PubMed Abstract | Publisher 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): 139–140. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAnders S, Huber W: Differential expression analysis for sequence count data. Genome Biol. 2010; 11(10): R106. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHardcastle TJ, Kelly KA: baySeq: empirical Bayesian methods for identifying differential expression in sequence count data. BMC Bioinformatics. 2010; 11: 422. 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\nLeng N, Dawson JA, Thomson JA, et al.: EBSeq: an empirical Bayes hierarchical model for inference in RNA-seq experiments. Bioinformatics. 2013; 29(8): 1035–1043. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZhou X, Lindsay H, Robinson MD: Robustly detecting differential expression in RNA sequencing data using observation weights. Nucleic Acids Res. 2014; 42(11): e91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDi Y, Schafer DW, Cumbie JS, et al.: The NBP negative binomial model for assessing differential gene expression from RNA-seq. Stat Appl Genet Mol Biol. 2011; 10(1): 1–18. Publisher Full Text\n\nRobinson MD, Smyth GK: Small-sample estimation of negative binomial dispersion, with applications to SAGE data. Biostatistics. 2008; 9(2): 321–332. PubMed Abstract | Publisher Full Text\n\nGeorge NI, Bowyer JF, Crabtree NM, et al.: An Iterative Leave-One-Out Approach to Outlier Detection in RNA-Seq Data. PLoS One. 2015; 10(6): e0125224. PubMed Abstract | Publisher Full Text | Free Full Text\n\nZwiener I, Frisch B, Binder H: Transforming RNA-Seq data to improve the performance of prognostic gene signatures. PLoS One. 2014; 9(1): e85150. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobinson MD, Oshlack A: A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 2010; 11(3): R25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaw CW, Chen Y, Shi W, et al.: voom: Precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 2014; 15(2): R29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBottomly D, Walter NA, Hunter JE, et al.: Evaluating gene expression in C57BL/6J and DBA/2J mouse striatum using RNA-Seq and microarrays. PLoS One. 2011; 6(3): e17820. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHuang da W, Sherman BT, Lempicki RA: Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 2009; 4(1): 44–57. PubMed Abstract | Publisher Full Text\n\nSzklarczyk D, Morris JH, Cook H, et al.: The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible. Nucleic Acids Res. 2017; 45(D1): D362–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHsu SD, Lin FM, Wu WY, et al.: miRTarBase: a database curates experimentally validated microRNA-target interactions. Nucleic Acids Res. 2011; 39(Database issue): 163–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShahjaman Md: Simulated Data for figure 1 (Version v1). 2018. http://www.doi.org/10.5281/zenodo.2212881\n\nShahjaman Md: miRNAs-target gene interactions using the outlying genes identified by the proposed method (Version v1.0.0). 2018. http://www.doi.org/10.5281/zenodo.2279921\n\nShahjaman Md: Biological process categories for 16 genes (Version v1.0.0). 2018. http://www.doi.org/10.5281/zenodo.2280012\n\nsnotjanu: snotjanu/OutMod-RnaSeq v1.0.0 (Version v1.0.0). 2018. http://www.doi.org/10.5281/zenodo.2279405"
}
|
[
{
"id": "47416",
"date": "29 Apr 2019",
"name": "Lei Li",
"expertise": [
"Reviewer Expertise RNA seq"
],
"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 manuscript, Shahjaman et al. presents a new median control chart method for controlling outliers in RNA-Seq data. They applied this method on simulated data and also a real mouse dataset. In general, a useful bioinformatics method for outlier detection will be valuable for biomarker detection. However, in the present state of the manuscript the robust and efficiency of this method is not clear. There are a few concerns that should be considered to improve the manuscript.\nOutlier methods has actually been implemented in a few tools such as DESeq2 (https://www.bioconductor.org/packages/devel/bioc/vignettes/DESeq2/inst/doc/DESeq2.html#approach-to-count-outliers). The authors need to compare their method with other available outlier detection methods? Also, what is the tool parameters for the authors use in their evaluation? The outlier performance may largely depend on the parameters used. The figure legend needs to be well described. What is the meaning of Figure 1A and Figure 1C, is that one for pre-control and one for after-medial control? Also, Figure 1c and Figure 1d? Why only use three, not five methods for real data performance evaluation? The released github code cannot reproduce the authors’ result.\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? Partly\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "48899",
"date": "06 Jun 2019",
"name": "Jun Li",
"expertise": [
"Reviewer Expertise statistics",
"biostatistics",
"bioinformatics",
"sequencing data analysis"
],
"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 author proposed a method for outlier detection and differential expression (DE) identification for RNA-seq data. While DE is surely an important problem in RNA-seq data analysis, the proposed method is based on a wrong mathematical model and thus makes no sense.\n\nThe author assumes the read count $y_{gik}$ follows a negative binomial model with mean $\\mu_{gi}$ depending only on the gene ($g$) and the condition $i$. Where is the sequencing depth? Normalizing/incorporating sequencing depth has been a central question in DE analysis and significant efforts have been made by many important papers in this field, but the author completely ignored this term. Similarly, without considering the sequencing depth, the null hypothesis the author wrote is also wrong. With the wrong model and wrong hypothesis for testing, the proposed method does not make any sense.\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? Yes\n\nAre the conclusions drawn adequately supported by the results? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/8-7
|
https://f1000research.com/articles/8-6/v1
|
03 Jan 19
|
{
"type": "Opinion Article",
"title": "Hypothesis: Cancer alloustasis",
"authors": [
"Wu Zhou"
],
"abstract": "Despite decades of study, there are still many unanswered questions about cancer metastasis. One of these questions is about the origin of the tumor cells that give rise to metastases. The currently accepted models of cancer metastasis are inconsistent with some clinical observations of the natural history of cancer and its response to therapy, challenging our understanding of the nature of metastasis. Here, we coin a term ‘cancer alloustasis’ to describe the tumor that forms de novo in secondary sites, but is not originated from primary site derived cells. Moreover, we present a complementary hypothesis suggesting that the progenitors of alloustasis are distinct with those of primary tumors and the outgrowth of alloustases is promoted by primary tumors.",
"keywords": [
"Origin of metastasis",
"Cancer alloustasis",
"Complementary hypothesis"
],
"content": "The Primary Site Derived hypothesis\n\nSince the days of the German pathologist Rudolf Virchow (1821–1902), the prevailing view for the origin of cancer metastasis has been the assumption that the anatomic progression of a malignant cell population is a stepwise movement of cells from the primary tumor to the regional lymph nodes and thence to more distant organs. Cancer metastases are believed to be the end result of a multistage process that includes local tissue invasion by primary tumor cells, intravasation into blood vessels or lymphatic system, survival during transit in circulation, arrest within a distant organ, extravasation, survival in the new tissue environment, and proliferation to produce metastatic colonization (Gupta & Massague, 2006).\n\nIn line of this hypothesis, Nowell proposed the clonal selection model, where only small sublines of cells within primary tumor acquire genetic permits grow as metastases under stepwise selection (Nowell, 1976). Whereas weiss emphasize on dynamic or compartmental heterogeneity, stating that the primary tumor is composed of a heterogeneous population of cells with metastatic capability, but metastatic phenotype is only a property of a given tumor cells due to epigenetic changes (Weiss, 1990). Other major models of metastatic growth include the clonal dominance model (Kerbel et al., 1988), metastasis gene transfer theory (Garcia-Olmo et al., 2004), the fusion model (Pawelek, 2005; Pawelek & Chakraborty, 2008) and the parallel progression theory (Klein, 2009). All these models share the common belief that metastasis emerge from primary tumor or primary lesion. I nominate these models for the origin of metastasis as the Primary Site Derived hypothesis (PSD hypothesis).\n\n\nThe phenomena that challenge PSD hypothesis\n\nThere are several complex biological phenomena that challenge this assumption of metastatic progression. One important clinical observation is of metastatic tumors from an unknown primary source. It is estimated that 2–6% of cancer patients (according to various literature reports) presenting to oncology units have cancer metastases in the absence of an identifiable primary tumor.\n\nIn these clinic cases, the patient is considered to have ‘unknown’ or ‘occult’ primary tumor and characterized as unknown primary origin (CUP) or unknown primary tumors (UPT) (Ettinger et al., 2011). The existence of CUP suggests that metastases may not always be derived from primary tumors (Greco, 2014; Oskarsson et al., 2014; Vanharanta & Massague, 2013).\n\nMoreover, several studies show that the formation of metastatic tumors is not definitely rely on the formation of primary tumor. The morphologically normal breast cells from genetically engineered mice were injected into the tail veins of other female mice. The results showed that tumors were formed resulted in the lungs but not in the breast after oncogene induction (Podsypanina et al., 2008). In addition, the cultured metastatic cells may even be detected in the primary organs rather than secondary sites, for example, the kidneys and heart (Friberg & Nystrom, 2015). Another uncertainty is whether metastasis is really the result of cancer cell dissemination to a secondary organ through the bloodstream or lymphatic vessels as circulating tumor cells (CTCs). For example, a study has estimated that the probability of not detecting any CTCs in blood from metastatic breast cancer patients is 0.6 (Shahriyari, 2016). Although this finding may be due to technical issues to detect CTCs in all patients, it may also suggest that other mechanisms may operate to create tumors in other organs.\n\n\nDoubts about the PSD hypothesis\n\nThe above experimental or clinical observations generate a follow-up doubt: Do the tumors in secondary sites develop always from the cancer cells preexisting within a primary tumor or absolutely from premalignant cells originated from primary lesion? It is important to note that our current understanding of the biology of cancer metastasis is strongly influenced by experimental models including in vitro and ex vivo systems, mimic in vivo models developed on Drosophila Melanogaster, zebrafish, fertilized chicken eggs and mices (van Marion et al., 2016). These models contribute to study the tumor progression especially combined with whole body imaging techniques, which enables the early detection of small metastases and track longitudinally their fate in the same animal. However, the clinical observations is frequently not consistent with the metastatic process in the animal models.\n\nIn this regard, the ideal model about the origin of secondary tumors should be summarized solely from clinical observation. Clinically, most secondary tumors occur later when the primary tumor is larger. Nevertheless, secondary tumors can occasionally form early in tumor progression when the primary tumor is still small or even undetectable. The parallel progression model (Klein, 2009) explain this phenomena as tumor cells depart the primary lesion before the acquisition of fully malignant phenotypes to undergo somatic progression and metastatic growth at a distant site. This model well illustrates the characteristic biology of early dissemination (Hosseini et al., 2016; Husemann et al., 2008; Klein et al., 2002) while remaining questions are left open. If parallel progression hypothesis is applicable for all secondary tumors, we would expect that the ever-increasing ability to detect secondary tumors at earlier stages should certainly lead to a major reduction in mortality. In fact, despite earlier diagnosis and improved treatment modalities and supportive care, age-adjusted mortality rates have not appeared to decrease.\n\nCancer is a complex disease and our current understanding of systemic cancer is insufficient. Unfortunately, neither prevailing models about the origin of secondary tumors is supported by direct and incontrovertible evidence. Therefore, I address the possibility that some tumors formed de novo in secondary sites are not coming from the primary tumor or primary lesion. These particular tumors, in a sense, grow up like ‘primary tumors’ and according to their own principles.\n\n\nThe OSD hypothesis\n\nI propose a model to describe the origin of cancer metastases in some patients and nominate this model as Own Site Derived hypothesis (OSD hypothesis). The major points that involved into the OSD hypothesis are:\n\nMetastasis is a Greek word meaning ‘displacement’, from μετά, meta, ‘next’, and στάσις, stasis, ‘placement’. It was coined in 1829 by Jean Claude Recamier and is now defined as ‘the transfer of disease from one organ, or part, to another not directly connected to it’ (Fidler & Balch, 1987). Since cancer metastasis is commonly considered as the secondary tumors derived from primary tumor, we create a term ‘cancer alloustasis’ to define the tumors developed from particular cancer cells that are not coming from primary tumor. This idea is inspired by changing μετά, meta, ‘next’ to άλλού, allou, ‘elsewhere’. Accordingly, ‘cancer alloustasis’ means ‘elsewhere placement of tumor’. ‘Alloustasis’ is a singular noun. Its plural, adjective and transitive verb is denoted as alloustases, alloustatic and alloustasize, respectively.\n\nCancer is a systemic disease. The origin of cancer cells is the result of clonal selection and evolution. It is possible that there are transformed cells arising from both primary and secondary organs before the occurrence of primary tumor. The fate of these progenitors of cancer is determined by their adaptability for the host environment. The tumor that acquires sustainable growth capability will continue to grow as a malignant lesion. Otherwise the transformed cells will maintain their premalignant state or be eliminated by their own microenvironment (Figure 1). According to the OSD hypothesis, the first malignant lesions are defined as primary tumors and the organ where they emerge from is described as primary organ. The transformed cells located into primary organ are distinct with those transformed cells located into secondary sites. Even after the outgrowth of primary tumor(s), the transformed cells in secondary sites do not emerge from a subpopulation of cells present within a primary tumor but rather selected out by their own microenvironment.\n\nThe premalignant cells arise from both primary and secondary organs before the outgrowth of primary tumor. The fate of these progenitors is determined by their adaptability for the host environment and their own microenvironment. a. The transformed cells arise from multiple organs of patient. b. The special transformed cell acquires priority selection. c. The selected transformed cell continues to grow as primary tumor and other transformed cells maintain their premalignant state or are eliminated by their own microenvironment.\n\nThe host environment of the patient is dynamic. The successful growth of the primary tumor may remodel the host environment of the patient’s whole body by excreting humoral factors (such as hormones or cytokines) or by an immune response against those tumor cells. It follows that primary tumors change the microenvironment of secondary site and induce the outgrowth of alloustasis which, in the same manner, produces secondary alloustases (so-called ‘alloustasis of alloustases’) (Figure 2). Thus, in a short time, a small primary tumor may produce a cascade of alloustases. This mechanism reveals why cancer exponentially increases the clinical impact to the host and causes severe diseases, for example paraneoplastic syndrome (Finora, 2003; Torrielli et al., 1971) and cachexia (Fearon & Moses, 2002).\n\na. The transformed cells arise from multiple organs of the patient. b. The special transformed cell acquires priority selection. c. The selected transformed cell continues to grow as primary tumor. The outgrowth of primary tumor promotes the formation of alloustasis. d. The outgrowth of alloustasis promotes the formation of alloustasis of alloustases.\n\nAs I mention before, cancer is a selective and evolutionary process. It is possible that there are multiple transformed cells arising from multiple sites among various organs. These transformed cells will influence the host environment of the whole body and directly or indirectly cooperate with each other to promote the progression of primary tumors. Finally, only part of them acquire the ability to continue growing as malignant lesions (Figure 3) which is observed clinically as a result of a ‘field effect’ (Freireich et al., 2005).\n\na. The transformed cells arise from multiple sites of different organs. b. Multiple transformed cells into the same organ are selected out and continue to grow as multiple primary tumors. c. The outgrowth of multiple primary tumors promotes the single or multiple alloustasis.\n\nAccording to the OSD model, the particular pre-cancerous cells are kept evolutionary and equally selected by host environment of the whole body prior to the formation of primary tumor. Moreover, we propose that the growth of malignant lesions (including primary tumors and alloustasis) is dynamically regulated by the fluctuation of host environment in some patient. After the outgrowth of malignant lesions, they may even be destroyed by their local environment (Figure 4). This can be applied for explanation of the CUP behavior and phenomena of regression of human cancers without treatment (spontaneous regression) which is first reported more than 100 years ago and is well documented for many types of cancer, albeit with low frequency (Harada et al., 2010; Kumar et al., 2010, Onuigbo, 2012, Strub et al., 2013).\n\na. The transformed cells arise from multiple organs of patient. b. The outgrowth of primary tumor promotes the formation of alloustasis. c–e. The malignant lesions of patient are dynamically regulated by the host environment and their own microenvironment. The primary tumor and alloustases are gradually eliminated or even promoted by the host.\n\n\nConcluding remarks\n\nHere I challenge the prevalent hypothesis of cancer metastasis, and consider that, in some patients, the ancestral cells of secondary tumors do not necessarily have to be cancer cells originating from a primary site. I propose a complementary OSD hypothesis to indicate that some tumors formed in situ into secondary sites (defined as alloustases) resulting from clonal selection and evolution for adapting the dynamic environment (tumor microenvironment and host environment of the whole body).\n\nAccording to the OSD hypothesis, the relationship of primary tumors and alloustases is mutually independent but interrelated. In a sense, the origin of alloustasis seems like that of a primary tumor. However, the hypothesis highlights that the progenitors of alloustases differ from those of the primary tumors and the progress of alloustasis is promoted by the formation of primary tumor. Meanwhile, the primary tumors and alloustases may exhibit parallel evolution, and they will cooperate or even compete with each other during the progress of cancer.\n\nI also emphasize the central significance of environment. The local microenvironment and systemic host environment of malignant lesions (both primary tumors and alloustases) are not stable but rather very dynamic. Hence the cancer lesions may be diminished or eliminated by the environment. Notably, we consider cancer alloustasis as a systemic disease. In that case, I think it is necessary to treat cancer as a systemic illness. More attention should be focused on systemic treatment and prevent of cancer as they relate to etiology. Perhaps the identification of systemically-acting carcinogens based on the systemically therapeutic strategies might be targeted clinically and would have a substantial impact on overall cancer mortality.\n\n\nData availability\n\nNo data is associated with this article.",
"appendix": "Grant information\n\nThis work was supported by the National Natural Science Foundation of China (No. 81572879).\n\n\nReferences\n\nEttinger DS, Agulnik M, Cates JM, et al.: NCCN Clinical Practice Guidelines Occult primary. J Natl Compr Canc Netw. 2011; 9(12): 1358–1395. PubMed Abstract | Publisher Full Text\n\nFearon KC, Moses AG: Cancer cachexia. Int J Cardiol. 2002; 85(1): 73–81. PubMed Abstract | Publisher Full Text\n\nFidler IJ, Balch CM: The biology of cancer metastasis and implications for therapy. Curr Probl Surg. 1987; 24(3): 129–209. PubMed Abstract | Publisher Full Text\n\nFinora K: Common paraneoplastic syndromes. Clin Tech Small Anim Pract. 2003; 18(2): 123–126. PubMed Abstract | Publisher Full Text\n\nFreireich EJ, Kurzrock R, Estrov Z: Metastasis--an alternative hypothesis. Cancer. 2005; 103(8): 1537–1539. PubMed Abstract | Publisher Full Text\n\nFriberg S, Nyström A: Cancer Metastases: Early Dissemination and Late Recurrences. Cancer Growth Metastasis. 2015; 8: 43–49. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGarcia-Olmo DC, Ruiz-Piqueras R, Garcia-Olmo D: Circulating nucleic acids in plasma and serum (CNAPS) and its relation to stem cells and cancer metastasis: state of the issue. Histol Histopathol. 2004; 19(2): 575–583. PubMed Abstract | Publisher Full Text\n\nGreco FA: Cancer of unknown primary site: still an entity, a biological mystery and a metastatic model. Nat Rev Cancer. 2014; 14(1): 3–4. PubMed Abstract | Publisher Full Text\n\nGupta GP, Massagué J: Cancer metastasis: building a framework. Cell. 2006; 127(4): 679–695. PubMed Abstract | Publisher Full Text\n\nHarada T, Sakaguchi T, Inaba K, et al.: [Spontaneous regression of hepatocellular carcinomas]. Nihon Shokakibyo Gakkai Zasshi. 2010; 107(3): 432–441. PubMed Abstract | Publisher Full Text\n\nHosseini H, Obradović MM, Hoffmann M, et al.: Early dissemination seeds metastasis in breast cancer. Nature. 2016; 540: 552–558. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHusemann Y, Geigl JB, Schubert F, et al.: Systemic spread is an early step in breast cancer. Cancer Cell. 2008; 13(1): 58–68. PubMed Abstract | Publisher Full Text\n\nKerbel RS, Waghorne C, Korczak B, et al.: Clonal dominance of primary tumours by metastatic cells: genetic analysis and biological implications. Cancer Surv. 1988; 7(4): 597–629. PubMed Abstract\n\nKlein CA: Parallel progression of primary tumours and metastases. Nat Rev Cancer. 2009; 9(4): 302–312. PubMed Abstract | Publisher Full Text\n\nKlein CA, Blankenstein TJ, Schmidt-Kittler O, et al.: Genetic heterogeneity of single disseminated tumour cells in minimal residual cancer. Lancet. 2002; 360(9334): 683–689. PubMed Abstract | Publisher Full Text\n\nKumar T, Patel N, Talwar A: Spontaneous regression of thoracic malignancies. Respir Med. 2010; 104(10): 1543–1550. PubMed Abstract | Publisher Full Text\n\nNowell PC: The clonal evolution of tumor cell populations. Science. 1976; 194(4260): 23–28. PubMed Abstract | Publisher Full Text\n\nOnuigbo WI: Spontaneous regression of breast carcinoma: review of English publications from 1753 to 1897. Oncol Rev. 2012; 6(2): e22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nOskarsson T, Batlle E, Massagué J: Metastatic stem cells: sources, niches, and vital pathways. Cell Stem Cell. 2014; 14(3): 306–321. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPawelek JM: Tumour-cell fusion as a source of myeloid traits in cancer. Lancet Oncol. 2005; 6(12): 988–993. PubMed Abstract | Publisher Full Text\n\nPawelek JM, Chakraborty AK: Fusion of tumour cells with bone marrow-derived cells: a unifying explanation for metastasis. Nat Rev Cancer. 2008; 8(5): 377–386. PubMed Abstract | Publisher Full Text\n\nPodsypanina K, Du YC, Jechlinger M, et al.: Seeding and propagation of untransformed mouse mammary cells in the lung. Science. 2008; 321(5897): 1841–1844. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShahriyari L: A new hypothesis: some metastases are the result of inflammatory processes by adapted cells, especially adapted immune cells at sites of inflammation [version 1; referees: 3 approved]. F1000Res. 2016; 5: 175. PubMed Abstract | Publisher Full Text | Free Full Text\n\nStrub B, Moron M, Meuli-Simmen C, et al.: [Merkel cell carcinoma of the right cheek and spontaneous regression--case report and literature review]. Zentralbl Chir. 2013; 138 Suppl 2: e113–116. PubMed Abstract | Publisher Full Text\n\nTorrielli A, Ribatto A, Gatto PL: [Paraneoplastic syndromes: critical review of the literature and report of cases]. Pathologica. 1971; 62(909): 105–127. PubMed Abstract\n\nvan Marion DM, Domanska UM, Timmer-Bosscha H, et al.: Studying cancer metastasis: Existing models, challenges and future perspectives. Crit Rev Oncol Hematol. 2016; 97: 107–117. PubMed Abstract | Publisher Full Text\n\nVanharanta S, Massagué J: Origins of metastatic traits. Cancer Cell. 2013; 24(4): 410–421. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWeiss L: Metastatic inefficiency. Adv Cancer Res. 1990; 54: 159–211. PubMed Abstract | Publisher Full Text"
}
|
[
{
"id": "44863",
"date": "20 Mar 2019",
"name": "Christoph A. Klein",
"expertise": [
"Reviewer Expertise cancer biology",
"metastasis"
],
"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 opinion article by Wu Zhou provides an interesting hypothesis about the origin of systemic cancer. It starts by opposing the Primary Side Derived (PSD) hypothesis of systemic cancer with the Own Side Derived (OSD) hypothesis. In a nutshell, the OSD hypothesis suggests that there is no migration from one side to another but that tumors arise in different organs in parallel. The first diagnosed tumor is then called the \"primary tumor\" whereas the other growths are called alloustasis, which means that there is another tumor at another site.\nThereby, the author provides a different concept of systemic cancer, in which the migration inherent to the standard metastasis model is replaced by growth promoting factors of the first, i.e. the leading lesion, which may support progression and diagnosis of the alloustasis. In a way, this is similar to the parallel progression model suggested some years ago (Klein NRC, 20091), in which early disseminated cancer that are genetically too immature to autonomously grow into a metastasis receive systemically acting, growth-promoting support from the primary tumor.\n\nCriticism:\n\nWhile the concept is interesting and thought provoking, I think the paper needs some more work before being listed in pubmed, since students and scientists in training need more information to fully judge its content.\n\n1) Text and nomenclature:\nI strongly recommend that the author revise the text with help of a native speaker. The understanding of such an article fully depends on a precise language. For example the sentence \"The transformed cells located into primary organ are distinct with those transformed cells located into secondary sites\" is unclear, imprecise and grammatically incorrect. There are numerous sentences like this, which reduce the clarity and scientific vigor of the concept. I also strongly recommend that the author spend some more thinking on how to adapt his nomenclature. For example, the term \"malignant\" indicates in the standard model that a proliferating lesion seeds cells to other sites and generates systemic cancer. Dissemination is defining malignancy. What would be the meaning of malignant in the alloustasis model? Since invasion and dissemination do not exist, how would the author differentiate malignant vs. benign tumors? My school knowledge about ancient Greek has become minimal, however I wonder whether \"alloustasis\" should be replaced by allestasis. \"Stasis\" is feminin and \"alle\" would be the corresponding form of the adjective.\n\n2) References / state of the art\nFor several statements, the author should provide the references. For example, I would not agree that early detection and local therapy does NOT impact on outcome. As such the statement that \"... the ever-increasing ability to detect secondary tumors at earlier stages should certainly lead to a major reduction in mortality. In fact, despite earlier diagnosis and improved treatment modalities and supportive care, age-adjusted mortality rates have not appeared to decrease\" is at best misleading if not incorrect. References should be provided to justify this view. As mentioned above, the arguments in favor of a metastasis model should be summarized. This includes a careful analysis of CUP. Recent work on the epigenetic tracking of tumor origin (see the Lancet Oncology paper of the Esteller group2) has generated some evidence in favor of a metastasis model. In fact, consideration of the genetic and epigenetic comparisons between primary tumours and what is currently judged as metastases, may be a useful starting point to define the need of an alloustasis concept.\n\n3) Explanatory power of the concept\nScientific theories and concepts aim to properly describe reality. Any change needs to be evaluated by its contribution to better explain observed phenomena. The author provides some arguments in this direction, however, this part needs to be substantially improved. For example, cancer of unknown primary is the major argument in favor of alloustasis so far. However, early dissemination and regression of the primary lesion explains CUP equally well. I would suggest that the author generates a table of all arguments in support of alloustasis. The article should also contain a list of arguments that support the \"metastasis\" concept. What is then needed is that the alloustasis model explains all/most metastasis-supporting phenomena equally well. Such an approach will reveal, for which cancer cases alloustasis may indeed provide a reasonable explanation and which cases would be better explained by a metastasis model. In line with this point, it would extremely helpful if the author provides a clear separation of the two models. For example, estrogen receptor expression in a breast cancer and in a metachronous brain lesion is currently taken as evidence for a brain metastasis from a hormone receptor positive breast cancer. How would the alloustasis model explain this? Or would the author agree that in this case metastasis has occurred? Similarly, epithelial growths in mesenchymal organs like lymph node or bone marrow: how are these explained in an alloustasis model given the fact that these organs do not contain epithelial cells that could give rise to an alloustatic epithelial cancer? When such confrontation of the two models has been elaborated, the author should check, which model is simpler, i.e. he should apply Ockham's razor. Finally, I would greatly appreciate, if the author could provide his thoughts about data that would falsify alloustasis. Would there be any experiment or finding that would lead to rejection of the concept? Falsification criteria are extremely helpful to generate new research, even if methods for such falsification are currently not available.\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": []
}
] | 1
|
https://f1000research.com/articles/8-6
|
https://f1000research.com/articles/7-620/v1
|
21 May 18
|
{
"type": "Opinion Article",
"title": "Lost in translation",
"authors": [
"Parashkev Nachev",
"Geraint Rees",
"Richard Frackowiak",
"Geraint Rees",
"Richard Frackowiak"
],
"abstract": "Translation in cognitive neuroscience remains beyond the horizon, brought no closer by supposed major advances in our understanding of the brain. Unless our explanatory models descend to the individual level—a cardinal requirement for any intervention—their real-world applications will always be limited. Drawing on an analysis of the informational properties of the brain, here we argue that adequate individualisation needs models of far greater dimensionality than has been usual in the field. This necessity arises from the widely distributed causality of neural systems, a consequence of the fundamentally adaptive nature of their developmental and physiological mechanisms. We discuss how recent advances in high-performance computing, combined with collections of large-scale data, enable the high-dimensional modelling we argue is critical to successful translation, and urge its adoption if the ultimate goal of impact on the lives of patients is to be achieved.",
"keywords": [
"Translation",
"high-dimensional inference",
"causality",
"neuroimaging",
"cognitive neuroscience",
"machine learning."
],
"content": "The question\n\nCognitive neuroscience is yet to produce applications of major clinical impact. If its relative immaturity is to blame, we need merely wait. But if its approach is fundamentally ill-suited, we could be left waiting forever. We must therefore consider how well the means of cognitive neuroscience support translational ends. Such consideration cannot be expected to emerge spontaneously from the field itself, for neuroscience evolves under the selective pressure of supposed understanding, not the collateral of mechanistic insight translation is widely perceived to be. Nor may we presume the obstacles to translation to be peculiar to each cognitive subfield and unlikely to be illuminated by a general analysis: it is possible they lie within the proximal, cardinal steps common to all of neuroscience. Here we examine this possibility, show it to be overwhelmingly likely, and outline how neuroscience must change if it is to deliver real-world patient impact.\n\n\nTranslation and individualisation\n\nMost societies give primacy to the individual person, imposing collective interests only with reluctance. This is especially true of healthcare, where the object of clinical action is archetypally the individual and the group only secondarily. To take a striking example, we could overnight revolutionize population outcomes in acute stroke by intervening with thrombolytic therapy at the kerbside, bypassing delays that hospital transfer for diagnostic computed tomographic scans inevitably introduce (Wardlaw et al., 1997). However, no matter how compelling the population statistics, such a manoeuvre is rendered unconscionable by the resultant death or greater disability of the 10% of patients with primary intracerebral haemorrhage (Qureshi et al., 2009). Even where the stakes are less sharply polarised, it remains difficult to implement any treatment whose individual benefit is only crudely probabilistic, for all interventions have a cost: both personal and financial. Moreover, since populations merely summarise effects on individuals, the greater the individual variation, the lesser the population-level impact. Both constitutively and politically, translational success or failure is thus critically dependent on our ability to individualise our interventions.\n\nHow is individuality determined? Consider by way of illustration that most personal part of the body, the face (Figure 1). Though one feature may sometimes be uniquely idiosyncratic, to distinguish a face from another generally requires the conjunction of many features, even when all redundancy is eliminated. Such irreducibly high intrinsic dimensionality is conveniently captured by the notion of minimum description length (Rissanen, 1978) – intuitively, the most compressed complete description of a system. This quantity sets a hard limit on the minimal complexity of any model that must distinguish one state or instance of a system from another to perform its task. No matter how clever the mathematics, a machine vision model tasked with (say) classifying the sex of a face will always perform badly when starved of input features because no small subset of features contains the necessary information; conversely, even a relatively unsophisticated model with sufficient capacity will perform well, given enough data (Parkhi et al., 2015; Schroff et al., 2015; Zhou et al., 2015). It should come as no surprise that face coding in the primate brain takes a high-dimensional approach, deriving identity by projecting a multiplicity of features onto a compacted representational space (Chang & Tsao, 2017). Now our concern is not individuation simpliciter but the individuation of causal mechanisms of predictive or prescriptive utility. For this we need a causally constrained extension of the concept of minimal description length: what we here term a minimal causal field. To see how this is specified requires a brief examination of biological causality.\n\nThe face of the Roman Emperor Hostilian (top left) is poorly described by the canonical face of all Roman Emperors (top right), which is—by definition—not identical with any of the individual faces from which it is derived. Furthermore, the individuality of a face is better captured by a low-precision, high-dimensional parameterisation (bottom left), than it is by a high-precision, low-dimensional parameterisation such as the inter-pupillary distance (bottom right). The photograph of Hostilian is reproduced with the kind permission of Dr William Storage.\n\n\nNeural causation\n\nWe have a natural intellectual predisposition to causal models with two cardinal features: economy and seriality (Hacker, 2007). This is a consequence partly of reasoning by analogy and partly of practicability. The intelligible, mechanistically pellucid processes we observe in the non-organic world and exploit in the machines we build tend to have few parameters of causal significance, arranged sequentially. It seems natural to apply the same approach to biology, indeed inevitable, for a causal model with (say) a thousand parameters is intellectually intractable. When we insist on identifying necessary and sufficient links within a more or less serial chain, it is because no other option has been open to us.\n\nBut whereas this notion of causation is adequate for understanding simple, serially organised systems, it does not scale with complexity. In complex systems, where a multiplicity of factors is jointly brought to bear on the outcome, each individual factor becomes an insufficient but necessary part of a set of factors that are unnecessary but sufficient for the result: an INUS condition (Mackie, 1974). To give an adequately explanatory account it is necessary to specify a causal field of many such INUS conditional factors that interact in complex ways (see Figure 2).\n\nDistributed causality is elegantly illustrated by the behaviour of artificial neural networks trained to transform an input into an output by optimising the weights of a stack of fully connected nodes. Here the input-output transformation is causally dependent on the nodes and their connections, for it cannot occur without most of them. But when the network is large, its dependence on any limited subset of nodes will be low. This is not because there is a reserve of unused nodes, but because the causality of the system is constitutionally distributed. Inactivating (in black) a set of nodes (large circles) or their connections (small circles) will thus degrade performance broadly in proportion to their number and not necessarily their identity. Causality thus becomes irreducible to any simple specification of necessity and sufficiency. Instead, each node becomes an insufficient but necessary part of an unnecessary but sufficient set of factors: an INUS condition. An adequate description of the causality of the system as a whole then requires specification of the entire causal field of factors: no subset will do, and no strong ranking need exist between them. If the architecture of real neural networks makes such causality possible—and it certainly does—we need to be capable of modelling it. But this is more than just a theoretical possibility. It is striking that encouraging distributed causal architectures through dropping nodes or connections during training dramatically improves the performance of artificial neural networks. And, of course, real neural substrates often exhibit remarkable robustness to injury, a phenomenon conventionally construed as “reserve”, but since no part of the brain lies in wait, inactive, distributed causality is a more plausible explanation.\n\nDo neural systems require such a complexity of causal specification? Consider the far simpler behaviour of artificial neural networks, such as deep-learning architectures in which layers of laterally connected units are hierarchically arranged in an end-to-end error-minimising stack (Goodfellow et al., 2016; LeCun et al., 2015). Taking the input-output transformation produced by such a network as its “function”, we can test the causal contribution of sets of network nodes by examining the functional consequences of deactivating them, essentially performing artificial neural network lesion-deficit mapping (Adolphs, 2016; Rorden & Karnath, 2004). When a trained network is subjected to such drop-out (Gal & Ghahramani, 2016; Le Cun et al., 1989; Srivastava et al., 2014; Wan et al., 2013), the degradation of any output is gradual, and often proportionate with the mass of deactivated nodes but varying with their identity in a complex manner that precludes the identification of a \"critical\" node, or even a clear ranking of the material contribution of individual nodes. Causality is constitutionally distributed in a way any conventional description simply cannot capture; only a causal field specification will do (Mackie, 1974).\n\nThe widespread use of drop-out in the deep-learning literature shows causally distributed architectures learn complex input-output transformations better than other systems examined to date (Goodfellow et al., 2016). They are also more robust to damage, an important consideration for any biological system. However, there is no need to appeal to plausibility in our argument: to the extent that a node contributes to function it must be causally relevant. It is inconceivable that the observed complexity of real neural systems is merely epiphenomenal to a much simpler underlying causal organisation. In any event, since we cannot assume the minimal causal field is small, we need to consider how to model it when it is irreducibly large.\n\n\nMapping causal fields\n\nIn seeking to understand the causality of any complex system, we must distinguish between causally relevant and incidental variation. Though rarely acknowledged, the approach to making such a distinction depends critically on a cardinal assumption about a system’s structure. If we assume the organisation of a particular brain network is fundamentally the same across people—i.e. it is monomorphous—individual variation may be treated as noise. The population mean will then be the best available guide to the fundamental mechanism and to its expression in individuals. This is an implicit assumption behind the vast majority of studies in cognitive neuroscience where a set of estimates, derived from a small group, are considered to reveal general truths about the brain. But if, at some causally critical level, the neural organisation is not the same across people—i.e. it is polymorphous—individual variation cannot be treated as noise and the population mean will be a poor guide, both to mechanism and individual behaviour (see Figure 3). The distinction between monomorphous and polymorphous organisation is crucial because it radically alters the optimal inferential approach. We suggest the common assumption of a monomorphous architecture of the brain is unjustified—both empirically and theoretically—and must be discarded, for the following reasons.\n\nWhere the fundamental architecture of a biological system is the same, our best guide will be the simple mean of the population, for each individual will differ from it randomly. Studying such monorphous systems is illustrated by adding random noise to an image of a specific watch mechanism, and then averaging across 45 noisy instances. The underlying architecture is thereby easily revealed. Where the solution in each individual differs locally, illustrated by taking a family of 45 different watch mechanisms of the same brand, the population mean is a very poor guide, for individual variability is no longer noise but the outcome of a plurality of comparably good solutions. We must instead define local regularities of organisation, here done by t stochastic neighbour embedding of the images into a two dimensional latent space, revealing characteristic features of each family of solutions. Given that neural systems are complex, stochastically initiated, and optimised by feedback, polymorphous architectures are likely to dominate, mandating a data-driven, neighbourhood-defining approach to modelling.\n\n\nThe genetic information gap\n\nA neural architecture can be shared across individuals only as far as it is identically specified by the genome, the environment, and their interaction. The constitutive variability of the environment leaves the genome as the primary driver of inter-individual homology. Genomic information content is information theoretically limited by the number of base pairs and the range of nucleotide options at each locus. If we implausibly (Rands et al., 2014) assume every locus is both functional and material to the operations of the brain, so that no section is redundant, we have only ~6 x 109 bits of information, roughly the content of an old compact disc. Even if all this information is used to specify the minimal causal field of a human brain, leaving none for the rest of the body, we remain unable to meet even the most conservative estimates of the brain’s complexity. A commonly offered prenatal estimate, ~1014 bits, derived from the number of synapses in the brain (Huttenlocher & Dabholkar, 1997; Tang et al., 2001), implausibly assumes a synapse can only encode one bit at any one time, and that neural connectivity is the only differentiator. This is equivalent to treating a neuron rather like a transistor in a modern computer-processing unit, distinguished from its neighbours only by the role assigned to it. In short, we are not faced with an information gap but more an information chasm. The conclusion is that a great deal of the functional architecture of the brain cannot be monomorphous, for the necessary information simply is not there.\n\n\nCreating polymorphous architectures\n\nThe brain cannot violate the laws of physics, so how can complexity arise from so relatively impoverished an initial specification? Theoretically the simplest approach is to inject randomness (Matsuoka, 1992) at the outset of development, allowing a complex order to emerge downstream through feedback learning.\n\nSuch stochastic initiation is evident in normal neural development, where as many cells face an orchestrated death, at great structural and energetic cost to the organism, as survive into adulthood (Lossi & Merighi, 2003). Seemingly playing a compound game of “Russian roulette cum musical chairs”, developing neurons are subjected to an environmentally dependent selection process, determined only once development is in play. The process is not fully specified in the genome, or else the redundant neurons would never be born. The biologically dominant prohibition of regeneration in the central nervous system, far from being mysterious, is necessary where the organising information emerges during development, and is therefore stored only in the final product itself.\n\nEqually, the ubiquity of neural feedback learning is evident in the way recurrence is so densely woven into the neural fabric. One-way brain pathways are an exception, not the rule (Bressler & Menon, 2010). It could not be otherwise, for learning—here neural learning—is the only way an order more complex than the initial genetic specifications could conceivably arise.\n\nNow a stochastically-initiated, feedback-learning system, with multiple tuneable parameters, will inevitably have many different solutions for the same target input/output transformation. It is therefore bound to be polymorphous. Crucially, there need be no mechanism for regularising such solutions across individuals to impose a higher, species-level order, for no such order need exist, even if it could be imposed. An organism adapts its structure in response to errors only within its own input-output transformations, not those of others; biology does not do federated learning (McMahan et al., 2016).\n\nThough our concern is to define the bounds of biological possibility our models must be able to cover, it is natural to seek empirical evidence of biological plausibility. The only credible evidence can come from a system that has been comprehensively characterised. Since our claim is about under-estimating complexity, we may as well pick a simple one. Consider, the gut of the lobster, or rather the stomatogastric sub-circuit, meticulously studied for decades by Eve Marder. Though absurdly simple anatomically, with only 30 neurons, and physiologically, with only regular peristaltic oscillation, the relation between the two is not only complex, but also polymorphous in precisely the way described. The same functional physiology can be arrived at from different individual neuronal “settings”, both across time in the same animal and across different animals (Marder & Bucher, 2007). We cannot presume that the rules of human functional brain organisation are any simpler.\n\n\nModelling polymorphous systems\n\nWe must confront the functional complexity of neural organisation before us if translation from mechanisms of disease to rational treatments is to be possible. How do we generate, estimate, and validate polymorphous neural models of potentially incomprehensible complexity?\n\nLet us begin with the last step first: validation. It is conventional to take goodness-of-fit, qualified by some statistical measure, as evidence for the plausibility and utility of a model. But this is of little use where the field of possible models is both vast and sparsely sampled. That our model shows a degree of fit with the data means little if uncountable very different models fit just as well or better. The practice is kin with awarding oneself a gold medal after finishing a race blind to the rest of the field. Nor is limited model comparison acceptable, for differentiating between a handful of models tells us little about the sea of possibilities from which they are drawn.\n\nRather, we need to quantify the individual predictive power of a model, across time or across individuals, in relation to the future state of a system, or some outcome measure of interest. Such prediction is naturally framed in standard terms of sensitivity and specificity, derived from a comprehensive spread of data the model has not seen (Dwork et al., 2015; Vapnik, 1998). A model with perfect predictive power cannot be improved upon, so its competitors may be reasonably dismissed. A model with imperfect predictive power is to be stratified by metrics, leaving as much or as little room for exploring others as its performance dictates.\n\nCrucially, if a given model is powerfully predictive, none of the constituent features can be treated as noise, no matter how random they may appear when viewed in isolation. This approach does not implicate any individual constituent feature mechanistically, because functionally irrelevant incidentals (data and/or features) may drive prediction, like correlation. But it does imply no component feature leading to a good prediction can be safely ignored.\n\nOf course, the richer the parameterisation of a model, the more susceptible it is to “overfitting” - the identification of coincidences of features in a dataset arising by chance with no predictive power beyond it (Hawkins, 2004). But that is no more reason for avoiding such an approach than the possibility of being dazzled is a reason for keeping one’s eyes permanently shut. It is in any event a practical, not a theoretical objection, addressable through the use of large-scale, fully inclusive datasets and high performance computing, as we discuss below.\n\nTo insist on intuiting a hypothesis as the first investigative step imposes a bias towards models couched in familiar concepts within a contemporary sphere of comfort. Where the hypothesis space is too large for our imaginations to traverse confidently, relying on intuition is not principled but hubristic. We need a formal hypothesis generation step, explicitly driven by exploratory analysis of data at sufficient scale and with adequate dimensional richness. The optimal scale and dimensionality will vary unknowably with any specific problem, but since both are likely to be very large, practical feasibility shall generally be the limit (Ghahramani, 2015).\n\nThe manner of model generation constrains subsequent model estimation. If the former requires high dimensionality so will the latter. We cannot assume the underlying causal field to be sparse, or that its components will be linearly separable. In attempting to compress the dimensionality of models—explicitly through the use of a feature selection step, or implicitly through the use of sparsity-promoting inferential methods—we need to watch the impact on individual predictive power, assessed over a sufficiently diverse sample. Where a smooth decrement in prediction performance is seen with feature reduction, the underlying system is likely to be polymorphous, and aggressive feature selection is likely to be counter-productive. Equally, we cannot reliably rank input features taken in isolation on their marginal contribution to predictability, for this necessarily ignores their interactions (Dramiński et al., 2008).\n\nIn short, models need to be complex enough to be tractable only with the highest capacity inferential architectures, such as the neurally inspired forms that have so rapidly grown to dominate the field of machine learning, notably in vision research. As in that case, this conclusion reveals two crucial problems, namely sensitivity to data scale and interpretability, both widely discussed in the literature (e.g. (Bzdok & Yeo, 2017)). Rather than rehearse the familiar difficulties they present, here we draw attention to a few unexpected possibilities they reveal.\n\n\nThe blessing of dimensionality\n\nWe have seen that complex, polymorphous systems require irreducibly many variables to achieve individually meaningful predictions. The resultant expansion of the parameter space under-determines models in proportion to the small scale of commonly available data. This familiar curse of dimensionality [25] makes good solutions hard to find and even harder to generalize, for the risk of purely accidental fits increases with the number of parameters.\n\nBut we should recognise that dimensionality also carries a blessing. Consider the parameterisations of contrasting dimensionality shown in Figure 1. Such individualisation as our low-dimensional parameterisation may achieve, here the inter-ocular distance, will be strongly dependent on the precision of measurement, for everyone is differentiated along a single dimension. In contrast with a high-dimensional parameterisation, such as a crudely pixelated rendition of the image, the precision of each individual variable is much less important, for the signal is conveyed in the covariance across variables. Crucially, since the structure of the underlying high-dimensional pattern is unlikely to resemble instrumental or other sources of noise, we can achieve greater individualisation with lower quality data. This is intuitively obvious in our ability to recognize faces from noisy, low-resolution images, robust not only to affine transforms of the data such as contrast, zoom, and skew, but also to fairly complex non-linear distortions.\n\nThe conventional resistance to using routinely acquired data on the grounds of noise and heterogeneity is only justified where the analysis is low dimensional. When measuring (say) total grey matter volume, it matters that one scanner will generate consistently greater estimates compared with another. But when extracting the high-dimensional variation of grey matter concentration across the brain, such effects will drop out as irrelevant affine shifts that leave the complex, individuating covariance patterns intact.\n\nPerhaps the most important objection to high-dimensional modelling—the scale of the data required—is thus addressable through collections for another purpose, obtained outside a research environment. In the domain of structural brain imaging, the obvious source is clinical imaging (Frackowiak & Markram, 2015). Since brain imaging is carried out to resolve diagnostic uncertainty towards normality almost as often as away from it, such data need not be restricted to the realm of pathology. Similarly, though smartphones may fall short of the precision of dedicated psychophysical devices, their ubiquity and critical mediating role in life enable the collection of rich, high-dimensional data on a vast scale (Teki et al., 2016).\n\nOf course, the correct balance between data size and data quality is an empirical question, to be settled case-by-case. But we cannot assume the former must be gated by the latter, and discount a high-dimensional approach simply because a conventional psychophysical laboratory cannot scale to thousands of participants. Rather, we must reconsider what we actually need to know, and what human activity may collaterally disclose it.\n\n\nLiving with opacity\n\nWhat use are high-dimensional models if they are too complex to understand? Where outcomes are highly variable, as is the norm in cognitive neurology, prediction is clinically invaluable, not simply because patients are consoled by accurate prognosis but because interventions need to be guided by their individually predicted responses. If a “black box” predictor is the best guide to a correct choice of treatment actuarially, it would be difficult to justify not following it merely because its operations cannot be paraphrased in intelligible prose.\n\nMoreover, clinical interventions are already primarily driven by “black boxes” - the contents of our heads are a good example. To give a reason for acting is not to specify a cause or to imply an underlying causal model, it is kin with pointing to a latent variable. It is only rarely, where very simple biological systems are concerned, never in the brain, that we have a perspicuous, mechanistic explanatory model available. That a human expert can cite a reason for his actions does not make his decision-making less opaque than that of a synthetic counterpart.\n\nIf a system requires a complex model to describe it, then it is complex. Translational science needs to adjust to this emotionally, not hopelessly attempt to change it intellectually. A causal field so intricate it can only be specified as an artificial neural network with a million parameters is explanatory, even if its incomprehensibility makes us hesitate to use a word stronger than predictive. We can no more hope to understand the brain shackled to simple, linear models, than a literary critic could hope to understand Shakespeare applying the basic rules of grammar alone.\n\n\nConcluding remarks\n\nUntil a decade ago, the foregoing analysis would have been unbearably nihilistic, for we had neither the data nor the computational tools to realise the alternative it urges. The ground is still new and uncertain, yet to be proven capable of supporting the structure we argue it is imperative we begin to erect on it. But if we wish to move beyond discussions of tractability or feasibility into translatable action, we must confront the single most striking fact about the brain - its immense complexity.\n\nThe difficulties are all the greater for being distributed across many intellectual, technological, even political domains, reaching deep into the foundations of the very notion of biological understanding. A cognitive neuroscience recast in the form we propose will have more in common with meteorology than horology. If so, then it will be because the fundamental nature of the brain has compelled it, for what we urge here above all is to let the data, not our own brains, speak first. And if effective prediction supplants defective understanding as a result, those outside the field, whose lives cognitive neuroscience and cognitive neurology seek ultimately to serve, will appreciate the exchange.\n\n\nData availability\n\nNo data is associated with this article.",
"appendix": "Competing interests\n\n\n\nNo competing interests were disclosed.\n\n\nGrant information\n\nPN is funded by a Health Innovation Challenge Fund award (HICF-R9-501, WT-103709) and the UCLH NIHR Biomedical Research Centre.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nAdolphs R: Human Lesion Studies in the 21st Century. Neuron. 2016; 90(6): 1151–1153. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBressler SL, Menon V: Large-scale brain networks in cognition: emerging methods and principles. Trends Cogn Sci. 2010; 14(6): 277–290. PubMed Abstract | Publisher Full Text\n\nBzdok D, Yeo BTT: Inference in the age of big data: Future perspectives on neuroscience. NeuroImage. 2017; 155: 549–564. PubMed Abstract | Publisher Full Text\n\nChang L, Tsao DY: The Code for Facial Identity in the Primate Brain. Cell. 2017; 169(6): 1013–1028.e14. PubMed Abstract | Publisher Full Text\n\nDramiński M, Rada-Iglesias A, Enroth S, et al.: Monte Carlo feature selection for supervised classification. Bioinformatics. 2008; 24(1): 110–117. PubMed Abstract | Publisher Full Text\n\nDwork C, Feldman V, Hardt M, et al.: STATISTICS. The reusable holdout: Preserving validity in adaptive data analysis. Science. 2015; 349(6248): 636–638. PubMed Abstract | Publisher Full Text\n\nFrackowiak R, Markram H: The future of human cerebral cartography: a novel approach. Philos Trans R Soc Lond B Biol Sci. 2015; 370(1668): pii: 20140171. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGal Y, Ghahramani Z: Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning. In PMLR. 2016; 1050–1059. [Accessed June 5, 2017]. Reference Source\n\nGhahramani Z: Probabilistic machine learning and artificial intelligence. Nature. 2015; 521(7553): 452–459. PubMed Abstract | Publisher Full Text\n\nGoodfellow I, Bengio Y, Courville A: Deep Learning. MIT Press, 2016. Reference Source\n\nHacker PMS: Human Nature: The Categorial Framework. John Wiley & Sons. 2007. Reference Source\n\nHawkins DM: The problem of overfitting. J Chem Inf Comput Sci. 2004; 44(1): 1–12. PubMed Abstract | Publisher Full Text\n\nHuttenlocher PR, Dabholkar AS: Regional differences in synaptogenesis in human cerebral cortex. J Comp Neurol. 1997; 387(2): 167–178. PubMed Abstract | Publisher Full Text\n\nLe Cun Y, Denker JS, Solla SA: Optimal Brain Damage. In Proceedings of the 2Nd International Conference on Neural Information Processing Systems. NIPS’ 89. (Cambridge, MA USA: MIT Press), 1989; 598–605. Reference Source\n\nLeCun Y, Bengio Y, Hinton G: Deep learning. Nature. 2015; 521(7553): 436–444. PubMed Abstract | Publisher Full Text\n\nLossi L, Merighi A: In vivo cellular and molecular mechanisms of neuronal apoptosis in the mammalian CNS. Prog Neurobiol. 2003; 69(5): 287–312. PubMed Abstract | Publisher Full Text\n\nMackie JL: The Cement of the Universe. Oxford, Clarendon Press, 1974. Reference Source\n\nMarder E, Bucher D: Understanding circuit dynamics using the stomatogastric nervous system of lobsters and crabs. Annu Rev Physiol. 2007; 69: 291–316. PubMed Abstract | Publisher Full Text\n\nMatsuoka K: Noise injection into inputs in back-propagation learning. IEEE Transactions on Systems, Man, and Cybernetics. 1992; 22(3): 436–440. Publisher Full Text\n\nMcMahan HB, Moore E, Ramage D, et al.: Communication-Efficient Learning of Deep Networks from Decentralized Data. arXiv: 1602.05629 [cs]. 2016. [Accessed May 13, 2018]. Reference Source\n\nParkhi OM, Vedaldi A, Zisserman A: Deep Face Recognition. In BMVC. 2015; 6. Publisher Full Text\n\nQureshi AI, Mendelow AD, Hanley DF: Intracerebral haemorrhage. Lancet. 2009; 373(9675): 1632–1644. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRands CM, Meader S, Ponting CP, et al.: 8.2% of the Human genome is constrained: variation in rates of turnover across functional element classes in the human lineage. PLoS Genet. 2014; 10(7): e1004525. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRissanen J: Modeling by shortest data description. Automatica. 1978; 14(5): 465–471. Publisher Full Text\n\nRorden C, Karnath HO: Using human brain lesions to infer function: a relic from a past era in the fMRI age? Nat Rev Neurosci. 2004; 5(10): 813–819. PubMed Abstract | Publisher Full Text\n\nSchroff F, Kalenichenko D, Philbin J: FaceNet: A Unified Embedding for Face Recognition and Clustering. 2015; 815–823. [Accessed December 7, 2016]. Publisher Full Text\n\nSrivastava N, Hinton G, Krizhevsky A, et al.: Dropout: A Simple Way to Prevent Neural Networks from Overfitting. J Mach Learn Res. 2014; 15: 1929–1958. Reference Source\n\nTang Y, Nyengaard JR, De Groot DM, et al.: Total regional and global number of synapses in the human brain neocortex. Synapse. 2001; 41(3): 258–273. PubMed Abstract | Publisher Full Text\n\nTeki S, Kumar S, Griffiths TD: Large-Scale Analysis of Auditory Segregation Behavior Crowdsourced via a Smartphone App. PLoS One. 2016; 11(4): e0153916. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVapnik V: Statistical learning theory. Wiley. 1998. Reference Source\n\nWan L, Zeiler M, Zhang S, et al.: Regularization of Neural Networks using DropConnect. 2013; 1058–1066. [Accessed June 5, 2017]. Reference Source\n\nWardlaw JM, Warlow CP, Counsell C: Systematic review of evidence on thrombolytic therapy for acute ischaemic stroke. Lancet. 1997; 350(9078): 607–614. PubMed Abstract | Publisher Full Text\n\nZhou E, Cao Z, Yin Q: Naive-Deep Face Recognition: Touching the Limit of LFW Benchmark or Not? arXiv: 1501.04690 [cs]. 2015. Reference Source"
}
|
[
{
"id": "34258",
"date": "19 Jun 2018",
"name": "Michel Thiebaut de Schotten",
"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\nIn their excellent and timely contribution Nachev, Rees and Frackowiak tackle current limitation in the models used to understand the brain functioning and the translation of this knowledge to the clinical practice. The text is engaging, and the message is clear. The authors did not limit their focus on the current problems but also provide clear new solutions and recommendations for future generations.\n\nMay I suggest the part on Dimensionality and individualisation to be linked up to a recent editorial entitled 'is a *single* brain model sufficient' (Thiebaut de Schotten and Shallice Cortex 2017)1. I think this is appropriate but I leave it as optional for the authors.\n\nAgain, thank you for this elegant contribution soon to become a classic in the field.\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": "4326",
"date": "02 Jan 2019",
"name": "Parashkev Nachev",
"role": "Author Response",
"response": "We thank Professor Thiebaut de Schotten for his appreciative comments, and for the reference now cited in the revised version."
}
]
},
{
"id": "35675",
"date": "23 Aug 2018",
"name": "Tor D. Wager",
"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\nIn “Lost in Translation”, Nachev et al. provide a bold and thought-provoking piece on the past and future of cognitive neuroscience. Their thesis is that the classic inferential tools of neuropsychology, which also formed some of the foundational building blocks of cognitive neuroscience, are inadequate to model and understand the complexity of the brain. New paradigms are needed, and Nachev et al. offer a fresh, theoretical argument for why we should embrace new concepts of distributed causality, nonlinearity, and models that move beyond averaging over individuals to better capture inter-individual variation. The implication is that only by embracing such models can cognitive neuroscience, and perhaps neuroimaging in particular, develop models that are accurate enough to predict individual performance and clinical status - in short, to be translationally useful.\nNachev et al.’s position has much to recommend it. One anchor point in their argument is that the brain processes that drive (cause) feeling and behavior are distributed across neurons and/or brain systems—and, if this is the case, then models and measures with coarse, distributed features (even if noisy) will outperform those with only a few, high-precision features. I resonate with this point. There is substantial evidence that the neural representations underlying multiple forms of cognition, emotion, and action are population codes distributed across large numbers of neurons and (in the case of fMRI) brain regions and systems (e.g., for review, see Kragel et al. 20181). In fMRI studies, distributed predictive models that include activity across regions and systems can dramatically outperform those based on even the best single brain regions (for recent reviews see, e.g., Arbabshirani et al. 20172; Bzdok & Meyer-Lindenberg 20173; Woo et al. 20174; Kragel et al. 2018, Figure 3 and text).\n\nUnits of analysis\nThe essential question when it comes to brain systems is “What are the units of analysis”? The authors imply that translational failures come from a localisationist approach, and provide a thought-provoking theoretical argument for why the brain - at least with respect to cognition, feeling, and behavior - ought to be treated as a system with broadly distributed causality. I agree in part. Lesioning isolated nodes in a neural network typically does not catastrophically, or selectively, impair performance largely because the single node (or neuron) is not the relevant unit of analysis. But if one were to lesion a layer, particularly a layer dedicated to a particular function as in the structured, brain-inspired networks of O’Reilly et al. (Aisa et al. 20085; O’Reilly et al. 20176), the effects on network behavior would be profound. Recent advances in opto- and chemogenetics allow for the targeting and activation/inactivation of distributed sets of neurons that collectively represent particular cognitions and actions (e.g., Ramirez et al. 20137), with strong effects on behavior. Likewise, when studying ecosystems, the loss of individual organisms selected at random has little effect on the behavior of the system as a whole; but loss of a species can have a profound effect. The species, but not the individual, can be characterized as having necessary and sufficient roles in the system’s behavior. Likewise, individual neurons are likely not necessary or sufficient for anything, but neural populations are.\n\nA related point is Nachev et al.’s critique of averaging over individuals, which also hinges on the issue of which units of analysis are averaged over. Nachev et al.’s “watch” example is an interesting case study. They show that averaging over visual images of watch mechanisms does not elucidate the nature of watches or produce anything like any of the individual watches. But the problem is not averaging per se - it is knowing what to average over. Pixels in an image of the watch mechanisms are simply not the right unit of analysis, so averaging over them is meaningless. However, the average watch has 2 gear wheels, 2 hands, and one battery; averaging over or otherwise characterizing the distributions over these properties makes sense.\nPerhaps we will discover that brain voxels are not the right features to average over, either (and I suspect that they are not!). Cognitive neuroscience converges with machine learning in that a big part of the endeavor is, and has always been, discovering the units (or features) and level of analysis that confer maximal ability to understand the mind and predict future behavior.\n\nMonomorphous and polymorphous\nAnother interesting contrast that Nachev et al. make is the distinction between monomorphous populations, whose individuals are identical, and polymorphous populations, whose individuals vary. Their central argument is that because human brains are polymorphous, we should not characterize them using averages across individuals; rather, we should focus on more individualized models.\n\nClearly, humans are a polymorphous bunch. But does this mean that population-level studies that characterize averages - or, alternatively, develop multivariate predictive models of behavior across individuals - are useless? Really, the brain is both monomorphous and polymorphous, at different levels of analysis. Identifying patterns of commonality does not mean that all variation is noise. If I were trying to describe to a space-faring alien what a car looks like, I would not assume that all cars are identical and the difference between a Tesla and a Toyota is “noise”. But neither would I assume that every car is completely different, which would preclude any sort of common description at all. The brain is similar in this respect. Virtually all of us have an occipital lobe, which contains a primary visual cortex. We have a primary motor cortex, a hippocampus, an inferior frontal junction, each of which plays consistent roles in behavior across individuals. In my lab’s work, we find that systems that track and predict the intensity of evoked pain experience are very similar across individuals - for example, the same brain pattern responds to painful events to some degree in 95% of participants (the 5% might well be largely noise; see Zunhammer et al., N = 603). But this does not mean that individual differences are unimportant! Rather, these baseline commonalities are a launching point for understanding the ‘variations on a theme’ that make individuals different from one another.\n\nLost and found\nA final reflection: It is true that cognitive neuroscience has not developed many translational applications that are used clinically or commercially (e.g., Woo et al. 2017). But this does not necessarily imply that cognitive neuroscience has failed so far. We should remember that the goal of cognitive neuroscience has been to understand the physiological representation of thought and behavior - one of the thorniest challenges in all of science, and a basic (not clinical) goal at that. Like many forms of basic science, the hope is that by better understanding how the mind works, without tying it immediately to any commercial venture or practical application, will yield new ways of thinking about the brain and mind, which in turn will inspire future applications that were previously unimagined. In my own lab’s work on pain and emotion, I have learned that the gaps between science and commercial application are not just about limitations in the science. Even if current cognitive neuroscience-based models could reliably diagnose mental and brain health conditions in individual people with perfect sensitivity and specificity, there would be gaps related to business development, marketing, public understanding and policy, economic cost/benefit ratios, equal access, insurance reimbursements, and more. I believe that in the past 2-3 years, cognitive and clinical neuroscience has succeeded in developing models of brain function that could be useful for characterizing dementia, depression, pain, autism, and more. Their clinical and commercial success will depend largely on what society wants to do with the science.\nThis does not, of course, take away from the point that translation is a worthy and useful goal, both from a humanitarian (e.g., Gabrieli et al. 20158) and scientific perspective. Not only can it advance clinical applications, translation also provides a concrete, objective yardstick against which to evaluate our understanding of the brain. Clearly, we have a great distance to go; the good news is that we're moving forward.\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? Partly\n\nAre the conclusions drawn balanced and justified on the basis of the presented arguments? Partly",
"responses": [
{
"c_id": "4325",
"date": "02 Jan 2019",
"name": "Parashkev Nachev",
"role": "Author Response",
"response": "We are grateful to Professor Wager for his thoughtful commentary on our paper, and are pleased he finds sympathy with the broad thrust of our argument. Our revision clarifies the points he raises, and we provide a few further thoughts below.Units of analysisWe agree there is no warrant for presuming a neural system must be intelligible at a neuroscientist's chosen pet scale of analysis. It is essential to explore a wide variety of scales, including those intelligible only through high-dimensional modelling, where—as we argue in detail in the paper—the critical organisation is mostly likely to be found. But it is also true that no single scale of analysis need be sufficient, for the organisation may be distributed across a multiplicity of them. Indeed, the kind of compact causal structure biology is likely to favour—for the information theoretic reasons we give—is constitutionally widely distributed. Of course, necessity and sufficiency is easily obtained at a whole brain level of analysis, but such causality is merely permissive rather than properly explanatory of how thought and behaviour really arise. Might there be some isolated \"intermediate\" level where causality is plain yet the underlying relations are revealed in sufficient detail? We doubt it, but the only way to tell is through the high-dimensional modelling approach we describe in the second half of the paper.AveragingIt is a criterion both of general explanation—to the satisfaction of the scientist—and individual prediction—to the satisfaction of the clinician—that one can generalise from known instances to unknown ones. In both cases a model learns from one set of data (even if it is not called learning in the conventional statistical literature) and extrapolates to another. When we criticise averaging, our target is specifically low-dimensional models whose form is intuitively specified. In a high-dimensional, necessarily data-driven model, averaging occurs too, but across a local neighbourhood, not across an arbitrarily defined group as a whole. And that neighbourhood is determined by the modelling process itself, guided by out-of-sample predictive performance rather than model statistics. The point is illustrated by the clock mechanism example given in Figure 3. TranslationWe are clinicians, and for us the value of thought is measured by its utility in action. Others may, of course, adopt a different perspective. But consider what the ultimate criterion of fidelity must always be. A neuroscientist may claim his beautifully perspicuous theory of the function of some aspect of the brain is explanatory because it fits a model with half a dozen dimensions derived from fifty people. But what do we tell the patients—often the majority—the model does not fit? That they are noise? This would be an acceptable excuse only if the predictive power of more complex models remains poor, and the unmodelled variance is indeed plausibly noise. Once a model with better—and of course generalisable—predictive power is found, the claim is fatally undermined, for what was previously discarded as noise is now explained. That the resultant explanation might be complex need not be to its detriment: generalisable individual predictive performance always trumps perspicuity. The primary concern of the clinician ought to be the scientist's too."
}
]
}
] | 1
|
https://f1000research.com/articles/7-620
|
https://f1000research.com/articles/7-1666/v1
|
19 Oct 18
|
{
"type": "Research Article",
"title": "Functional proteomic profiling reveals KLK13 and TMPRSS11D as active proteases in the lower female reproductive tract",
"authors": [
"Carla M.J. Muytjens",
"Yijing Yu",
"Eleftherios P. Diamandis",
"Carla M.J. Muytjens",
"Yijing Yu"
],
"abstract": "Background: Cervical-vaginal fluid (CVF) hydrates the mucosa of the lower female reproductive tract and is known to contain numerous proteases. The low pH of CVF (4.5 or below in healthy women of reproductive age) is a uniquely human attribute and poses a challenge for the proteolytic functioning of the proteases identified in this complex biological fluid. Despite the abundance of certain proteases in CVF, the proteolytic activity and function of proteases in CVF is not well characterized. Methods: In the present study, we employed fluorogenic substrate screening to investigate the influence of pH and inhibitory compounds on the proteolytic activity in CVF. Activity-based probe (ABP) proteomics has evolved as a powerful tool to investigate active proteases within complex proteomes and a trypsin-specific ABP was used to identify active proteases in CVF. Results: Serine proteases are among the most abundant proteins in the CVF proteome. Labeling human CVF samples with the trypsin-specific ABP revealed serine proteases transmembrane protein serine 11D and kallikrein-related peptidase 13 as active proteases in CVF. Furthermore, we demonstrated that the proteolytic activity in CVF is highly pH-dependent with an almost absolute inhibition of trypsin-like proteolytic activity at physiological pH levels. Conclusions: These findings provide a framework to understand proteolytic activity in CVF. Furthermore, the present results provide clues for a novel regulatory mechanism in which fluctuations in CVF pH have the potential to control the catalytic activity in the lower female reproductive tract.",
"keywords": [
"cervical-vaginal fluid",
"serine proteases",
"activtiy-based probe proteomics"
],
"content": "Introduction\n\nApproximately 2–4% of the genes in the human genome account for proteases, and in addition, over 150 different protease inhibitors are encoded in the human genome1. Proteases irreversibly modify proteins via peptide bond hydrolysis and thereby impact on both protein structure and function. As a result, proteases are implicated in nearly every significant biological process.\n\nCervical-vaginal fluid (CVF) hydrates the mucosa of the vagina and lower part of the cervix. Situated directly at the interface between the external environment and the internal milieu, CVF plays key roles in sexual intercourse, conception and provides a barrier against microbial invasion2–6. Almost 8% of all proteins identified in CVF are proteases, which highlights the tremendous proteolytic potential of this complex biological fluid7. Interestingly, human CVF differs from other mammals in that it is uniquely acidic with levels of 4.5 or less in healthy women of reproductive age8,9. The low pH is estrogen dependent and results from the conversion of glycogen to lactic acid by the vaginal microbiome or, alternatively, the vaginal lumen is acidified via active proton secretion across the vaginal epithelial membrane10–12. The acidity levels in CVF pose a substantial challenge for protease activity in this environment. Despite the high abundance of proteases in CVF, their activity and functioning in the lower female reproductive tract remains to be elucidated.\n\nFunctional proteomics using an activity-based probe (ABP) coupled to mass spectrometry has emerged as a powerful tool to study active proteases within complex proteomes13,14. ABPs, designed to target certain classes of active proteases, have been successfully applied to investigate ABP-mediated profiling of complex proteomes, inhibitor screening and enzymatic target identification15,16.\n\nIn the current study, ABP profiling of CVF resulted in the identification of transmembrane protein serine 11D (TMPRSS11D) and kallikrein-related peptidase 13 (KLK13) as active trypsin-like proteases. In addition, we observed a near complete cessation of trypsin-like proteolytic activity at physiological pH levels in CVF. Therefore, we propose a novel mechanism in which fluctuations in CVF pH have the potential to regulate the catalytic activity in the lower female reproductive tract. Understanding the protease activity and functioning in CVF can provide important insights in their mechanism of action and ultimately lead to novel protease targeted therapeutics.\n\n\nMethods\n\nCVF samples were collected from 10 healthy, non-pregnant female volunteers using a softcup collection device (Instead Inc. San Diego, CA) as described previously7,17. Informed consent was obtained from all participants prior to sample collection. The collection and use of human samples was approved by the Research Ethic Board at Mount Sinai Hospital in Toronto (MSH REB 16-0137-E). All women abstained from sexual activity and douching at least 3 days prior to sample collection and had no self-reported vaginal infections. The sample was collected after 90 minutes via centrifugation (Eppendorf centrifuge, model 5415R) of the softcup inside a 50mL conical tube at 200xg for 2 minutes. CVF was transferred to a 1.5mL tube and centrifuged at 15,000xg for 20 minutes to remove cellular debris. Total protein of each CVF sample was determined using a Bradford protein assay (Thermo Fisher Scientific, #1856209)). The samples were stored at -80°C until further use.\n\nThe estimation of protease abundance in the CVF proteome was based on the CVF proteome dataset from healthy non-pregnant women7. Proteases were identified by comparison with the Merops database and grouped according to their catalytic type1. The log transformed MS1 area data were used as a proxy for the relative abundance of each protein within the dataset. CVF proteins and proteases were subsequently ranked according to the relative abundance and plots were produced using R (R version 3.5.0).\n\nThe trypsin-like proteolytic activity in CVF was measured by incubating 0.1µg total protein CVF sample with 0.5mM of the fluorogenic t-butoxycarbonyl-tri-peptide-7amino-4-methylcoumarin (AMC) synthetic peptide (Bachem, #4003460) with a valine (V), proline (P), arginine (R) tripeptide sequence. Fluorescence was measured every minute for 15 minutes on a Wallac Envision fluorometer (Perkin-Elmer) set at 355nm for excitation and 460nm for emission. All reactions were performed at 37°C in female reproductive tract buffer (20mM potassium phosphate, 60mM sodium chloride)18. To determine the influence of pH on the proteolytic activity in CVF the buffers were adjusted using NaOH and HCl so that pH ranged from 9.0 to 3.5 in 0.5 increments. Trypsin-like proteolytic activity in CVF was also measured in the presence of 1 mM phenylmethylsulfonyl fluoride (PMSF), 10mM ethylenediaminetetraacetic acid (EDTA) or 5nM soybean trypsin inhibitor (STI). Proteolytic activity was quantified by calculating the rate of substrate hydrolysis after subtraction of the fluorescence of enzyme-free reactions. All experiments were performed in triplicate on 2 separate days. Figures were prepared using GraphPad Prism (version 7.00 for Windows).\n\nA trypsin-specific, biotin-conjugated activity-based probe (ABP) was used to label the active enzymes in CVF specifically and covalently19. The probe was kindly provided by Professor Brendon Gilmore (Queen’s University Belfast, Ireland). Active enzymes were labeled by incubating 1 µg total protein of CVF sample in 5 µL (in 10mM Tris, pH 8.0) with 1 µL 10x proteinase assay buffer (500mM Tris-HCl, pH8.0, 2% NP40, 15mM CaCl2), 1 µL of 1 mM ABP and 2 µL H2O. Additionally, CVF sample was incubated with 10mM EDTA, 1mM PMSF and 5nM STI prior to the labeling of the active enzymes with the ABP. After a 2-hour incubation at room temperature, the reaction was stopped using SDS sample buffer (31.25mM Tris-HCl pH 6.8, 12.5% glycerol, 1% SDS, 0.005% Bromphenol Blue (Biorad, #161-0747) containing 50mM DTT, followed by boiling for 10 minutes to denature the proteins. Samples were loaded on a 4–15% Mini-Protean TGX precast gel (Biorad, #456-1086/1083) and SDS gel electrophoresis was performed at 200V for approximately 30 minutes. Proteins were transferred using a 0.2 µm PVDF membrane (BioRad, #1704156) in a Trans-blot Turbo Blotting System (Biorad). The membranes were blocked with 5% casein in TBS-T (0.05% Tween-20 in 50mM Tris, 150 mM NaCl, pH 7.5) at room temperature for 2 hours, followed by a wash with TBS-T (3 × 10 minutes). The membrane was subsequently incubated with streptavidin-horseradish peroxidase (SA-HRP) (Jackson Immunoresearch, #016-030-084) in a 1:10,000 dilution in 1% casein in TBST-T for an hour, followed by an extensive wash with TBS-T (3 × 15 minutes, 3 × 5 minutes). A total of 125 µL of chemiluminescence substrate (GE healthcare, #45000875) was added per square centimeter of membrane. The membrane was then placed into an autoradiography cassette, exposed for 6 minutes and developed using x-ray film.\n\nCVF samples (N= 5) were fractionated using size exclusion chromatography (SEC) in order to isolate the fractions with demonstrated proteolytic activity. A total of 250µg total protein per sample was loaded onto a TSK gel G3000SW column (Tosoh Bioscience) in 0.1M Na2HPO4/Na2SO4 (pH 6.8) and fractionated with a constant flow rate of 0.5mL/min. Fractions were collected every minute throughout the gradient and subsequently analyzed for proteolytic activity using the fluorescent substrate and ABP as described previously. The trypsin-like proteolytic activity in the SEC fractions was quantified via comparison to standard curve prepared with porcine trypsin (Sigma, 16,000 BAEE U/mg, #T0303).\n\nThe fractionated CVF samples were prepared for gel electrophoresis by intermixing with SDS sample buffer followed by heating to 95°C in order to denature the proteins. Samples were loaded on a 4–15% Mini-Protean TGX precast gel (Biorad, #456-1086/1083) and SDS gel electrophoresis was performed at 200V. After completion of gel electrophoresis, the proteins in the gel were visualized using a Coomassie stain (Biorad, #1610787) and bands matching the protease-probe complexes were excised and prepared for mass spectrometry. The samples were concentrated with Omix C18MB tips (Agilent Technologies, #A5700310K) and eluted with 5µL buffer B (0.1% formic acid (FA), 65% acetonitrile (ACN)). To each eluted sample, 60µL of buffer A (0.1% FA, water) was added, of which 18 µL were loaded from a 96-well microplate autosampler onto a C18 trap column using the EASY-nLC1000 system (Thermo Fisher Scientific, #LC120) and running buffer C (0.1 % FA in water). The trap column consisted of IntegraFrit capillary (inner diameter 150 µm, New Objective, #IF15010xxxx) cut to 3.2 cm in length and packed in-house with 5µm Pursuit C-18 (Agilent Technologies). Peptides were eluted from the trap column at 300nL/min with an increasing concentration of buffer D (0.1% FA in ACN) over a 30-minute gradient onto a resolving 15 cm long PicoTip Emitter (75µm inner diameter, 8µm tip, New Objective) packed in-house with 3µm Pursuit C-18 (Agilent Technologies). The liquid chromatography setup was coupled online to a Q Exactive Plus (Thermo Fisher Scientific) mass spectrometer using a nanoelectrospray ionization source (Thermo Fisher Scientific) with capillary temperature set to 275°C and spray voltage of 2kV. A 60-minute data-dependant acquisition (DDA) method was setup on the Q Exactive Plus. The full MS1 scan from 400–1500 m/z was acquired in the Orbitrap at a resolution of 70,000 in profile mode with subsequent fragmentation of top 12 parent ions using the HCD cell and detection of fragment ions in the Orbitrap using centroid mode at a resolution of 17,500. The following MS method parameters were used: MS1 Automatic Gain Control (AGC) target was set to 3e6 with maximum injection time (IT) of 100ms, MS2 AGC was set to 1e5 with maximum IT of 50ms, isolation window was 1.6 Da, underfill ratio 2%, intensity threshold 4e4, normalized collision energy was set to 27, charge exclusion was set to fragment only 2+,3+ and 4+ charge state ions, apex trigger was deactivated, peptide match set to preferred and dynamic exclusion set to 45 seconds.\n\nThe mass spectrometry RAW files were uploaded into MaxQuant ver. 1.5.2.8 and searched with built-in Andromeda search engine20,21. The search was performed with oxidation of methionine and N-terminal protein acetylation as variable modifications, carbamidomethylation of cysteine as a fixed modification, trypsin as the digestive enzyme with a maximum of 2 allowed missed cleavages. The first peptide search tolerance for both samples was 20ppm against a small ‘human first-search’ database for the purpose of mass recalibration, whereas the main search was performed at 4.5 ppm against the Human SwissProt protein database (January 2015 release). The database was reversed to calculate the peptide and protein level false-discovery rate (set at 1%). Proteases were selected from the protein list by comparing the dataset to the Merops database. Candidate active proteases were subsequently selected based on relative protein abundance within the sample, catalytic type, molecular size of the mature protein and substrate specificity.\n\nValidation of TMPRSS11D was performed by preparing 10µg total protein unfractionated CVF sample and SEC fractionated CVF samples for gel electrophoresis as described previously. A total of 20µg HepG2 cell lysate (abcam, #ab7900) was included as a positive control as per manufacturer’s instructions. Following gel electrophoresis and transfer of the proteins to a PDVF membrane, the membrane was blocked in 5% milk in TBS-T at 4°C overnight. The membrane was probed with anti-TMPRSS11D antibody (ab127031, abcam, RRID:AB_11129135) diluted 1 in 500 in 1% milk in TBS-T for 90 minutes at room temperature. Subsequently, the membrane was washed 3 times for 10 minutes with TBS-T and incubated with peroxidase conjugated goat anti rabbit secondary antibody (Jackson Immunoresearch, #111-035-045) diluted 1 in 3000 in 1% milk in TBS-T for 45 minutes. The membrane was washed 4 times for 10 minutes with TBS-T prior to incubation with chemiluminescence substrate, exposure and development.\n\nRecombinant active KLK13 was produced using a Pichia pastoris yeast expression system. Briefly, a PCR-amplified DNA fragment encoding the mature KLK13 isoform (36-277 aa) was cloned into the pPIC9 expression vector, in-frame with its α-secretion signal and the alcohol oxidase AOX1 gene. Purified mature KLK13 pPIC9 DNA construct was confirmed by sequencing using the 5’-AOX1, 3’AOX1 and the α-secretion signal vector-specific primers and the NCBI BLAST align program. The KLK13-pPIC9 construct was linearized with SacI enzyme and transformed into the KM71 P. pastoris strain via electroporation. A stable KM71 transformant was grown in 1 liter of buffered glycerol-complex medium (BMGY) media. After 2 days, yeast culture was centrifuged, and the cell pellet was resuspended in 300 ml of BMMY media (A600 = 10). Recombinant KLK13 expression was induced with 1% methanol for 5 days at 30°C in a shaking incubator (250 rpm). Recombinant KLK13 was purified from culture supernatant by ultraconcentration, serial dialysis, and centrifugation procedures, followed by cation-exchange chromatography using an automated ÄKTA FPLC system on a pre-equilibrated 5-ml cation-exchange HiTrap high performance Sepharose HP-SP column (GE Healthcare). Recombinant KLK13 was further purified using 10mL cation-exchange Source15S TricornTM column (GE Healthcare). The collected fractions were pooled, concentrated, analyzed by SDS-PAGE and Western blotting and stored at −80°C until further use. The concentration of purified protein was measured using a Coomassie protein assay and KLK13 ELISA as described previously22. The N-termini of the bands identified on SDS-PAGE gels were analyzed with N-terminal Edman sequencing.\n\nUnfractionated CVF and SEC fractionated CVF samples were prepared for gel electrophoresis as described previously. A total of 1µg in-house produced recombinant KLK13 was included as a positive control. Following gel electrophoresis and protein transfer, the membrane was blocked in 5% milk in TBS-T at 4°C overnight. The membrane was probed with a in-house produced monoclonal anti-KLK13 antibody diluted 1 in 50 in 1% milk in TBS-T for 90 minutes at room temperature. Subsequently, the membrane was washed 3 times for 10 minutes with TBS-T and incubated with peroxidase conjugated goat anti mouse secondary antibody (Jackson Immunoresearch, #115-035-146) 1 in 3000 diluted in 1% milk in TBS-T for 45 minutes. The membrane was washed 4 times for 10 minutes with TBS-T prior to incubation with chemiluminescence substrate (GE healthcare, #45000875), exposure and development.\n\nAn immunocapture method was developed to confirm the presence and activity of KLK13 in CVF. In this assay, 500ng of in-house produced monoclonal KLK13 antibody was coated overnight on a 96-well plate in coating buffer (50mM Tris, 0.05% Tween-20, pH 7.8). The plate was washed 3 times with washing buffer (50mM Tris, 150mM NaCl, 0.05% Tween-20, pH 7.8) and blocked with 1% bovine serum albumin (BSA) in activity buffer (100mM NaH2PO4/Na2HPO4, pH 8.5) for 90 minutes. The plate was washed 3 times with washing buffer prior to incubation with CVF samples (N=3) in a 1 in 5, 1 in 10 and 1 in 50 dilutions for 3 hours at room temperature. Activity buffer without active enzyme, IgG coated wells, and 6nM in-house produced recombinant KLK13 were included as a negative and positive controls respectively. The wells were washed extensively 6 times with washing buffer to remove any unbound protein. A final 0.5mM VPR-AMC substrate (Bachem, #4003460) was added to the wells and the resulting increase in fluorescence was measured in 1-minute intervals on a Wallac Envision fluorometer set at 355nm for excitation and 460nm for emission. Proteolytic KLK13 activity in CVF was semi-quantified by comparing the rate of substrate hydrolysis after subtraction of the fluorescence of enzyme-free reactions to the recombinant KLK13 activity.\n\nThe influence of pH on the proteolytic activity of KLK13 was tested by measuring the cleavage of fluorogenic substrate by 6nM recombinant KLK13 in buffer with pH ranging from 3.5 to 10.0 in 0.5 increments. Fluorescence was measured every minute for 15 minutes on a Wallac Envision fluorometer (Perkin-Elmer) set at 355nm for excitation and 460nm for emission. Proteolytic activity was quantified by calculating the rate of substrate hydrolysis after subtraction of the fluorescence of enzyme-free reactions relative to proteolytic activity at pH 8.0. All reactions were performed at 37°C in triplicate on 2 separate days.\n\n\nResults\n\nA total of 85 proteases were identified in the CVF proteome based on samples from healthy, non-pregnant women of reproductive age7. As shown in Figure 1, proteases of different catalytic classes are present in the entire relative abundance spectrum of the CVF proteome. Multiple serine proteases rank among the most abundant proteins in CVF. As can be seen in Figure 2A, the trypsin-like proteolytic activity in CVF is optimal at slightly alkaline pH levels (pH 7.5-9). Almost half of the maximal trypsin-like proteolytic activity is CVF remains at pH 6.0, whereas there is negligible trypsin-like proteolytic activity at pH 4.5 or below. The presence of PMSF and STI significantly reduced the trypsin-like proteolytic activity in CVF (p<0.05), whereas EDTA did not impact the proteolytic activity in CVF significantly (Figure 2B).\n\nVisualization of the ranked relative abundance of all proteins in CVF (1,065 proteins, grey dots) highlighting serine proteases (36 proteins, red dots), cysteine proteases (13 proteins, blue dots), threonine proteases (13 proteins, green dots) metalloproteinases (10 proteins, yellow dots). Aspartic acid proteases (1 protein), mixed catalytic type proteases (3 proteins) and unassigned proteases (4 proteins) were also identified in CVF.\n\n(A) Trypsin-like proteolytic activity in CVF is highly pH dependent as demonstrated by the cleavage of fluorogenic trypsin-like VPR-AMC substrate by proteases in CVF. Grey areas represent the CVF pH during infections and post-menopausal status (pH 5–6.5) and after sperm ejaculation (pH 7.2) (B) Trypsin-like proteolytic activity is significantly inhibited by PMSF and STI, whereas EDTA has no significant effect.\n\nActive proteases present in CVF from healthy, non-pregnant women of reproductive age were visualized using the ABP probe, which binds trypsin-like active proteases in a covalent and specific manner19. As illustrated in Figure 3A, trypsin-like protease activity in CVF is present at 2 separate molecular weight (MW) bands at approximately 28 and 35kDa. Apart from differences in the intensity of the bands, the observed expression pattern was nearly identical among the different CVF samples. Next, we investigated whether the trypsin-like protease activity in CVF can be attributed to proteases of certain catalytic types. Complete cessation of trypsin-like activity in CVF was accomplished by incubation of the CVF samples with the irreversible protease inhibitor PMSF, targeting both serine and cysteine proteases, prior to coupling to the ABP (Figure 3B). Additionally, incubation of CVF with reversible serine protease inhibitor STI resulted in a near complete inhibition of the protease activity in CVF compared to the uninhibited control condition (Figure 3C). Inhibition of metalloproteinases via incubation with EDTA did not affect the trypsin-like proteolytic activity in CVF (Figure 3B). As expected, no active proteases were detected in the control condition without probe addition (Figure 3D).\n\n(A) Active trypsin-like proteases in CVF (N=10) were coupled to the biotinylated ABP probe and visualized following gel electrophoresis. Trypsin-like proteolytic activity was present in two separate bands (28kDa and 35kDa) in each of the samples. (B) Co-incubation of CVF with PMSF prior to coupling to the ABP probe resulted in the inhibition of active trypsin-like proteolytic activity, whereas co-incubation with EDTA did not effect active trypsin-like proteolytic activity (C) Incubation with STI prior to labeling CVF with the ABP probe inhibited the trypsin-like proteolytic activity in CVF compared to the control condition (D) Negative control condition incubated without ABP probe showed no trypsin-like proteolytic activity in experimental samples.\n\nSEC fractionation of CVF resulted in the successful separation of the high and low MW proteases as demonstrated by the split in proteolytic activity measured in fractionated CVF using the fluorogenic VPR-AMC substrate (Figure 4A) and subsequent visualization of active proteases in CVF using the ABP (Figure 4B). Both the high and low MW bands were excised and resolved separately with mass spectrometry resulting in the identification of approximately 100 proteins for the high MW and low MW bands (Figure 4C–D). Candidate active proteases were selected based on relative protein abundance among the sample proteome, serine proteolytic catalytic type, correct molecular size and trypsin-like substrate specificity. This approached resulted in the identification of KLK13 as a candidate protease for the high MW band and TMPRSS11D as the candidate active protease in the low MW band.\n\n(A) trypsin-like activity was quantified in fractions F36 – F46 in SEC fractionated CVF using VPR-AMC substrate (B) Visualization of separation of high (F36-F38) and low (F42-F46) MW band containing trypsin-like proteolytic activity in SEC fractionated CVF (C) Ranked relative abundance of proteases (red dots) in high MW fractions across CVF proteome (D) Ranked relative abundance of proteases (red dots) in low MW fractions across CVF proteome. For more comments see text.\n\nA Western blot showed TMPRSS11D at the correct MW in unfractionated CVF samples (Figure 5). Furthermore, TMPRSS11D is detected in the CVF fractions corresponding to the low MW band as expected.\n\nTMPRSS11D is detected in unfractionated CVF and in F42-F46 of SEC fractionated CVF corresponding with the proteolytic activity visualized with the ABP in the low MW band. HepG2 cell lysate is included as a positive control.\n\nMature, active recombinant KLK13 was produced in a P. pastoris expression system. The KLK13 750bp PCR construct was successfully ligated into a eukaryotic pPIC9 vector as confirmed by DNA sequencing. The pPIC9-klk13 vector was subsequently linearized by SacI enzyme digestion and electroporated into P. pastoris KM71 cells. Gel electrophoresis of the KLK13 protein, isolated from the methanol induced yeast cell culture medium in a multi-step purification and concentration process, resulted in the visualization of KLK13 (Figure 6A). PNGaseF treatment revealed that the higher MW band is the result of glycosylation at 225NRT. (Figure 6B). KLK13 identity was confirmed by mass spectrometry analysis.\n\n(A) Silver staining of reduced SDS-PAGE of KLK13 following purification with band visible at 30kDa. The band was found to contain KLK13 sequences by mass spectrometry and recognized by anti-KLK13 antibody by western blot (WB) (B) Coomassie staining of reduced SDS-PAGE of purified KLK13 showing PNGase F treatment increasing the 28kDa band intensity.\n\nKLK13 is present in unfractionated CVF at multiple sizes that correspond with a smaller, potentially degraded form of KLK13, KLK13 bound to endogenous inhibitors and KLK13 in its active mature form (Figure 7A). Analysis of the fractionated CVF sample demonstrated that the presence of KLK13 at the correct MW in the fractions corresponding with the high MW band of Figure 3 and Figure 4. The cleavage of the fluorescent substrate by immunocaptured KLK13 showed that at least a proportion of KLK13 is present in CVF in its mature, active form, which is not bound to endogenous inhibitors (Figure 7B). The proteolytic activity of recombinant KLK13 was demonstrated to be highly pH-dependent with maximum activity observed at pH 8.0 and minimal activity at the physiological pH observed in CVF (<6) (Figure 7C).\n\n(A) Western blot with KLK13 monoclonal antibody detected bands with multiple molecular sizes in unfractionated CVF. KLK13 was detected as a single band around 30-35kDa in SEC fractionated CVF F36-38 corresponding with the proteolytic activity visualized with the ABP probe in the high MW band. In-house produced recombinant KLK13 was included as a positive control (B) Immunocapture assay demonstrating the dilution dependent presence of KLK13 in CVF in active form and not bound to endogenous inhibitors (C) Effect of pH on KLK13 trypsin-like proteolytic activity showing pH dependency with optimum activity observed at pH 8.0.\n\n\nDiscussion\n\nIn-depth proteomic analysis of CVF of healthy, non-pregnant women previously revealed the presence of numerous proteases, including a large number of serine proteases, in this fluid. Indeed, multiple serine proteases are among the most abundant proteins in the CVF proteome and incubation of CVF with serine protease inhibitors significantly decreased proteolytic activity in CVF. Both findings highlight the functional importance of the serine protease family in CVF.\n\nIn the current study, we performed a functional proteomic approach coupling a biotinylated ABP probe with trypsin-like specificity with mass spectrometry, to profile the active proteases in CVF. This powerful approach resulted in the identification of serine proteases KLK13 and TMPRSS11D as major active proteolytic enzymes in CVF.\n\nTMPRSS11D, better known as human airway trypsin-like protease or HAT, displays trypsin-like activity and belongs to the type II transmembrane serine proteases23. TMPRSS11D is synthesized as an inactive zymogen which requires proteolytic cleavage to release the soluble extracellular 27kDa mature active form24,25. This transmembrane protein is expressed in the squamous epithelial cells of the vagina and the glandular cells of the endometrium and was detected in the CVF proteome7,26. Mostly studied in the context of respiratory diseases, TMPRSS11D is a versatile protein whose functions include fibrinolysis, PAR-2 receptor activation and the processing of viral glycoproteins27. Interestingly, activation of the PAR-2 receptor negatively affects the structural integrity of the vaginal epithelium prompting the hypothesis that TMPRSS11D is involved in barrier function modulation in the lower female reproductive system28–31. Additionally, TMPRSS11D is differentially expressed in the peri-implantation uterine luminal epithelium in mice where it is putatively involved in morphological changes required for the initial attachment of the embryo to the luminal epithelium32. It is unknown whether TMPRSS11D plays a similar role in the human endometrial environment.\n\nKLK13 belongs to the kallikrein-related peptidase family; a group of 15 secreted serine proteases with either tryptic or chymotryptic substrate specificity. KLKs are expressed throughout the human body with the highest expression levels observed in the fluids and tissues of the male and female reproductive systems33,34. Studies on KLK13 have predominantly focused on its applicability as a cancer biomarker, leaving the physiological role of this protein not well understood33,35–39. In vitro assays showed that despite being present in active form and unbound to endogenous inhibitors, KLK13 activity will largely be inhibited due to the low pH in CVF.\n\nKLK13 is the highest abundant KLK member among a cassette of KLKs that are co-expressed in CVF33. The concentration of KLK13 in CVF is approximately 12 mg/L (data not shown); only KLK2 and KLK3 are found at higher concentrations in the male reproductive system. Both KLK5 and KLK12 have been studied in the lower female reproductive tract and are thought to be involved in the desquamation of the vaginal epithelium, facilitation of sperm transport towards the oocyte and the processing of antimicrobial peptides in CVF40,41. KLK13 could also be involved in these processes either directly or, based on the detection of multiple members of the KLK family in CVF, via participation in a proteolytic cascade in which proteases activate consecutive proteases thereby amplifying the proteolytic potential42.\n\nIn conclusion, the current study provides insights on the proteolytic functioning in the lower female reproductive tract. Fitting in the framework that serine proteases are the predominant source of trypsin-like proteolytic activity in CVF, ABP proteomics resulted in the identification of KLK13 and TMPRSS11D as active proteases in this complex biological fluid. Both proteolytic enzymes exert their effects extracellularly in CVF which led us to believe that the activity of these enzymes will be largely inhibited by the low pH43. Based on the highly pH-dependent trypsin-like proteolytic activity in this complex biological fluid, the possibility that pH acts as a regulatory mechanism to control proteolytic activity in CVF was raised. As can be seen by the grey overlay in Figure 2A, proteolytic activity is absent at physiological pH levels (4.5 or below). It is possible however that conditions characterized by an elevated pH would allow for moderate to high levels of protease activity, and more specifically KLK13 activity, in CVF. It is known that the vaginal pH shows a dramatic rise to 7.2 within seconds after sperm ejaculation44. It is therefore not unlikely that KLK13 could be activated just after ejaculation during sexual intercourse, participates in a proteolytic cascade with proteases present in the ejaculate and initiates events associated with fertility. Additionally, CVF pH is elevated during certain vaginal infections and in postmenopausal women45,46. These conditions provide a window of opportunity for proteolytic activity in the lower female reproductive system. The current study provides a framework for future investigations aimed at understanding proteolytic functioning within the complexities of the lower female reproductive system.\n\n\nData availability\n\nUnderlying data is available from figshare:\n\nDataset 1. Ranked relative abundance proteases in CVF. https://dx.doi.org/10.6084/m9.figshare.7137920.v147\n\nDataset 2. Trypsin-like proteolytic activity in CVF. https://dx.doi.org/10.6084/m9.figshare.7137938.v148\n\nDataset 3. Visualization of active trypsin-like proteases in CVF. https://dx.doi.org/10.6084/m9.figshare.7137941.v149\n\nDataset 4. Trypsin-like proteolytic activity in size exclusion (SEC) fractionated CVF. https://dx.doi.org/10.6084/m9.figshare.7137944.v150\n\nDataset 5. Validation of TMPRSS1D in CVF. https://dx.doi.org/10.6084/m9.figshare.7137947.v151\n\nDataset 6. Production of recombinant active of protein KLK13 protein. https://dx.doi.org/10.6084/m9.figshare.7137953.v152\n\nDataset 7. Validation of KLK13 presence in CVF. https://dx.doi.org/10.6084/m9.figshare.7137962.v153\n\nAll data is available under a CC by 4.0 Universal licence",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nAcknowledgements\n\nThe authors would like to thank Professor Brendan Gilmore (Queen’s University Belfast) for providing the activity-based probe and Dr. Katerina Oikonomopoulou (University Health Network, Toronto) for her advice and suggestions during the preparation of this manuscript.\n\n\nReferences\n\nRawlings ND, Barrett AJ, Finn R: Twenty years of the MEROPS database of proteolytic enzymes, their substrates and inhibitors. Nucleic Acids Res. 2016; 44(D1): D343–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKim SO, Oh KJ, Lee HS, et al.: Expression of aquaporin water channels in the vagina in premenopausal women. J Sex Med. 2011; 8(7): 1925–30. PubMed Abstract | Publisher Full Text\n\nTraish AM, Botchevar E, Kim NN: Biochemical factors modulating female genital sexual arousal physiology. J Sex Med. 2010; 7(9): 2925–46. PubMed Abstract | Publisher Full Text\n\nAmaral E, Perdigão A, Souza MH, et al.: Postcoital testing after the use of a bio-adhesive acid buffering gel (ACIDFORM) and a 2% nonoxynol-9 product. Contraception. 2004; 70(6): 492–7. PubMed Abstract | Publisher Full Text\n\nSuarez SS, Pacey AA: Sperm transport in the female reproductive tract. Hum Reprod Update. 2006; 12(1): 23–37. PubMed Abstract | Publisher Full Text\n\nYarbrough VL, Winkle S, Herbst-Kralovetz MM: Antimicrobial peptides in the female reproductive tract: a critical component of the mucosal immune barrier with physiological and clinical implications. Hum Reprod Update. 2015; 21(3): 353–77. PubMed Abstract | Publisher Full Text\n\nMuytjens CMJ, Yu Y, Diamandis EP: Discovery of Antimicrobial Peptides in Cervical-Vaginal Fluid from Healthy Nonpregnant Women via an Integrated Proteome and Peptidome Analysis. Proteomics. 2017; 17(21): 1600461–9. PubMed Abstract | Publisher Full Text\n\nWitkin SS, Ledger WJ: Complexities of the uniquely human vagina. Sci Transl Med. 2012; 4(132): 132fs11. PubMed Abstract | Publisher Full Text\n\nRavel J, Gajer P, Abdo Z, et al.: Vaginal microbiome of reproductive-age women. Proc Natl Acad Sci U S A. 2011; 108 Suppl 1: 4680–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBoskey ER, Telsch KM, Whaley KJ, et al.: Acid production by vaginal flora in vitro is consistent with the rate and extent of vaginal acidification. Infect Immun. 1999; 67(10): 5170–5. PubMed Abstract | Free Full Text\n\nGorodeski GI: Effects of estrogen on proton secretion via the apical membrane in vaginal-ectocervical epithelial cells of postmenopausal women. Menopause. 2005; 12(6): 679–84. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGorodeski GI, Hopfer U, Liu CC, et al.: Estrogen acidifies vaginal pH by up-regulation of proton secretion via the apical membrane of vaginal-ectocervical epithelial cells. Endocrinology. 2005; 146(2): 816–24. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYang P, Liu K: Activity-based protein profiling: recent advances in probe development and applications. Chembiochem. 2015; 16(5): 712–24. PubMed Abstract | Publisher Full Text\n\nBarglow KT, Cravatt BF: Activity-based protein profiling for the functional annotation of enzymes. Nat Methods. 2007; 4(10): 822–7. PubMed Abstract | Publisher Full Text\n\nSerim S, Haedke U, Verhelst SH: Activity-based probes for the study of proteases: recent advances and developments. ChemMedChem. 2012; 7(7): 1146–59. PubMed Abstract | Publisher Full Text\n\nMayers MD, Moon C, Stupp GS, et al.: Quantitative Metaproteomics and Activity-Based Probe Enrichment Reveals Significant Alterations in Protein Expression from a Mouse Model of Inflammatory Bowel Disease. J Proteome Res. 2017; 16(2): 1014–26. PubMed Abstract | Publisher Full Text | Free Full Text\n\nNorth BB, Oldham MJ: Preclinical, clinical, and over-the-counter postmarketing experience with a new vaginal cup: menstrual collection. J Womens Health (Larchmt). 2011; 20(2): 303–11. PubMed Abstract | Publisher Full Text | Free Full Text\n\nValore EV, Park CH, Igreti SL, et al.: Antimicrobial components of vaginal fluid. Am J Obstet Gynecol. 2002; 187(3): 561–8. PubMed Abstract | Publisher Full Text\n\nPan Z, Jeffery DA, Chehade K, et al.: Development of activity-based probes for trypsin-family serine proteases. Bioorg Med Chem Lett. 2006; 16(11): 2882–5. PubMed Abstract | Publisher Full Text\n\nCox J, Mann M: MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat Biotechnol. 2008; 26(12): 1367–72. PubMed Abstract | Publisher Full Text\n\nCox J, Neuhauser N, Michalski A, et al.: Andromeda: a peptide search engine integrated into the MaxQuant environment. J Proteome Res. 2011; 10(4): 1794–805. PubMed Abstract | Publisher Full Text\n\nKapadia C, Chang A, Sotiropoulou G, et al.: Human kallikrein 13: production and purification of recombinant protein and monoclonal and polyclonal antibodies, and development of a sensitive and specific immunofluorometric assay. Clin Chem. 2003; 49(1): 77–86. PubMed Abstract | Publisher Full Text\n\nWysocka M, Spichalska B, Lesner A, et al.: Substrate specificity and inhibitory study of human airway trypsin-like protease. Bioorg Med Chem. 2010; 18(15): 5504–9. PubMed Abstract | Publisher Full Text\n\nKato M, Hashimoto T, Shimomura T, et al.: Hepatocyte growth factor activator inhibitor type 1 inhibits protease activity and proteolytic activation of human airway trypsin-like protease. J Biochem. 2012; 151(2): 179–87. PubMed Abstract | Publisher Full Text\n\nYasuoka S, Ohnishi T, Kawano S, et al.: Purification, characterization, and localization of a novel trypsin-like protease found in the human airway. Am J Respir Cell Mol Biol. 1997; 16(3): 300–8. PubMed Abstract | Publisher Full Text\n\nUhlén M, Fagerberg L, Hallström BM, et al.: Proteomics. Tissue-based map of the human proteome. Science. 2015; 347(6220): 1260419. PubMed Abstract | Publisher Full Text\n\nMenou A, Duitman J, Flajolet P, et al.: Human airway trypsin-like protease, a serine protease involved in respiratory diseases. Am J Physiol Lung Cell Mol Physiol. 2017; 312(5): L657–L68. PubMed Abstract | Publisher Full Text\n\nChokki M, Eguchi H, Hamamura I, et al.: Human airway trypsin-like protease induces amphiregulin release through a mechanism involving protease-activated receptor-2-mediated ERK activation and TNF alpha-converting enzyme activity in airway epithelial cells. FEBS J. 2005; 272(24): 6387–99. PubMed Abstract | Publisher Full Text\n\nLi Y, Feng C, Wei X, et al.: Activation of protease-activated receptor-2 disrupts vaginal epithelial barrier function. Cell Biol Int. 2014; 38(11): 1247–51. PubMed Abstract | Publisher Full Text\n\nMatsushima R, Takahashi A, Nakaya Y, et al.: Human airway trypsin-like protease stimulates human bronchial fibroblast proliferation in a protease-activated receptor-2-dependent pathway. Am J Physiol Lung Cell Mol Physiol. 2006; 290(2): L385–95. PubMed Abstract | Publisher Full Text\n\nMiki M, Nakamura Y, Takahashi A, et al.: Effect of human airway trypsin-like protease on intracellular free Ca2+ concentration in human bronchial epithelial cells. J Med Invest. 2003; 50(1–2): 95–107. PubMed Abstract\n\nXiao S, Diao H, Zhao F, et al.: Differential gene expression profiling of mouse uterine luminal epithelium during periimplantation. Reprod Sci. 2014; 21(3): 351–62. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShaw JL, Diamandis EP: Distribution of 15 human kallikreins in tissues and biological fluids. Clin Chem. 2007; 53(8): 1423–32. PubMed Abstract | Publisher Full Text\n\nMuytjens CM, Vasiliou SK, Oikonomopoulou K, et al.: Putative functions of tissue kallikrein-related peptidases in vaginal fluid. Nat Rev Urol. 2016; 13(10): 596–607. PubMed Abstract | Publisher Full Text\n\nNohara K, Yamada K, Yamada L, et al.: Expression of kallikrein-related peptidase 13 is associated with poor prognosis in esophageal squamous cell carcinoma. Gen Thorac Cardiovasc Surg. 2018; 66(6): 351–7. PubMed Abstract | Publisher Full Text\n\nDettmar L, Ahmed N, Kotzsch M, et al.: Advanced high-grade serous ovarian cancer: inverse association of KLK13 and KLK14 mRNA levels in tumor tissue and patients' prognosis. J Cancer Res Clin Oncol. 2018; 144(6): 1109–18. PubMed Abstract | Publisher Full Text\n\nLin Q, Mao W, Wu Q, et al.: Downregulation of KLK13 promotes the invasiveness and metastasis of oesophageal squamous cell carcinoma. Biomed Pharmacother. 2017; 96: 1008–15. PubMed Abstract | Publisher Full Text\n\nTokas T, Avgeris M, Alamanis C, et al.: Downregulated KLK13 expression in bladder cancer highlights tumor aggressiveness and unfavorable patients' prognosis. J Cancer Res Clin Oncol. 2017; 143(3): 521–32. PubMed Abstract | Publisher Full Text\n\nGueugnon F, Barascu A, Mavridis K, et al.: Kallikrein-related peptidase 13: an independent indicator of favorable prognosis for patients with nonsmall cell lung cancer. Tumour Biol. 2015; 36(7): 4979–86. PubMed Abstract | Publisher Full Text\n\nShaw JL, Diamandis EP: A potential role for tissue kallikrein-related peptidases in human cervico-vaginal physiology. Biol Chem. 2008; 389(6): 681–8. PubMed Abstract | Publisher Full Text\n\nShaw JL, Petraki C, Watson C, et al.: Role of tissue kallikrein-related peptidases in cervical mucus remodeling and host defense. Biol Chem. 2008; 389(12): 1513–22. PubMed Abstract | Publisher Full Text\n\nYoon H, Laxmikanthan G, Lee J, et al.: Activation profiles and regulatory cascades of the human kallikrein-related peptidases. J Biol Chem. 2007; 282(44): 31852–64. PubMed Abstract | Publisher Full Text\n\nMa B, Forney LJ, Ravel J: Vaginal microbiome: rethinking health and disease. Annu Rev Microbiol. 2012; 66: 371–89. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFox CA, Meldrum SJ, Watson BW: Continuous measurement by radio-telemetry of vaginal pH during human coitus. J Reprod Fertil. 1973; 33(1): 69–75. PubMed Abstract | Publisher Full Text\n\nCaillouette JC, Sharp CF Jr, Zimmerman GJ, et al.: Vaginal pH as a marker for bacterial pathogens and menopausal status. Am J Obstet Gynecol. 1997; 176(6): 1270–5; discussion 1275–7. PubMed Abstract | Publisher Full Text\n\nMilsom I, Arvidsson L, Ekelund P, et al.: Factors influencing vaginal cytology, pH and bacterial flora in elderly women. Acta Obstet Gynecol Scand. 1993; 72(4): 286–91. PubMed Abstract | Publisher Full Text\n\nMuytjens C: Ranked relative abundance proteases in CVF. 2018. http://www.doi.org/10.6084/m9.figshare.7137920.v1\n\nMuytjens C: Trypsin-like proteolytic activity in CVF. 2018. http://www.doi.org/10.6084/m9.figshare.7137938.v1\n\nMuytjens C: Visualization of active trypsin-like proteases in CVF. 2018. http://www.doi.org/10.6084/m9.figshare.7137941.v1\n\nMuytjens C: Trypsin-like proteolytic activity in size exclusion (SEC) fractionated CVF. 2018. http://www.doi.org/10.6084/m9.figshare.7137944.v1\n\nMuytjens C: Validation of TMPRSS1D in CVF. 2018. http://www.doi.org/10.6084/m9.figshare.7137947.v1\n\nMuytjens C: Production of recombinant active of protein KLK13. 2018. http://www.doi.org/10.6084/m9.figshare.7137953.v1\n\nMuytjens C: Validation of KLK13 presence in CVF. 2018. http://www.doi.org/10.6084/m9.figshare.7137962.v1"
}
|
[
{
"id": "39959",
"date": "01 Nov 2018",
"name": "Georgios Pampalakis",
"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 study uses functional proteomics and specifically Activity-Based Probe proteomics to identify the active proteases in cervicovaginal fluid (CVF). CVF is an interesting biological fluids because it contains high level of protease (and up to now their activation status was not known) and very low pH (<4.5). The study is well designed. There are a few minor comments that will help the readers to understand better the findings.\nSpecifically:\n\nWhat is the structure of the ABP used in the present study? (the ref 19 describes the synthesis of more than 1 ABP for probing trypsin-like activities). For their proteomic identification of ABP-targets, the authors cut the gel at the position of ABP signaling and performed a mass spectrometry analysis. In their analysis why did the authors not try to identify the tryptic fragment that contains the active-site serine covalently attached to the ABP? In this way, the authors could identify other potential proteases that bind the ABP (therefore they are active in the sample). In Figure 3C why does the inactive KLK13 sample show a signal? What is the nature of inactive KLK13 (e.g. mutant at active site Ser?)\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? Yes",
"responses": []
},
{
"id": "40431",
"date": "22 Nov 2018",
"name": "Céline Deraison",
"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 presented study aimed to characterize active serine proteases in cervicovaginal fluid (CVF). Active serine proteases were profiled using activity based probe from human CVF highlighting 2 close bands. A whole proteomic strategy was performed using human CVF and 2 proteases were deeper characterised, KLK3 and TMPRSS11D. CVF is a particular fluid since the pH is very low and also numerous proteases. The study is well designed and based on human samples, increasing the interest of the study. The manuscript is well written. However different points will need to be discussed or explained.\n\nHow could we explain why using ABP only two bands are highlighted (figure3) since 36 serine proteases were identified on MS/MS study (Figure1)? A clear correlation between activity of KLK13 and TMPRSS11D and their presence in CVF is not well established. Why authors did not perform MS/MS approach on active proteases enriched sample after ABP trapping? Authors claimed that KLK3 active form was present in CVF sample. However in figure 3 and 4 it not clear that immunoband from recombinant active KLK13 and higher band from ABP treated CVF are at the same size. In addition, why active recombinant KLK3 showed two bands after ABP trapping? Why inactive mutant KLK13 is also trapped by ABP? What is the identity of mutation? Why KLK13 is always present on gels in figure 3? Authors suggested that KLK13 is active in CVF sample since proteolytic activity is detected with immunocaptured fraction using antibody against KLK13. “The cleavage of the fluorescent substrate by immunocaptured KLK13 showed that at least a proportion of KLK13 is present in CVF in its mature, active form, which is not bound to endogenous inhibitors.In addition, the used antibody was produced in the lab. Could authors provide a bibliographic reference illustrating the specificity of the antibody since the group of KLK contains some very similar proteins. In figure 4, it could be informative to highlight KLK13 and TMPRSS11D on graphs. On this figure, it is not clear how is ranked the intensity of each protein. Does it correspond to the quantity of peptides identified by MS/MS for each protein? In discussion section, 2 sentences implied that a direct identification of active protease targeted by ABP was performed. Please, correct these sentences (‘This powerful approach resulted in the identification of serine proteases KLK13 and TMPRSS11D as major active proteolytic enzymes in CVF. ABP proteomics resulted in the identification of KLK13 and TMPRSS11D as active proteases in this complex biological fluid.)\n\nMinor comments:\nFigure 2: The graph represents the “relative trypsin-like activity”. What is the reference? What does NRT mean? The biological concentration of KLK13 is compared to KLK2 and KLK3 in sperm (“The concentration of KLK13 in CVF is approximately 12 mg/L (data not shown); only KLK2 and KLK3 are found at higher concentrations in the male reproductive system). It would be interesting to support this sentence with a bibliographic 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? 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": "4290",
"date": "31 Dec 2018",
"name": "Eleftherios P Diamandis",
"role": "Author Response F1000Research Advisory Board Member",
"response": "How could we explain why using ABP only two bands are highlighted (figure3) since 36 serine proteases were identified on MS/MS study (Figure1)? The ABP recognizes only active serine proteases, whereas the MS/MS study aims to identify all proteases in CVF regardless of their functional status. The difference in the number of identified proteases (36 in MS/MS versus 2 clusters using the ABP) is most likely due to state of the proteases. A number of proteases will be present in CVF, but in zymogen form or closely bound to endogenous inhibitors. These proteases will be identified using MS, but would not be visible on the ABP blot since these proteins do not exert proteolytic activity. A clear correlation between activity of KLK13 and TMPRSS11D and their presence in CVF is not well established. Why authors did not perform MS/MS approach on active proteases enriched sample after ABP trapping? The availability of the ABP was a limiting factor in this study thereby preventing us to perform MS/MS on an ABP-enriched samples. We aimed to establish the correlation by showing that TMPRSS11D is present in CVF at the correct MW and fractionated CVF in the expected fractions. For KLK13, we also showed that KLK13 is present in CVF and fractionated CVF at the correct MW and in the expected fractions. Additionally, we using the immunocapture assay we demonstrated that (a proportion of) KLK13 is present in CVF in active form without its prodomain or bound to endogenous inhibitors. Authors claimed that KLK3 active form was present in CVF sample. However in figure 3 and 4 it not clear that immunoband from recombinant active KLK13 and higher band from ABP treated CVF are at the same size. In addition, why active recombinant KLK3 showed two bands after ABP trapping? Why inactive mutant KLK13 is also trapped by ABP? What is the identity of mutation? Why KLK13 is always present on gels in figure 3? Authors suggested that KLK13 is active in CVF sample since proteolytic activity is detected with immunocaptured fraction using antibody against KLK13. “The cleavage of the fluorescent substrate by immunocaptured KLK13 showed that at least a proportion of KLK13 is present in CVF in its mature, active form, which is not bound to endogenous inhibitors.In addition, the used antibody was produced in the lab. Could authors provide a bibliographic reference illustrating the specificity of the antibody since the group of KLK contains some very similar proteins. The data in figure 3 and 4 demonstrating KLK13 in CVF is supported by additional evidence from the mass spectrometry analysis indicating that KLK13 is abundantly present in the bands corresponding to the higher MW band from the ABP-treated CVF. The active recombinant KLK13 shows 2 bands after ABP trapping most likely due to glycosylation. This affects the MW of the protein, while at the same time still allowing the activity of the protease. KLK13 is visible as 2 bands: one glycosylated and one not glycosylated. A logical explanation for the presence of KLK13 throughout the gel in figure 4 is due to the high abundance of this protein in CVF. It is frequently observed that high abundance proteins will not separate clearly in the expected fractions, but rather are present throughout the fractions. We expect that this is the case for KLK13 in fractionated CVF. KLK13 was inactivated not by introducing a mutation, but via incubation with the potent irreversible serine protease inhibitor PMSF. There are two explanations that clarify the small band present in the inactivated KLK13. It is possible that the band in inactive KLK13 is due to the proximity to the active recombinant KLK13 and is a result of spillover from the KLK13 lane into the inactive KLK13 lane. Alternatively, it is possible that KLK13 was not completely inactivated by PMSF resulting in the greatly diminished band shown on the blot. We have provided the reference in the manuscript for the production of the monoclonal KLK13 antibody used in this study. In figure 4, it could be informative to highlight KLK13 and TMPRSS11D on graphs. On this figure, it is not clear how is ranked the intensity of each protein. Does it correspond to the quantity of peptides identified by MS/MS for each protein? We agree that figure 4 will be improved by highlighting KLK13 and TMPRSS11D. We have adjusted the figure in the manuscript accordingly. The proteins were ranked according to the intensity of each protein for which we used the average MS1 area of each protein as a proxy measure of protein abundance. We have adjusted the legend of the figure to reflect this change. In discussion section, 2 sentences implied that a direct identification of active protease targeted by ABP was performed. Please, correct these sentences (‘This powerful approach resulted in the identification of serine proteases KLK13 and TMPRSS11D as major active proteolytic enzymes in CVF. ABP proteomics resulted in the identification of KLK13 and TMPRSS11D as active proteases in this complex biological fluid.) We agree with the reviewer and have changed these sentences in the manuscript accordingly. Minor comments: Figure 2: The graph represents the “relative trypsin-like activity”. What is the reference? For figure 2A the trypsin-like activity is relative to the trypsin-like activity measured at pH 9.0. This is the pH at which maximum trypsin-like activity was observed. In figure 2B, the trypsin-like activity is presented as relative to the activity in CVF without any inhibitors. We have adjusted the wording in the legend to better reflect this. What does NRT mean? The 225NRT stands for the N-R-T amino acids in the KLK13 sequence. We have changed the wording in the manuscript to reflect that the glysocylation site is 225N. The biological concentration of KLK13 is compared to KLK2 and KLK3 in sperm (“The concentration of KLK13 in CVF is approximately 12 mg/L (data not shown); only KLK2 and KLK3 are found at higher concentrations in the male reproductive system). It would be interesting to support this sentence with a bibliographic reference. We have included the appropriate reference to support the biological concentrations of KLKs in the CVF and seminal plasma."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1666
|
https://f1000research.com/articles/7-1177/v1
|
02 Aug 18
|
{
"type": "Research Article",
"title": "The dynamic side of the Warburg effect: glycolytic intermediates as buffer for fluctuating glucose and O2 supply in tumor cells",
"authors": [
"Johannes H.G.M. van Beek"
],
"abstract": "Background: Tumor cells show the Warburg effect: high glucose uptake and lactate production despite sufficient oxygen supply. Otto Warburg found this effect in tissue slices and in suspensions of Ehrlich ascites tumor cells. Remarkably, these ascites tumor cells can transiently take up glucose an order of magnitude faster than the steady high rate measured by Warburg for hours. Methods: The purpose of the transiently very high glucose uptake is investigated here with a computational model of glycolysis, oxidative phosphorylation and ATP consumption which reproduces short kinetic experiments on the ascites tumor cells as well as the long-lasting Warburg, Crabtree and Pasteur effects. The model, extended with equations for glucose and O2 transport in tissue, is subsequently used to predict metabolism in tumor cells during fluctuations of tissue blood flow resulting in cycling hypoxia. Results: The model analysis suggests that the head section of the glycolytic chain in the tumor cells is partially inhibited in about a minute when substantial amounts of glucose have been taken up intracellularly; this head section of the glycolytic chain is subsequently disinhibited slowly when concentrations of glycolytic intermediates are low. Based on these dynamic characteristics, simulations of tissue with fluctuating O2 and glucose supply predict that tumor cells greedily take up glucose when this periodically becomes available, leaving very little for other cells. The glucose is stored as fructose 1,6-bisphosphate and other glycolytic intermediates, which are used for ATP production during\n\nO2 and glucose shortages. Conclusions: The head section of glycolysis which phosphorylates glucose may be dynamically regulated and takes up glucose at rates exceeding the Warburg effect if glucose levels have been low for some time. The hypothesis is put forward here that dynamic regulation of the powerful glycolytic enzyme system in tumors is used to buffer oxygen and nutrient fluctuations in tissue.",
"keywords": [
"glycolysis",
"cancer",
"hypoxia",
"cycling hypoxia",
"nutrient shortage",
"computational model",
"cancer metabolism",
"oxidative phosphorylation",
"nutrient fluctuation"
],
"content": "Introduction\n\nCancer cells often show high lactate production despite sufficient oxygen supply, a phenomenon discovered by Otto Warburg1, and an example of widespread metabolic reprogramming in cancer2–4. Warburg’s favorite experimental system to study this effect were suspensions of mouse Ehrlich ascites tumor cells (EATC)1,5,6, which showed high aerobic glycolytic rates persisting for hours, at least when glucose concentrations remained high. These EATC were later also used by Warburg’s contemporaries to study the kinetics of metabolic responses in the first seconds and minutes after glucose addition to cells previously depleted of glucose7,8, showing that glucose uptake is much higher in the first minute than averaged over one hour. The results of these experiments were explained by Chance and Hess with a mathematical model, which may have been the first digital computer model of a metabolic system7,9. Their model contained some biochemical assumptions that are now considered untenable. In the present study, a small computational model is developed that economically reproduces the experimental results of the kinetic as well as the steady-state experiments on Ehrlich ascites tumor cells, and furthermore provides a testable model of the dynamic regulation of energy metabolism in the ascites tumor cells. Analysis of the model suggests that the head section of glycolysis can sequester glucose at very high capacity, but is downregulated quickly to steady-state Warburg effect levels if glucose has been taken up. However, the glycolytic head section is disinhibited slowly if glycolytic intermediates are depleted in the cells.\n\nBecause the metabolic model reproduces the behavior of the ascites tumor cells well for conditions with variable glucose levels, it is subsequently used to investigate the possible physiological role of this dynamic metabolic regulation in the tumor cells. Blood flow and the supply of oxygen and nutrients is often fluctuating in tumor tissue, a phenomenon referred to as cycling hypoxia10–12. To investigate the role of the dynamic regulation of metabolism, the computational model is extended with equations for oxygen and glucose transport in tumor tissue with cycling blood flow. The simulations reported here suggest that tumor cells can store glucose-derived metabolites to maintain ATP and carbon substrate levels during periodic oxygen and glucose shortages, as are commonly found in tumor tissue11,13. As a result, cells with lower glycolytic capacity than tumor cells have sufficient energy supply at constant blood flow, but their energy supply fails in conditions with fluctuating blood flow where tumor cells with high glycolytic capacity still do well.\n\n\nMethods\n\nThe simplified computational model developed and applied in this study comprises glycolysis, oxidative phosphorylation, ATP consumption and their interactions in the tumor cell (Figure 1). The goal of the model is to reconstruct the glucose uptake behavior and the dynamic balance of ATP, phosphorylated metabolites, glucose-derived metabolites and NADH/NAD redox status in the cell, especially in the first minute after a challenge. In addition, it also reproduces three effects which persist on the order of an hour or longer: i) the Warburg effect1,6: high glycolytic rate despite abundant oxygen availability; ii) the Pasteur effect7: increase in glycolytic rate when oxygen is depleted; and iii) the Crabtree effect6,14: decrease in oxygen uptake after addition of glucose. The computational model consists of rate equations for the head and tail part of glycolysis, oxidative phosphorylation and lactate dehydrogenase which together determine the rate of change of the key metabolites in the model, captured in a system of ordinary differential equations. The model is not meant to be a detailed reconstruction of the enzyme reactions involved and their regulatory mechanisms, but focuses on reproduction of the metabolic responses of the cell which are measured experimentally. Nevertheless, this small model reproduces the three steady effects and a range of kinetic data with satisfactory quantitative approximation.\n\nIn the head section of glycolysis, 2 ATP are spent to phosphorylate glucose, resulting in phosphorylated glycolytic intermediates (PGI) with fructose 1,6-bisphosphate (FBP) as major species. In the tail section of glycolysis four ATP, two reduced nicotinamide adenine dinucleotide (NADH) and two pyruvate molecules are produced per metabolized FBP and two inorganic phosphate (Pi) molecules are taken up. Pyruvate molecules can be converted to lactate while producing oxidized NAD. Pyruvate and NADH are also substrates for mitochondrial oxidative metabolism. ATP is used for growth, proliferation and maintenance tasks such as ion pumping. Increased NADH concentration reduces flux in the tail section. Signals from the PGI pool inhibit the head section with a time delay.\n\nChance and Hess7,9 already had developed a digital computer model to explain measurements of transients in glucose metabolism and mitochondrial respiration in Ehrlich ascites tumor cells. This was probably the first digital model of a biochemical system ever published. However, the model’s assumptions are not compatible with present biochemical knowledge: oxidative phosphorylation, for instance, was assumed to occur via a phosphorylated high energy intermediate and not via a chemiosmotic mechanism, and mitochondria were assumed to retain synthesized ATP until an uncoupling agent was applied. Therefore a new model was developed here.\n\nAlthough glycolysis has been extensively studied, it is presently still difficult to construct a fully detailed accurate model of this pathway15. Therefore, a simplified representation of glycolysis by a head and tail section is used, similar to that in old conceptual models16. This approach is also taken in recent computational17,18 models for yeast glycolysis to investigate robustness, efficiency, oscillations, and failure to start up. Consequently, the new model incorporates a parsimonious description capturing the essential kinetic properties of the glycolytic system in mammalian cells. Two kinetic equations represent the head and tail sections of glycolysis upstream and downstream of fructose 1,6-bisphosphate (FBP). These two equations make it possible to calculate the time course of the FBP pool, which can be directly compared with measurements in the experimental data sets. FBP usually also is the most abundant species of the phosphorylated glycolytic intermediates (PGI). The new model presented here further incorporates a simple description of oxidative phosphorylation in the mitochondria, which responds to ADP, inorganic phosphate (Pi) and oxygen concentrations. This equation is compatible with biochemical knowledge and has been used to investigate the functional significance of the creatine kinase energy buffer system in muscle19. The equations are discussed in detail in the Supplementary Material. The state variables of the model are given in Supplementary Table 1 and the metabolic fluxes in Supplementary Table 2.\n\nThe head section of glycolysis comprises the hexokinase, glucose 6-phosphate isomerase and phosphofructokinase enzymes, which catalyze the double phosphorylation of hexose. The most abundant phosphorylated glycolytic intermediate is FBP, which is directly represented in the model. However, the other phosphorylated glycolytic intermediates (PGI), consisting of glucose 6-phosphate, fructose 6-phosphate, dihydroxyacetone phosphate, 3-phosphoglycerate, etc., are taken into account in the storage of glucose-derived metabolites. They are lumped with FBP in the total PGI pool with a model parameter representing the fixed ratio between the sum of all phosphorylated glycolytic intermediates and FBP. In this way the total PGI content is taken into account in the time-dependent mass balance calculations. The rate of the glycolytic head section depends on glucose and ATP concentrations. The interaction of glucose and ATP in determining the rate of the head section is modelled similarly as in kinetic equations for mammalian hexokinase20,21, a major site of glycolytic rate limitation in cancer cells22.\n\nIn tumor cells there is strong negative feedback of glucose 6-phosphate (G6P) on hexokinase, the first enzyme of the head section of glycolysis22. In addition to feedback by G6P, feedback by FBP has also been reported in Ehrlich ascites tumor cells23. The feedback control on the head section of glycolysis by downstream intermediates shows a clear time delay and affects the glycolytic rate in the head section with a half time of order 10 s24,25. Binding of G6P to hexokinase also may lead to translocation of this enzyme with a similar time course26. The delayed negative feedback from the PGI pool on hexokinase is represented in the present model by a second order reaction of PGI with the head section, governed by a second order forward rate constant and a first order backward rate constant (see Eq. 22 in Supplementary Text). The forward reaction inactivates the head section and the backward reaction reactivates the inactivated head section. Representation in this simple form adequately describes the time delay of activation and reactivation. The activation state of the head section is represented by the active fraction, Factive. The delay in inhibition of the head section reproduces the overshoot in FBP concentration after glucose addition to the cell suspension, whereas previously ATP trapping in the mitochondria7,9 or complex regulatory interactions between two compartmentalized glycolytic systems had to be hypothesized16 to account for the time course of glucose uptake and FBP.\n\nThe tail section of glycolysis in the model is downstream of the FBP pool. It consists of the glycolytic enzymes aldolase, triose phosphate isomerase, glyceraldehyde 3-phosphate dehydrogenase (GAPDH), phosphoglycerate kinase, phosphoglycerate mutase, enolase and pyruvate kinase. Input reactants for the tail section are FBP, NAD+, ADP and inorganic phosphate (Pi), while its products are pyruvate, NADH and ATP. Equation 2 in the Supplementary Text represents the tail section in a lumped fashion. Each reactant which influences the reaction rate is represented by a Michaelis-Menten constant, while NADH, which is a product of the GAPDH reaction, negatively affects the forward net reaction rate in the tail section21,27.\n\nThe equation for the lactate dehydrogenase equation, pyruvate + NADH ⇌ lactate + NAD+, was taken from Lambeth and Kushmerick28. ATP consumption for maintenance, growth and cell function correlates linearly with the fall in adenine nucleotide concentration (ATP+ADP) in the experimental data, as found in the experiments of Figure 2. Incorporating this relation in the model reproduces the steep decline in ATP hydrolysis which was found after acutely giving glucose to cells which had been deprived of glucose for some time.\n\nGlucose concentration was zero at t<0, and the cells respired on endogenous substrates, such as lactate. Glucose was added at t=0. Data for experiments (dots) and model fit (lines). Left hand column: low initial glucose concentration (92 µM) was added at t=0 to a suspension of 2.2 volume percent tumor cells. Right hand column: a higher glucose concentration (776 µM) was added at t=0 to a suspension of 2.9 volume percent tumor cells. Contents of fructose 1,6-bisphosphate (FBP), ATP, total glucose taken up and total lactate produced since t=0 are given in µmol/ml cell volume. The rate of O2 consumption is given in µmol/liter intracellular water/sec.\n\nThe equations determining rates of change of metabolite levels represent balances for key players in the model: the balance of phosphoryl groups in the ATP, ADP and FBP pools, which play a central role in energy metabolism; the balance of carbon metabolites representing the distribution and storage of glucose; the balance of reduction of NAD to NADH and the reverse oxidation reaction, i.e. the NADH/NAD redox balance.\n\nThe present model provides only a coarse representation of regulatory mechanisms active in vivo, but it fulfills the goal of reproducing a broad range of measurements on EATC, both the average glucose and oxygen uptake measurements during 1 hour in Warburg’s laboratory5,6, as well as kinetic responses of glucose and oxygen uptake, lactate production, FBP and ATP levels measured in the first seconds and minutes following glucose addition7,29–32. The model is subsequently used to investigate what the physiological role is of high expression levels of glycolytic enzymes for the survival and growth of cancer cells.\n\nThe model equations are all given in the Supplementary Text, where the assumptions underlying the model are discussed further. The computational methods for integrating the system of ordinary differential equations, for parameter estimation and for uncertainty analysis are also given in the Supplementary Text.\n\n\nResults\n\nA data set was assembled consisting of representative experiments from the literature to be used to estimate parameters for the model of metabolic responses of Ehrlich ascites tumor cells (see Supplementary Text). The data sets are exemplary, but they are representative of results measured in many laboratories6–8,16,29–41. All selected experiments were done at 37°C on Ehrlich ascites cells that had been grown in mice. During the experiments, aerated tumor cell suspensions were diluted in buffer solution. Cells and suspension had been depleted of glucose for some time and were respiring on endogenous substrates such as lactate, which was abundantly present. At t=0, glucose was added to the suspension. Two kinetic data sets for the first 3–5 min consist of responses to addition of 92 µM and 776 µM glucose to cells which had been grown in ascites fluid in mice and suspended in media without glucose.\n\nThese measured responses of glucose-depleted EATC to addition of low concentrations of glucose are shown in Figure 2. The model is calibrated (Supplementary Text) on these data sets8,32, which are representative of results in several laboratories7,29–31,39,41. After adding 92 µM glucose initially32, glucose was soon exhausted (Dataset 1, experiment 1)42. After adding 776 µM glucose8, the glucose uptake rate was ~295 µM/s initially and lactate production rose to 157 µM/s in 5 s (Dataset 1, experiment 2)42. Glucose uptake was subsequently reduced by >90% within 90 s7,8,30,39. According to the model, this decline is caused by delayed feedback inhibition on the head section of glycolysis.\n\nDuring the first 20 s mitochondrial respiration is stimulated (Figure 2, right); after 30 s respiration is reduced appreciably below the initial value found before glucose addition7,29. Both the simulation and direct calculation of the mass balance of the measured phosphate metabolites shows a ~70% decline in ATP hydrolysis in the first minute, correlating with the amount of ATP plus ADP broken down to AMP, adenosine, inosine etc. This breakdown is reflected in the decreased ATP level after glucose addition (Figure 2, right). After the initial breakdown, adenine nucleotide levels recover in 0.5–1 hour27,43. The reduction of respiration after glucose addition is initially strongly determined by the reduced ATP hydrolysis.\n\nHalf of the glucose taken up in the first minute after addition is stored as PGI, mainly FBP. Subsequently, FBP declines (Figure 2, right), reflecting the delayed negative feedback on the head section of glycolysis, and settles at still appreciable levels. At 5 minutes after glucose addition, 18% of the total glucose taken up is found intracellularly as PGI, 43% has been excreted as lactate and 34% is stored intracellularly in other forms, e.g. glycogen, nucleosides and amino acids.\n\nModel predictions were subsequently compared with experiments not used for parameter estimation (Supplementary Text): Warburg’s laboratory measured 63±14 (SD) µM/s lactate production and 19±7 µM/s O2 consumption in EATC during 1 hour aerobic incubation with glucose6; the simulation predicts 52.5 µM/s lactate production and 19.8 µM/s oxygen consumption (Dataset 1; exp 3)42. Simulation further predicts that lactate production is increased by 61% during anoxia (Pasteur effect; Dataset 2)44; for comparison, in Warburg’s laboratory lactate production increased by 61±32% (SD) when oxidative phosphorylation was blocked6. Above 200 μM added glucose concentration, the peak FBP content levels off, both in experiments32,45 and in silico (Dataset 1, experiment 4)42. This is consistent with the estimated Km,glucose of 51 µM for the head section (Supplementary Table 3) and Km,glucose values reported for hexokinase, 46–78 µM20. The fast FBP and lactate accumulations measured at 5 and 10 sec8,45 after glucose addition agree with the simulations: tumor cells store for instance ~700 µM FBP intracellularly in 10 s if the initial extracellular glucose concentration is merely 77 µM (Dataset 1, experiment 5)42, demonstrating their high capacity to seize glucose.\n\nThe simulations reproduce the persistent inhibition of respiration by glucose, known as the Crabtree effect14: the average reduction over 1 hour after adding 11 mM glucose is 44% (Dataset 1, experiment 3)42, while a 30±12% (SD) reduction was measured in Warburg’s laboratory6. While the decline of respiration in the first minutes after glucose addition (Figure 2) is mainly caused by reduced ATP hydrolysis, the persisting high glycolytic ATP synthesis6 continues to keep ADP concentration and respiration reduced much longer (Dataset 1: exp 3)42.\n\nSimulations predict that ATP levels decline by 30% after glucose addition at low pyruvate concentrations because of breakdown to AMP, inosine etc. (Figure 2), but when 5 mM pyruvate is added, the predicted decline of ATP is merely 0.1% and the FBP peak decreases by 21% (Dataset 3)46; a similar pattern is seen experimentally27.\n\nIn short, the present small model economically integrates experimental data and biochemical knowledge, and quantitatively reproduces experimental results on the Warburg effect, Pasteur effect, Crabtree effect and kinetic experiments with addition of glucose. The model simulations show that after a period of glucose depletion, glucose uptake is much faster than measured for the steady Warburg effect, and that fructose 1,6-bisphosphate accumulates and can be quickly taken up in the cell’s biomass and consumed by the tail end of glycolysis where ATP is synthesized. This time-course is the consequence of inhibition of the head section of glycolysis in about 1 minute when glycolytic intermediates accumulate, and slow disinhibition of the glycolytic head section when glycolytic intermediate levels are low. A second mechanism for energy homeostasis suggested by the model consists of reduction of ATP usage, and underlies the first phase of the Crabtree effect.\n\nNext the role that the dynamic regulatory mechanisms captured in the computational model may play in tumor cell physiology is considered. ATP synthesis during hypoxia has long been considered a possible role for the glycolytic system underlying the Warburg effect. The O2 saturation of hemoglobin in capillaries in tumor tissue is often low or zero47. O2 concentrations are low in tumor tissue48 as well as in the ascites fluid in mice where EATC were grown5. Tumor blood flow sometimes stops temporarily49 and many blood vessels are not perfused over extensive periods50. Fluctuations in tumor blood flow may lead to cycling hypoxia11,51 and periodic glucose shortages. If O2 is still available when glucose is depleted, ATP can be synthesized by oxidative phosphorylation, burning lactate, fatty acids or glutamine52. If glucose is still present, glycolysis can synthesize ATP if O2 is depleted; however, the environment in solid tumors contains a glucose concentration in the order of a few hundred μM, and in many cases even <100 μM53. Cells die when anoxia is combined with glucose depletion for substantial periods of time1. Figure 2 suggests that tumor cells can store FBP and other PGI during periods of sufficient glucose supply during high blood flow in tissue (“times of abundance”). Periods of low blood flow lead to depletion of O2 and glucose (“times of famine”), and the cells can then use the stored PGI to synthesize ATP. For each FBP molecule metabolized in the tail part of glycolysis, 4 ATP molecules are synthesized (Figure 1). Stored FBP can reach ≥5000 µM, with additionally ≥1200 µM 6-carbon units stored as other PGI species (Dataset 1)42. This enables the synthesis of at least 4 × (5000+1200) μM = 25 mM ATP from PGI, potentially sustaining a high rate of ATP hydrolysis in EATC for >2 min, even after glucose and oxygen are depleted. The reduction of ATP consumption in the model, also seen experimentally in vitro, provides an additional protective mechanism: protein, DNA and RNA synthesis are presumably reduced first when ATP levels fall, followed by sodium and calcium ion pumping54–57. Warburg established experimentally that one-fifth of the normal growth energy supplied for 24 hours preserved the transplantability of tumor cells1. Reduced ATP hydrolysis required for maintaining cell viability may therefore be supported much longer than 2 min (probably at least 10 min) from FBP and other PGI stores.\n\nThe functioning of the FPB storage system of tumor cells is difficult to study experimentally in vivo. This may require metabolic measurements at a spatial resolution sufficient to distinguish low and high glycolytic cells. High time resolution to resolve the transient metabolic responses and experimental control of fluctuating O2 and nutrient supply is probably also needed. While experimental tests are challenging, the functioning of dynamic glycolytic regulation in tissue may be investigated with computational simulation.\n\nThere are limitations to experimental approaches, but the functioning of FBP buffering in vivo can be predicted with the present metabolic model, extended with well-known equations for glucose and O2 transport by blood flow and diffusion to simulate tumor tissue (Supplementary Text). The model equations for tissue transport are described in the Supplementary Text. Figure 3 shows a simulation of a hypothetical situation in tissue with blood flow fluctuating around a low average value. Similar fluctuations in blood flow are common in tumor tissue10,11,49–51. Blood flow rate, diffusion distance and plasma metabolite concentrations were set to values found in experiments on tumors implanted in rats58, while the metabolic characteristics of the simulated cells are set as determined in EATC in vitro (see above). O2 and glucose concentrations become virtually zero during the low blood flow phase, and the head section of glycolysis (Figure 3, dashed curve) and oxidative phosphorylation (blue curve) both stop. ATP synthesis from the stored FBP is quickly upregulated to replace reduced oxidative phosphorylation (red curve) and keeps ATP levels and ATP synthesis virtually constant near the level found at constant high blood flow (Dataset 4a)59. The effect of ATP synthesis by the FBP buffer mechanism is investigated by uncoupling glycolytic flux in the tail section from the associated phosphorylation of ADP. This uncoupling leads to an immediate decrease in adenine nucleotide levels and ATP hydrolysis is subsequently reduced, owing to the second homeostatic mechanism in the model. This prevents progressive imbalance of ATP hydrolysis and consumption, albeit at a lower turnover rate.\n\n(A) Model simulation of tumor cell metabolism in tissue during cycling blood flow, demonstrating ATP synthesis buffered from fructose 1,6-bisphosphate (FBP) stores. All cells have the full tumor glycolytic capacity. ATP synthesis by oxidative phosphorylation (blue line) fails periodically during low blood flow because of low oxygen supply. Glycolytic ATP synthesis by direct throughput of FBP from head to tail section fails because of glucose depletion (dashed black line). A burst of ATP synthesis from the stored fructose 1,6-bisphosphate (FBP) and other phosphorylated glycolytic intermediates (red curve) maintains ATP levels during glucose and O2 shortages. A steady state was reached after the transition at t=0 to cycling blood flow. ATP synthesis from decreasing levels of FBP was uncoupled between 3505 and 3550 seconds, leading to an immediate fall in ATP level. (B) Scheme of energy and nutrient buffering during fluctuating O2 and glucose supply. During high blood flow, FBP and other phosphorylated glycolytic intermediates are stored in the tumor cells. At low blood flow glucose and O2 are depleted. Flux in the tail part of glycolysis is maintained by use of previously stored FBP, which is replenished if blood flow increases. If blood flow stops for a long time, the intracellular FBP store is depleted.\n\nThe transition from constant to cycling blood flow was simulated (Figure 4 and Dataset 4b)59, with 80% of the cell volume consisting of tumor cells with full glycolytic capacity while the remaining 20% consists of cells with glycolytic capacity reduced to 10%. As long as blood flow is constant, ATP levels and ATP hydrolysis for cell functioning are maintained in both cell types. When blood flow starts to fluctuate, ATP concentration and ATP usage are well maintained in the cells with full glycolytic capacity. However, in the cells with 10% of the tumor glycolytic capacity, FBP buffering is appreciably decreased and adenine nucleotide levels and ATP hydrolysis fall quickly after blood flow fluctuations start. The low-capacity glycolytic cells sustain a lower rate of ATP turnover during cycling blood flow. Uncertainty analysis shows that the model predictions are sufficiently constrained (Supplementary Figure 1, Supplementary Figure 2 and Supplementary Text)60.\n\nIn this simulation, 80% of the cell volume had the full tumor glycolytic capacity; 20% of the cell volume had 10% of the full glycolytic capacity. The two top rows show tissue conditions experienced by both cell types. Blood flow and diffusion flux of glucose and O2 from the microvessel into tissue are given (top row). O2 and glucose concentrations seen by both cell types are given in the second row. Simulation for cells ~18 μm from the microvessel. High ATP consumption, >160 μM/s, was maintained at constant blood flow. See legend to Figure 3 for description of ATP synthesis fluxes. When blood flow started to fluctuate at t=0, ATP synthesis from the decline in stored fructose 1,6-bisphosphate (FBP) and other phosphorylated glycolytic intermediates (red curve) maintained ATP levels and high ATP hydrolysis rates in cells with full tumor glycolytic capacity; however, there was a drop in ATP level and ATP hydrolysis rate in the cells with reduced glycolytic capacity.\n\nATP turnover was well maintained at a constant blood flow, even for cells at merely 1.5% of the tumor glycolytic capacity which are representative of many normal cell types (Supplementary Figure 3 and Dataset 4c)1,59. However, FBP buffering was weak and ATP turnover strongly decreased during blood flow cycling. Cells at full tumor glycolytic capacity take up 50 μM/s glucose averaged over a flow cycle, while cells at 1.5% glycolytic capacity take up only 2 μM/s glucose. ATP synthesis from the FBP buffer is very low and the storage of glucose-derived metabolites for growth is compromised. Tumor cells with high glycolytic capacity take much more than their fair share of glucose.\n\nThe response to cycling blood flow in Figure 4 is influenced by two homeostatic mechanisms: FBP buffering and adaptation of ATP turnover. If ATP hydrolysis is made insensitive to the cell’s adenine nucleotide status and adaptation of ATP turnover therefore ineffective, the FBP buffering mechanism alone can still prevent the collapse of ATP during blood flow stops if the full tumor glycolytic capacity is active in the simulation (see Supplementary Figure 4 and Supplementary Text). However, glycolytic capacity reduction below the tumor level leads to compromised ATP concentration and ATP hydrolysis during flow stops. PGI stores accumulated in highly glycolytic cells during periods of high blood flow are often several-fold larger than maximal tissue glucose content (Dataset 5)61, which underscores their importance for energy and nutrient buffering.\n\nThese simulations address conditions in tumor tissue with cycling hypoxia and nutrient shortages caused by cycling blood flow. In the next section it is considered how hypoxia and low glucose concentrations can also be caused by large diffusion distances in the ascites fluid in the murine peritoneal cavity in which the ascites cells were grown in the laboratories of Warburg6, Chance29, Coe32 and others.\n\nWarburg observed that glucose and O2 concentrations were very low in the ascites fluid in the abdomen of mice in which he was growing EATC at a high cell density5; others reported ~300 µM glucose in this environment62–64. These low glucose concentrations are still sufficient for virtually maximal glucose consumption by EATC; however, glucose diffusion into the ascites fluid measured in vivo has limited capacity62, so that it can provide only a small fraction of this maximal consumption. Simulations of EATC in ascites fluid with the present model provide an explanation for this paradox: diffusion gradients over distances of hundreds of μm cause the glucose concentration in most of the ascites fluid in the peritoneal cavity to be far below the average concentration measured in fluid samples (Dataset 6, Dataset 7)65,66. Details of the calculation and results are given in the Supplementary Text. It is therefore plausible that most tumor cells in the peritoneal cavity are exposed to low glucose concentrations and consequently have very low metabolic rates.\n\nTumor cells in the ascites fluid shift position because of body movements and intestinal peristalsis64 which leads to quick changes in O2 and glucose concentrations. The cells are therefore exposed to fluctuating high and low nutrient concentrations (Supplementary Text). Greedy glucose uptake followed by storage of glucose-derived metabolites and buffering of ATP by a high-capacity glycolytic system may provide selective advantages to highly glycolytic tumor cells proliferating in environments with low and fluctuating glucose supply such as ascites fluid. This may favor the evolution of a high-capacity dynamically regulated glycolytic system in the tumor cells. Similar consideration may apply to cells in solid tumor environments which also often show low and fluctuating oxygen and nutrient supply.\n\n\nDiscussion\n\nThe present small computational model reproduces the three effects named after Warburg, Pasteur and Crabtree, respectively, which persist for an hour or more; at the same time the model captures the kinetic behavior in the first minutes after glucose addition and it is consistent with biochemical knowledge. This new concise model gives a new, testable explanation of the dynamic behavior of tumor cell metabolism, replacing the model of Chance and Hess7,9. Although the latter model has large historical value as the first digital model of a biochemical system, it contains assumptions which are biochemically untenable. Despite the present model’s explanation of a broad range of in vitro experimental data, further testing and refinement is necessary. Better understanding of the differential regulation of the head and tail sections of glycolysis is desirable. This requires experimental data revealing how the duration and extent of glucose depletion and the concentration of glycolytic intermediates affect the dynamic regulation of the head section of glycolysis in tumor cells. Although the details of the model deserve further investigation, it represents the experimental responses of ascites tumor metabolism in terms of glucose uptake, lactate production, FBP accumulation and ATP synthesis well.\n\nThe decrease of ATP consumption, correlating with the change in adenine nucleotide pool status (ATP+ADP), is required to fit the measured data in Figure 2. It should be noted that the decrease in ATP+ADP corresponds quantitatively with the accumulation of AMP, inosine, adenosine etc.27,43. The mechanism of this decrease of ATP hydrolysis requires further investigation. A useful extension would further be to model how the breakdown of ADP to AMP, inosine etc. helps to maintain the free energy of ATP hydrolysis under energetic stress by increasing the ATP/ADP ratio27,67.\n\nThe model predicts the metabolic responses in the tissue situation and provides a plausible and testable explanation why tumor cells benefit from a dynamically regulated uptake capacity of glucose that exceeds the requirements of the steady Warburg effect. The model predicts that tumor cells in tissue efficiently gulp glucose at low extracellular concentrations, and store it for the dynamic buffering of ATP and nutrients during periods of low blood flow. The model predicts that the high glucose-gulping capacity is ready for immediate action during times of famine, and is partially inhibited with some delay during times of feast, presumably to prevent overloading of the tumor cells with glucose products, while providing a time window of high uptake capacity. A remaining question is whether the time window, which provides a high capacity of glucose uptake provided by balance of inhibition and disinhibition of the head section of glycolysis, may be optimal for some cycling blood flow frequencies and not for others.\n\nExperimental interventions in the dynamic regulation of the head section of glycolysis may be employed to test the importance of the dynamic regulatory mechanism for tumor cell proliferation and growth. It is conceivable that such interventions could be beneficial for the treatment of tumors, limiting the competitiveness of tumor cells against normal tissue and immune cells.\n\nWhen tumor cells have been deprived of glucose for some time and are subsequently exposed to glucose, they can invest ~600 µM/s ATP for many seconds to sequester glucose (Dataset 1)42. For comparison, human vastus lateralis muscle consumes ~1000 µM/s for 6 s during maximal sprint performance68. The high glucose uptake capacity of tumor cells tends to keep tissue glucose concentrations low, making it difficult for competing cells with a lower glucose uptake capacity to take up sufficient glucose when supply is low and fluctuating. This may be the driving force for the evolution of Ehrlich ascites cells and tumor cells evolving in solid tumors to a state with high and dynamically regulated glucose metabolic uptake. Cells with higher glycolytic capacity also maintain higher levels of phosphorylated glycolytic intermediates to provide building blocks for macromolecular synthesis and cell growth, in addition to the dynamic ATP buffering. The hypothesis is therefore put forward here that the nutritional and energetic buffering mediated by dynamic regulation of high-capacity glucose metabolism by the glycolytic chain may give tumor cells a selective advantage over cells with lower glycolytic capacity under conditions of fluctuating oxygen and glucose supply.\n\n\nData availability\n\nDataset 1. Model simulation results for 5 experiments on suspensions of Ehrlich ascites tumor cells in vitro. Experiment 1: low glucose concentration added (92 µM); Experiment 2: higher glucose concentration added (776 μM); Experiment 3: one hour aerobic incubation with high concentration of glucose (11.1 mM); Experiment 4: maximum FBP content following addition of a range of glucose amounts; Experiment 5: accumulation of lactate and FBP after 5 and 10 s at two low initial glucose levels; Experiments 1–5 are described in Supplementary Text: Calibrating the computational model with experimental data. DOI: https://doi.org/10.5256/f1000research.15635.d21254442.\n\nDataset 2. Simulation results of incubation of Ehrlich ascites tumor cells at 11 mM glucose without oxygen, simulating experiments in Warburg’s laboratory. See description Experiment 6 in Supplementary Text: Testing the computational model with additional experimental data. DOI: https://doi.org/10.5256/f1000research.15635.d21254544.\n\nDataset 3. Simulation results of incubation of Ehrlich ascites tumor cells in vitro with 5 mM pyruvate and 10 mM glucose. See description of Experiment 7 in Supplementary Text: Testing the computational model with additional experimental data. DOI: https://doi.org/10.5256/f1000research.15635.d21254646.\n\nDataset 4. Simulations of tumor tissue including fluctuating blood flow, diffusion and tumor cell metabolism. ATP hydrolysis is high initially and strongly reduced when energy status is compromised. Simulation for tissue with a maximal diffusion distance of 40 µm. Result for the tissue layer at 15–20 μm from the blood vessel is given. Blood flow is constant for t≤0 and starts to fluctuate sinusoidally at t=0, periodically reaching zero for a moment, but not fully stopping.\n\nFor t ≤ 0: blood flow = offset.\n\nFor t>0: blood flow = offset - amplitude ∙ sin(2πt/Tperiod).\n\noffset = 4.4 ml/l intracellular H2O/s, amplitude = 4.4 ml/l/s, flow ≥ 0.\n\nWorksheet A. Simulations of tumor cells (100% of cell volume at 100% of the glycolytic capacity). From 3505–3550 sec the contribution to ATP synthesis in the tail part of glycolysis derived from falling stores of fructose 1,6-biphosphate (FBP) and other GPI is uncoupled and therefore not contributing to total ATP synthesis.\n\nWorksheet B. Simulations of tumor cells (80% of cell volume) and a second cell type with 10% of tumor glycolytic capacity (20% of volume) in tissue with fluctuating blood flow.\n\nWorksheet C. Simulations of tumor cells (80% of cell volume) and a second cell type with 1.5% of tumor glycolytic capacity (20% of volume) in tissue with fluctuating blood flow.\n\nSee Supplementary Text for details. DOI: https://doi.org/10.5256/f1000research.15635.d21254759.\n\nDataset 5. Simulations of tumor tissue with metabolism, diffusion and fluctuating low blood flow with long flow stops. Maximal ATP hydrolysis 100 µM/s. In the second (“Glycolytic capacity 100%”) and penultimate (“FBP buffering uncoupled”) worksheet all cells had the full glycolytic capacity of tumor cells. In the rest of the simulations, 95% of cell volume is occupied by tumor cells with glycolytic capacity at 100% of tumor cell level. A second cell type with lower glycolytic capacity occupies the remaining 5% of cell volume. Simulation for 8 tissue layers of width 5 μm, resulting in a maximal diffusion distance of 40 µm. Result is given for the tissue layer at 15–20 μm from the blood vessel.\n\nBlood flow is constant for t≤0 and starts to fluctuate sinusoidally at t=0, periodically stopping fully for ~2 min; for t ≤ 0: blood flow = offset; for t>0: blood flow = offset - amplitude ∙ sin(2πt/Tperiod).\n\noffset = 2.2 ml/l intracellular H2O/s, amplitude = 3.5 ml/l/s, flow ≥ 0.\n\nSix different simulations with different glycolytic capacities in the second cell type are given.\n\nWorksheet “Glycolytic capacity 100%”: all cells 100% of tumor cell level; worksheet “Glycolytic capacity 50%”: Second cell type: glycolytic capacity 50% of tumor cell level; worksheet “Glycolytic capacity 30%”: Second cell type: glycolytic capacity 30% of tumor cell level; worksheet “Glycolytic capacity 10%”: Second cell type: glycolytic capacity 10% of tumor cell level; worksheet “Glycolytic capacity 1.5%”: Second cell type: glycolytic capacity 1.5% of tumor cell level; worksheet “FBP buffering uncoupled”: Glycolytic ATP synthesis depending on falling stores of fructose 1,6-biphosphate (FBP) and other GPI uncoupled, glycolytic capacity 100% of tumor level for all cells; worksheet “Parameters”: the parameters representing the glycolytic capacities in the simulations above. DOI: https://doi.org/10.5256/f1000research.15635.d21254861.\n\nDataset 6. Simulation of diffusion of glucose from the peritoneum into ascites fluid not containing cells during 3 min. An experiment by Kemp and Mendel is simulated62, see Supplementary Text. Tumor cells and metabolism were absent in this simulation. The injected ascites fluid initially contained 167 µM glucose and the time course of glucose concentrations was simulated in sixty three stacked fluid layers with an increment of 10 µm per layer. DOI: https://doi.org/10.5256/f1000research.15635.d21254965.\n\nDataset 7. Simulation of steady-state diffusion gradients in a suspension of Ehrlich ascites tumor cells (25% vol/vol) in ascites fluid in the peritoneal cavity. This simulates conditions under which Erhlich ascites cells were grown in Warburg’s laboratory5,6 with a maximal diffusion distance of 630 μm from blood vessel into ascites fluid. This simulation resolves a paradox discussed by Kemp and Mendel62. Sixty three layers of ascites fluid with a radius increment of 10 µm per layer were simulated. DOI: https://doi.org/10.5256/f1000research.15635.d21255066.\n\n\nSoftware availability\n\nSource code available from: https://github.com/jhvanbeek/Metabolic-model-DSWE.\n\nArchived source code at time of publication: http://dx.doi.org/10.5281/zenodo.132239169.\n\nLicense: GNU General Public License v3.0.",
"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\nI thank SURFsara (https://www.surfsara.nl) for using the Lisa Compute Cluster where many of the computer calculations were performed.\n\n\nSupplementary materials\n\nSupplementary Text. Description of the computational model. This file also includes kinetic rate equations and differential equations for metabolite concentration changes with detailed description. In addition, it contains a description of the transport equations for oxygen and metabolites used for simulation of metabolism in tissue with cycling blood flow; also included are descriptions of computational methods, parameter optimization, analysis of prediction uncertainty, and of the experimental data sets used to calibrate and test the model. Description of simulation of cells growing in ascites fluid.\n\nClick here to access the data.\n\nSupplementary Figure 1. Uncertainty in prediction of the simulation results for tissue with cycling blood flow. The simulation in Figure 4 was repeated with parameter sets from 45 MCMC runs replicated with independent random seeds. The lines give the median (black) and the estimated 5 and 95% quantiles (red) of the prediction. This reveals the uncertainty in the model prediction of Figure 4. Model simulations are for tissue during blood flow cycling. Eighty percent of the cells had the full tumor glycolytic capacity (left column); 20% of the cell volume had 10% of the full glycolytic capacity (right column). The two top rows show tissue conditions experienced by both cell types. Blood flow was initially constant at 0.264 ml/ml intracellular H2O/min. For t>0 a sinus wave with amplitude 0.264 ml/ml/min was superimposed. High ATP consumption, >160 μM/s, was maintained at constant blood flow. When blood flow started cycling at t=0, ATP synthesis from the decline in fructose 1,6-bisphosphate (FBP) and other phosphorylated glycolytic intermediates (bottom row) maintained ATP levels and ATP hydrolysis rates in cells with full tumor glycolytic capacity. However, there was a sharp drop in ATP levels and hydrolysis rate in the cells with reduced glycolytic capacity. The 5% and 95% quantiles show that the predicted response pattern is reproducible for a broad range of parameter sets which reflect the potential experimental variation in the data used for parameter optimization.\n\nClick here to access the data.\n\nSupplementary Figure 2. Variation in simulation results for the in vitro experiments (see Figure 2). The time courses were calculated using the final parameter sets from 45 MCMC ensembles, replicated with independent random seeds (same as in Supplementary Figure 1). All 45 results for individual parameter sets were plotted (blue lines). The time courses for the distinct parameters sets correspond within a narrow range.\n\nClick here to access the data.\n\nSupplementary Figure 3. Simulation with cells at low glycolytic capacity during the start of cycling blood flow. Simulation as in Figure 4, but with 20% of the cells at 1.5% (rather than 10%) of the full glycolytic capacity. Model simulations of tissue during cycling blood flow: 80% of the cell volume had the full tumor glycolytic capacity; 20% of the cells had 1.5% of the full capacity, representative of many normal body cell types. The two top rows show tissue conditions experienced by both cell types. Blood flow was constant at 0.264 ml/ml intracellular H2O/min before t=0. For t>0 a sinus wave with amplitude 0.264 ml/ml intracellular H2O/min was superimposed. High ATP consumption, >160 μM/s, was maintained at constant blood flow. Concentrations and fluxes at ~18 μm from the microvessel are given. See the legend of Figure 3 for description of ATP synthesis fluxes. When blood flow started to fluctuate, ATP synthesis from the decline in fructose 1,6-bisphosphate (FBP) and other phosphorylated glycolytic intermediates (red curve) maintained ATP levels and high ATP hydrolysis rates in cells with full tumor glycolytic capacity. By contrast, glycolytic fluxes were low and there was a sharp drop in ATP levels in the cells with reduced glycolytic capacity.\n\nClick here to access the data.\n\nSupplementary Figure 4. Simulations of tumor metabolism during cycling and stopping blood flow. In the simulation on the second row and the bottom row all cells had the full tumor glycolytic capacity. In the third through penultimate row ATP concentrations and fluxes are given for cells which had their glycolytic capacity changed to a percentage of the full tumor cell glycolytic capacity, as indicated above the rows. These cells with reduced glycolytic capacity constitute 5% of the total cell volume; the remaining 95% of the cells had the full tumor glycolytic capacity (100%). Results are calculated at ~38 μm from the blood vessel. Blood flow and fluxes of O2 and glucose carried into the tissue by the arterial blood are common to all simulations (top row); the other rows each represent a separate simulation. Blood flow was initially constant and started to fluctuate at t=0, including flow stops. All ATP fluxes and blood flow are expressed per volume of intracellular H2O. The ATP synthesis flux is partitioned in oxidative phosphorylation (blue) and two contributions by the tail section of glycolysis, with fructose 1,6-bisphospate (FBP) either directly fed from the head part of glycolysis (direct glycolytic throughput: dashed line) or taken from decreasing FBP stores (FBP buffering: red line). When ATP synthesis depending on phosphorylated glycolytic intermediate stores was uncoupled (bottom row), ATP levels collapsed during flow stops, although glycolytic capacity was 100% in all cells.\n\nClick here to access the data.\n\nSupplementary Table 1. State variables in the model of tumor cell metabolism.\n\nClick here to access the data.\n\nSupplementary Table 2. Metabolic fluxes in the model of tumor cell metabolism.\n\nClick here to access the data.\n\nSupplementary Table 3. Parameters in the model of tumor cell metabolism.\n\nClick here to access the data.\n\n\nReferences\n\nWarburg O: On the origin of cancer cells. Science. 1956; 123(3191): 309–314. PubMed Abstract | Publisher Full Text\n\nSchulze A, Harris AL: How cancer metabolism is tuned for proliferation and vulnerable to disruption. Nature. 2012; 491(7424): 364–373. PubMed Abstract | Publisher Full Text\n\nVander Heiden MG, Cantley LC, Thompson CB: Understanding the Warburg effect: the metabolic requirements of cell proliferation. Science. 2009; 324(5930): 1029–1033. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMayers JR: Metabolic markers as cancer clues. Science. 2017; 358(6368): 1265. PubMed Abstract | Publisher Full Text\n\nWarburg VO, Hiepler E: Versuche mit Ascites-Tumorzellen. Z Naturforschung. 1952; 7b: 193–194. Reference Source\n\nTiedemann H: [The metabolism of the ascites tumor of the mouse]. Z Gesamte Exp Med. 1952; 119(3): 272–279. PubMed Abstract | Publisher Full Text\n\nChance B, Hess B: Spectroscopic evidence of metabolic control. Science. 1959; 129(3350): 700–708. PubMed Abstract | Publisher Full Text\n\nLee IY, Strunk RC, Coe EL: Coordination among rate-limiting steps of glycolysis and respiration in intact ascites tumor cells. J Biol Chem. 1967; 242(9): 2021–2028. PubMed Abstract\n\nChance B, Garfinkel D, Higgins J, et al.: Metabolic control mechanisms. 5. A solution for the equations representing interaction between glycolysis and respiration in ascites tumor cells. J Biol Chem. 1960; 235: 2426–2439. PubMed Abstract\n\nDewhirst MW, Cao Y, Moeller B: Cycling hypoxia and free radicals regulate angiogenesis and radiotherapy response. Nat Rev Cancer. 2008; 8(6): 425–437. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDewhirst MW: Relationships between cycling hypoxia, HIF-1, angiogenesis and oxidative stress. Radiat Res. 2009; 172(6): 653–665. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCárdenas-Navia LI, Mace D, Richardson RA, et al.: The pervasive presence of fluctuating oxygenation in tumors. Cancer Res. 2008; 68(14): 5812–5819. PubMed Abstract | Publisher Full Text\n\nDewhirst MW, Secomb TW: Transport of drugs from blood vessels to tumour tissue. Nat Rev Cancer. 2017; 17(12): 738–750. PubMed Abstract | Publisher Full Text\n\nCrabtree HG: Observations on the carbohydrate metabolism of tumours. Biochem J. 1929; 23(3): 536–545. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Eunen K, Kiewiet JA, Westerhoff HV, et al.: Testing biochemistry revisited: how in vivo metabolism can be understood from in vitro enzyme kinetics. PLoS Comput Biol. 2012; 8(4): e1002483. PubMed Abstract | Publisher Full Text | Free Full Text\n\nCoe EL, Greenhouse WV: Possible regulatory interactions between compartmentalized glycolytic systems during initiation of glycolysis in ascites tumor cells. Biochim Biophys Acta. 1973; 329(2): 171–182. PubMed Abstract | Publisher Full Text\n\nChandra FA, Buzi G, Doyle JC: Glycolytic oscillations and limits on robust efficiency. Science. 2011; 333(6039): 187–192. PubMed Abstract | Publisher Full Text\n\nvan Heerden JH, Wortel MT, Bruggeman FJ, et al.: Lost in transition: start-up of glycolysis yields subpopulations of nongrowing cells. Science. 2014; 343(6174): 1245114. PubMed Abstract | Publisher Full Text\n\nHettling H, van Beek JH: Analyzing the functional properties of the creatine kinase system with multiscale 'sloppy' modeling. PLoS Comput Biol. 2011; 7(8): e1002130. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMulquiney PJ, Kuchel PW: Model of 2,3-bisphosphoglycerate metabolism in the human erythrocyte based on detailed enzyme kinetic equations: equations and parameter refinement. Biochem J. 1999; 342 Pt 3: 581–596. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMarín-Hernández A, Gallardo-Pérez JC, Rodríguez-Enríquez S, et al.: Modeling cancer glycolysis. Biochim Biophys Acta. 2011; 1807(6): 755–767. PubMed Abstract | Publisher Full Text\n\nMarín-Hernández A, Rodríguez-Enríquez S, Vital-González PA, et al.: Determining and understanding the control of glycolysis in fast-growth tumor cells. Flux control by an over-expressed but strongly product-inhibited hexokinase. FEBS J. 2006; 273(9): 1975–1988. PubMed Abstract | Publisher Full Text\n\nGumaa KA, McLean P: Sequential control of hexokinase in ascites cells. Biochem Biophys Res Commun. 1969; 35(6): 824–831. PubMed Abstract | Publisher Full Text\n\nGumaa KA, McLean P: A possible interrelationship between binding of hexokinase and the site of ATP formation in Krebs ascites cells. Biochem Biophys Res Commun. 1969; 36(5): 771–779. PubMed Abstract | Publisher Full Text\n\nKosow DP, Rose IA: Origin of the delayed feedback control of glucose utilization in ascites tumor cells. Biochem Biophys Res Commun. 1972; 48(2): 376–383. PubMed Abstract | Publisher Full Text\n\nRose IA, Warms JV: Mitochondrial hexokinase. Release, rebinding, and location. J Biol Chem. 1967; 242(7): 1635–1645. PubMed Abstract\n\nGlaser G, Giloh H, Kasir J, et al.: On the mechanism of the glucose-induced ATP catabolism in ascites tumour cells and its reversal by pyruvate. Biochem J. 1980; 192(3): 793–800. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLambeth MJ, Kushmerick MJ: A computational model for glycogenolysis in skeletal muscle. Ann Biomed Eng. 2002; 30(6): 808–827. PubMed Abstract | Publisher Full Text\n\nChance B, Hess B: Metabolic control mechanisms. III. Kinetics of oxygen utilization in ascites tumor cells. J Biol Chem. 1959; 234: 2416–2420. PubMed Abstract\n\nHess B, Chance B: Metabolic control mechanisms. VI. Chemical events after glucose addition to ascites tumor cells. J Biol Chem. 1961; 236: 239–246. PubMed Abstract\n\nCoe EL: Correlation of glycolytic and respiratory events after addition of a small amount of glucose to Ehrlich ascites carcinoma. Cancer Res. 1966; 26: 269–275. Reference Source\n\nCoe EL, Ibsen KH, Dixon M, et al.: Glycolysis of small amounts of glucose by Ehrlich ascites carcinoma cells. Cancer Res. 1966; 26(2): 276–281. PubMed Abstract\n\nHess B, Chance B: Phosphorylation efficiency of the intact cell. I. Glucose-oxygen titrations in ascites tumor cells. J Biol chem. 1959; 234: 3031–3035. PubMed Abstract\n\nCoe EL: Correlations between adenine nucleotide levels and the velocities of rate-determining steps in the glycolysis and respiration of intact Ehrlich ascites carcinoma cells. Biochim Biophys Act. 1966; 118(3): 495–511. PubMed Abstract | Publisher Full Text\n\nCoe EL: Transient isolation of the hexokinase reaction from the glycolytic sequence on initiation of glycolysis in ascites tumor cells. Biochem Biophys Res Commun. 1970; 38(6): 1105–1112. PubMed Abstract | Publisher Full Text\n\nLee IY, Coe EL: Theoretical phosphorylation rates after addition of a small amount of glucose to intact ascites tumor cells. Biochim Biophys Acta. 1967; 131(3): 441–452. PubMed Abstract | Publisher Full Text\n\nWu R, Racker E: Regulatory mechanisms in carbohydrate metabolism. III. Limiting factors in glycolysis of ascites tumor cells. J Biol Chem. 1959; 234(5): 1029–1035. PubMed Abstract\n\nWu R, Racker E: Regulatory mechanisms in carbohydrate metabolism. IV. Pasteur effect and Crabtree effect in ascites tumor cells. J Biol Chem. 1959; 234(5): 1036–1041. PubMed Abstract\n\nLonberg-Holm KK: A direct study of intracellular glycolysis in Ehrlich's ascites tumor. Biochim Biophys Acta. 1959; 35: 464–472. PubMed Abstract | Publisher Full Text\n\nWilhelm G, Schulz J, Hofmann E: pH-dependence of aerobic glycolysis in ehrlich ascites tumour cells. FEBS Lett. 1971; 17(1): 158–162. PubMed Abstract | Publisher Full Text\n\nWilhelm G, Schulz J, Hoffmann E: [pH-dependence of glycolysis and respiration in Ehrlich ascites tumor cells]. Acta Biol Med Ger. 1972; 29(1): 1–16. PubMed Abstract\n\nvan Beek JHGM: Dataset 1 in: The Dynamic Side of the Warburg Effect: Glycolytic Intermediates as Buffer for Fluctuating Glucose and O2 Supply in Tumor Cells. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15635.d212544\n\nOvergaard-Hansen K: Metabolic regulation of the adenine nucleotide pool. I. Studies on the transient exhaustion of the adenine nucleotides by glucose in Ehrlich ascites tumor cells. Biochim Biophys Acta. 1965; 104(2): 330–347. PubMed Abstract | Publisher Full Text\n\nvan Beek JHGM: Dataset 2 in: The Dynamic Side of the Warburg Effect: Glycolytic Intermediates as Buffer for Fluctuating Glucose and O2 Supply in Tumor Cells. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15635.d212545\n\nSaha J, Coe EL: Evidence indicating the existence of two modes of glucose uptake in Ehrlich ascites tumor cells. Biochem Biophys Res Commun. 1967; 26(4): 441–446. PubMed Abstract | Publisher Full Text\n\nvan Beek JHGM: Dataset 3 in: The Dynamic Side of the Warburg Effect: Glycolytic Intermediates as Buffer for Fluctuating Glucose and O2 Supply in Tumor Cells. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15635.d212546\n\nVaupel P, Manz R, Müller-Klieser W, et al.: Intracapillary HbO2 saturation in malignant tumors during normoxia and hyperoxia. Microvasc Res. 1979; 17(2): 181–191. PubMed Abstract | Publisher Full Text\n\nHorsman MR, Vaupel P: Pathophysiological Basis for the Formation of the Tumor Microenvironment. Front Oncol. 2016; 6: 66. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDewhirst MW, Kimura H, Rehmus SW, et al.: Microvascular studies on the origins of perfusion-limited hypoxia. Br J Cancer Suppl. 1996; 74: S247–S251. PubMed Abstract | Free Full Text\n\nFisher DT, Muhitch JB, Kim M, et al.: Intraoperative intravital microscopy permits the study of human tumour vessels. Nat Commun. 2016; 7: 10684. PubMed Abstract | Publisher Full Text | Free Full Text\n\nPigott KH, Hill SA, Chaplin DJ, et al.: Microregional fluctuations in perfusion within human tumours detected using laser Doppler flowmetry. Radiother Oncol. 1996; 40(1): 45–50. PubMed Abstract | Publisher Full Text\n\nMartinez-Outschoorn UE, Peiris-Pagés M, Pestell RG, et al.: Cancer metabolism: a therapeutic perspective. Nat Rev Clin Oncol. 2017; 14(1): 11–31. PubMed Abstract | Publisher Full Text\n\nGullino PM, Grantham FH, Courtney AH: Glucose consumption by transplanted tumors in vivo. Cancer Res. 1967; 27(6): 1031–1040. PubMed Abstract\n\nSoboll S, Oh MH, Brown GC: Control of oxidative phosphorylation, gluconeogenesis, ureagenesis and ATP turnover in isolated perfused rat liver analyzed by top-down metabolic control analysis. Eur J Biochem. 1998; 254(1): 194–201. PubMed Abstract | Publisher Full Text\n\nButtgereit F, Brand MD: A hierarchy of ATP-consuming processes in mammalian cells. Biochem J. 1995; 312(Pt 1): 163–167. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAinscow EK, Brand MD: Top-down control analysis of ATP turnover, glycolysis and oxidative phosphorylation in rat hepatocytes. Eur J Biochem. 1999; 263(3): 671–685. PubMed Abstract | Publisher Full Text\n\nWieser W, Krumschnabel G: Hierarchies of ATP-consuming processes: direct compared with indirect measurements, and comparative aspects. Biochem J. 2001; 355(Pt 2): 389–395. PubMed Abstract | Publisher Full Text | Free Full Text\n\nVaupel P, Günther H, Grote J, et al.: [Respiratory gas exchange and glucose metabolism of tumors (DS-carcinosarcoma) in vivo. I. Experimental investigations of the parameters determining the supply conditions]. Z Gesamte Exp Med. 1971; 156(4): 283–294. PubMed Abstract | Publisher Full Text\n\nvan Beek JHGM: Dataset 4 in: The Dynamic Side of the Warburg Effect: Glycolytic Intermediates as Buffer for Fluctuating Glucose and O2 Supply in Tumor Cells. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15635.d212547\n\nGutenkunst RN, Waterfall JJ, Casey FP, et al.: Universally sloppy parameter sensitivities in systems biology models. PLoS Comput Biol. 2007; 3(10): 1871–1878. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Beek JHGM: Dataset 5 in: The Dynamic Side of the Warburg Effect: Glycolytic Intermediates as Buffer for Fluctuating Glucose and O2 Supply in Tumor Cells. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15635.d212548\n\nKemp A, Mendel B: How does the Ehrlich ascites tumour obtain its energy for growth? Nature. 1957; 180(4577): 131–132. PubMed Abstract | Publisher Full Text\n\nBurgess EA, Sylvén B: Changes in glucose and lactate content of ascites fluid and blood plasma during growth and decay of the ELD ascites tumour. Br J Cancer. 1962; 16: 298–305. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKlein G: [Some recent studies on the production and growth characteristics of ascites tumors; a review]. Z Krebsforsch. 1956; 61(2): 99–119. PubMed Abstract | Publisher Full Text\n\nvan Beek JHGM: Dataset 6 in: The Dynamic Side of the Warburg Effect: Glycolytic Intermediates as Buffer for Fluctuating Glucose and O2 Supply in Tumor Cells. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15635.d212549\n\nvan Beek JHGM: Dataset 7 in: The Dynamic Side of the Warburg Effect: Glycolytic Intermediates as Buffer for Fluctuating Glucose and O2 Supply in Tumor Cells. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.15635.d212550\n\nKroll K, Kinzie DJ, Gustafson LA: Open-system kinetics of myocardial phosphoenergetics during coronary underperfusion. Am J Physiol. 1997; 272(6 Pt 2): H2563–H2576. PubMed Abstract | Publisher Full Text\n\nGray SR, De Vito G, Nimmo MA, et al.: Skeletal muscle ATP turnover and muscle fiber conduction velocity are elevated at higher muscle temperatures during maximal power output development in humans. Am J Physiol Regul Integr Comp Physiol. 2006; 290(2): R376–382. PubMed Abstract | Publisher Full Text\n\njhb: jhvanbeek/Metabolic-model-DSWE: First release. (Version v1.0.0). F1000 Research. Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1322391"
}
|
[
{
"id": "36758",
"date": "11 Sep 2018",
"name": "Ranjan K. Dash",
"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\nThis is an interesting original research article based on computational modeling to understand several questions related to cancer cell metabolism, specifically on the Warburg effects and related phenomena (high glucose uptake and lactate production by cancer cells despite sufficient oxygen supply; also very high transient glucose uptake one order of magnitude faster than the high steady state glucose uptake). In doing this, the author developed a new computational model integrating simplified \"phenomenological\" models of key lumped reactions in glycolysis, mitochondrial ATP synthesis coupled to oxygen consumption (oxidative phosphorylation; OxPhos), and cytosolic ATP consumption. I am not an expert in cancer cell metabolism, so I do not know how much experimental data are available for testing and calibrating the mathematical model. I trust the author has carefully considered all the important experimental data that are available for the calibration of the model. For example, I see the author has considered some key experimental data sets showing transients in several important variables in the mathematical model governing cellular metabolism in cancer cells (Figure 2). This is surely an important data sets for model calibration, as it depicts the transients of fructose biphosphate (FBP) content, ATP content, glucose uptake, lactate production, and oxygen consumption for two different glucose stimulation conditions in cancer cells (low glucose and high glucose conditions). These data show distinct characteristics in these important variables for these two perturbation conditions. Based on the calibrated model, the author used model simulations to gather several interesting insights into the cancer cell metabolism. Finally, the author has very well recognized the limitations of the model. I still wanted to point out few further limitations of the modeling.\nAs I mentioned above, the modeling of the lumped reaction processes in this integrated model are highly phenomenological. The author considered two lumped reactions of the whole glycolysis process (termed as \"head\" and \"tail\" portion of the glycolysis), one reaction representing OxPhos, and one reaction for ATP consumption. I wonder what is the philosophy behind the formulated flux expressions for these lumped reactions. I see the author has not given any thermodynamic consideration in the modeling of these lumped reactions (e.g. reversible reaction fluxes satisfying the Haldane constraint, relating kinetic parameters to the Gibbs free energy of the lumped reactions). The author has also not included the stoichiometry of the biochemical species in the modeled reaction fluxes. For example, the modeled lumped reaction fluxes do not include square terms considering two ATP and ADP are involved in the lumped phosphorylation-dephosphorylation reactions. Also I am wondering if the author has individually parameterized these reactions fluxes prior to integration and testing the integrated model with the dynamic data in Figure 2.\nBesides these few modeling limitations, the author has done a great job in developing, calibrating and testing the model, and coming up with interesting conclusions regarding the operation of this complex metabolic system in cancer cell. I am not sure how these findings can be experimentally tested, but I would invite the author to put some of his thoughts on further testing experimentally these interesting model predictions. Besides, I have few other suggestions:\n1. I would suggest the author to show the model simulated reaction fluxes along with the model fittings in Figure 2.\n2. I would suggest the author using the acronym EATC throughout after it is defined.\n3. I did not see how the effect of blood flow is integrated into the model. I thought the model is for isolated cell experiments in a cuvette.\n4. The whole experimental system (cells in buffer) is considered as a single compartment. I am wondering how the model prediction would alter if one considers compartmentation (e.g. extracellular, cytosol, and mitochondria as separate compartments with transport of species in and out of the compartments).\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": "4295",
"date": "28 Dec 2018",
"name": "Hans van Beek",
"role": "Author Response F1000Research Advisory Board Member",
"response": "Answer to reviewer Dash: Thank you very much for your constructive criticism. I addressed all of your suggestions and questions, despite your approval of the manuscript without reservations. I think this led to considerable improvement of the manuscript. Details are given below. … the modeling of the lumped reaction processes in this integrated model are highly phenomenological. The author considered two lumped reactions of the whole glycolysis process (termed as \"head\" and \"tail\" portion of the glycolysis), one reaction representing OxPhos, and one reaction for ATP consumption. I wonder what is the philosophy behind the formulated flux expressions for these lumped reactions. The philosophy behind the flux expressions for the head and tail sections of glycolysis is explained in considerably greater detail now in the main text and Supplementary Text. This led to rewriting of the Methods section in the main text, in particular the first eight paragraphs, as well as of the Supplementary Text, in particular the first two and a half pages. The reasons for modeling lumped reaction processes rather than detailed enzymatic equations are addressed in the first and third paragraphs of the Methods section of the main text, in the first two paragraphs of the Supplementary Text and in response to a question by your fellow reviewer, Dr Dai. In addition, a full section was added to the Supplementary Text, entitled “Reasons for Employing a Coarse-grained Rather than Detailed Model of Glycolysis”. Here some of the main considerations are given regarding the equations for the head and tail section of glycolysis. The flux equation for the lumped reaction of the head section is made to depend on the concentration of the substrates which are used in this section: ATP and glucose. The equation is similar as used for hexokinase by Marin-Hernandez et al. (1) and by Mulquiney and Kuchel (2). This seems a relevant equation because hexokinase is the entry point of the head section and an important control point of rate limitation in this section (3), and this is the point where glucose and ATP are processed by the same enzyme. (ATP is of course also involved in the subsequent phosphofructokinase reaction.) Product inhibition of hexokinase is important in Ehrlich ascites tumor cells (4). In Mulquiney and Kuchel’s model inhibitors of hexokinase bind to the glucose-bound form of the enzyme and binding is assumed there to be instantaneously in equilibrium. In the present model, inhibition is independent of the substrate-binding state and is not instantaneous, leading to slow inactivation (see main text, fifth paragraph of Methods and in the Supplementary Text above Eq. 1). Whether this simplified description of the head section is satisfactory was in the end judged by its description of the data: the fit to the input and output of the head section, glucose uptake and fructose 1,6-bisphosphate, appears good. The tail section of glycolysis was represented by an equation which depends on the input reactants of the tail section: FBP, ADP and NAD+. Each is represented in the equation by a factor with a Michaelis-Menten constant. NADH, which is a product of the GAPDH reaction, not only negatively affects the forward net reaction rate in enzyme kinetic equations by increasing reverse flux in GAPDH (1, 5), but NADH is also an inhibitor of GAPDH in EATC (6). These negative effects of high NADH levels are represented by a factor with an inhibition constant. The factor depending on FBP not only reflects the saturable dependence of the tail part on phosphorylated substrate, but also allosteric activation of the pyruvate kinase enzyme in the tail section (7, 8). Therefore the complex kinetics of the enzymes in the tail part is represented by an approximation containing a limited number of parameters that are estimated from the experimental data on the intact system. The equation roughly reflects the kinetics of glyceraldehyde 3-phosphate dehydrogenase, which is an important limiting step in the tail part of glycolysis (9). The effect of ADP as a substrate for the reactions in the glycolytic tail section (phosphoglycerate kinase and pyruvate kinase) is also reflected by a factor with a Michaelis-Menten type constant. Discussion of the equations for ATP consumption is given in the Supplementary Text, covering about one page around Equations 5 and 6. The equation for oxidative phosphorylation was described and used previously by others and myself: see discussion in Van Beek (10) and Hettling and Van Beek (11) where a detailed model of oxidative phosphorylation described by Vendelin et al. (12) and Korzeniewski (13) is also discussed. These considerations have now been extended in the main text and the Supplementary Text. I see the author has not given any thermodynamic consideration in the modeling of these lumped reactions (e.g. reversible reaction fluxes satisfying the Haldane constraint, relating kinetic parameters to the Gibbs free energy of the lumped reactions). Your fellow reviewer, Dr Dai, gave a similar comment. I have therefore included the following answer also in the answer to her. Glycolysis is considered by biochemists to be essentially irreversible, in particular the hexokinase, phosphofructokinase and pyruvate kinase reactions (see for instance the textbook by Berg, Tymocko, Stryer. Biochemistry, Sixth Edition, 2007, e.g. page 460, ref. (14)). In addition the glyceraldehyde 3-phosphate dehydrogenase reaction may also be an important limiting step in cells showing the Warburg effect (9, 15). To produce glucose in the reverse direction (fructose 1,6-bisphosphate to glucose) the enzyme glucose 6-phosphatase is required to bypass the irreversible hexokinase reaction. Glucose 6-phosphatase’s activity in Ehrlich ascites cells is low (16), while its activity in the liver, which can produce glucose, is high. The tail section of glycolysis is also considered to be virtually irreversible and two additional enzymes (pyruvate carboxylase and phoshoenolpyruvate carboxykinase) are required to enable conversion of pyruvate in the direction of glucose, bypassing pyruvate kinase (see (14), p. 460-463). I had implemented two existing detailed models of the glycolytic chain and used these to explore the reversibility of the reactions. The first model was based on the model of Lambeth and Kushmerick (5), extended with the equation for hexokinase given by Mulquiney and Kuchel (2). The second model was based on the model of Marin-Hernandez et al. (1) for cancer cells. Both models represent the ten enzymatic steps in the glycolytic chain and incorporate both the forward and reverse reactions for all the steps. All these equations take the Haldane relation into account, and constraints based on thermodynamics are obeyed. Simulations with both models showed that at least one reaction in the head as well as the tail section of the glycolytic chain showed reverse fluxes that were extremely small relative to the forward fluxes. The results are discussed in more detail in the revised Supplementary Text, section titled “Effect of Reverse Fluxes on the Glycolytic Model”. These simulations illustrate the biochemical concept that glycolysis is essentially irreversible. In order to reduce the model as much as possible, the reverse reactions in glycolysis were therefore neglected. However, the lactate dehydrogenase reaction is included in fully reversible form and lactate is taken up for mitochondrial metabolism if glucose is depleted in the simulations. Mitochondrial oxidative phosphorylation coupled to oxygen consumption is also modeled as being essentially irreversible, as generally accepted by biochemists. However, these considerations point to a limitation: the model is not suitable for determining the magnitude of small reversed fluxes in the glycolytic chain under conditions where glycolytic products (pyruvate, NADH, ATP) are high and glucose concentration is very low. Reverse fluxes may be revealed by isotope exchange experiments. These considerations are now included in the Methods section of the main paper (8th paragraph) and the Supplementary Text (section title “Effect of Reverse Fluxes on the Glycolytic Model”). The author has also not included the stoichiometry of the biochemical species in the modeled reaction fluxes. For example, the modeled lumped reaction fluxes do not include square terms considering two ATP and ADP are involved in the lumped phosphorylation-dephosphorylation reactions. Also I am wondering if the author has individually parameterized these reactions fluxes prior to integration and testing the integrated model with the dynamic data in Figure 2. The kinetic rate equations are not based on mass action kinetics, but represent saturable enzyme kinetics with Michaelis-Menten type constants. When 2 ATP are consumed to phosphorylate glucose to fructose 1,6-bisphosphate, this occurs in two separate enzyme steps in the head section of the glycolytic chain which both show saturable kinetics. Therefore the rate equation was modeled with saturable kinetics without square terms, resembling the kinetics of hexokinase, a major control point of rate limitation in the glycolytic head section in cancer cells (3). These considerations have now been more extensively explained in the revised manuscript. However, note that the stoichiometry of the biochemical reactions was accounted for accurately in the differential equations that describe the mass balances of the species. For instance, the stoichiometry of the lumped head section reaction is glucose + 2 ATP -> fructose 1,6-bisphosphate + 2 ADP (for clarity the reaction is given without electric charges and H+ ions). The stoichiometric coefficient 2 is accounted for in algebraic Eq. 10 (Supplementary Text), and the total stoichiometry of the reaction is embedded in differential equations 16, 19, 20 and 21. This also applies to other reactions. I had not individually parameterized the rate equations prior to integrating them in the model. The consideration was that no data are available for the isolated head section or isolated tail section. If the enzymes for both these sections were isolated, their behavior would likely be affected by the isolation procedure and their isolated state. A further consideration was that the behavior of head or tail section would be best represented by the data on the intact cell, where the input and output of the head section is for instance well represented in the data by its input (glucose uptake) and output (FBP). This is discussed in the second paragraph in the Supplementary Text. I am not sure how these findings can be experimentally tested, but I would invite the author to put some of his thoughts on further testing experimentally these interesting model predictions. Here follow some ideas on further testing of the model of metabolic responses of Ehrlich ascites cells in vitro: the model predicts that a major part of the glucose taken up in an experiment as in Fig. 2 is stored in the cell. The time course, target molecules and amounts of storage could be investigated using labeled glucose (e.g. with carbon-13) and the time course of incorporation of label in pyruvate, lactate, glycolytic intermediates and side reactions of glycolysis (e.g. serine synthesis, pentose phosphate pathway) may be investigated - a prediction of the model is that there is no initial stimulation of respiration after addition of a high concentration of glucose if a high concentration of pyruvate is simultaneously present; the predicted time courses of lactate production and respiration during one hour can be measured. - the time course of metabolites not yet used for parameter estimation may be measured. An interesting candidate is NADH, whose time course might be measured by its autofluorescence. ATP and/or ADP concentrations can be measured by fluorescent probes or NMR spectroscopy. Glucose may be measured using a fluorescent probe. - the effect of partial inhibition or lowered expression of the head or tail portion of glycolysis on the time course of glucose uptake, FBP accumulation and lactate production can be simulated and tested experimentally. the model predicts that feedback from glycolytic intermediate pools limits the initially very high glucose uptake and FBP accumulation: this may be investigated by searching for chemicals which block the inhibition. Binding of the chemicals may be used to identify potential binding sites. The levels of glycolytic intermediates may be manipulated to identify potential candidates for feedback inhibition. Enzymes isolated from the Ehrlich ascites tumor cells may be isolated to test for their sensitivity to inhibition by glycolytic intermediates. This also applies to hexokinase bound to mitochondrial membranes. the time course of disinhibition after glucose and glycolytic intermediate depletion is still relatively imprecisely determined: this may be tested and fine-tuned by altering duration and extent of glucose depletion (zero glucose or slightly higher). genetic modification of genes may reveal key controllers of the dynamic regulation of glycolysis in the ascites tumor cells measurement of the rate of phosphoryl group transfer by using NMR saturation transfer or 18O labeling measurements Possible tests of the prediction of responses of tumor cells in vivo are: perfuse a cancer tissue with a pump, imposing fluctuating blood flow to be compared with constant blood flow. The preparation used by Vaupel (17) may be well suited because it has a distinct blood vessel supply and homogeneous cell population. Measure NADH fluorescence and tissue oxygen concentration continuously. Measure metabolite concentrations as function of the phase of the blood flow cycle. If sufficient cell mass is available: measure phosphoryl group transfer rate by saturation transfer NMR. Repeat this after partial inhibition of glycolysis, possibly at the level of glyceraldehyde 3-phosphate dehydrogenase. If different cell types are present: measure NADH fluorescence and metabolites with fluorescent labels in individual cells. Measure markers of cell death or apoptosis after a period of fluctuating blood flow in comparison with a period of constant blood flow. remove glucose from the perfusate but retain lactate. The prediction is this will be adequate to sustain tumor cell metabolism with constant flow, but it will cause an energy crisis during cycling blood flow. measure growth and proliferation of tumor tissue with and without inhibition of the feedback loop in case a suitable inhibitor has been identified. set up a system for perfusion of a cell culture containing different cell types with different glycolytic capacities. Perfuse with constant oxygen and glucose supply and with fluctuating oxygen and glucose supply. Measure markers of cell damage after a period of constant supply and after a period of fluctuating supply. Measure intracellular concentrations with fluorescent probes. Such probes are for instance available for ATP and glucose. These suggestions for experimental testing have now been added to the main text (Discussion, second paragraph for testing metabolic model in vitro, and ninth paragraph for testing predictions for tissue in vivo). Besides, I have few other suggestions: 1. I would suggest the author to show the model simulated reaction fluxes along with the model fittings in Figure 2. I followed this excellent suggestion. I have added a figure 2B with the simulated reaction fluxes applying to the responses in the previous Fig. 2, now designated Fig. 2A. 2. I would suggest the author using the acronym EATC throughout after it is defined. I have followed your suggestion in the revised manuscript. 3. I did not see how the effect of blood flow is integrated into the model. I thought the model is for isolated cell experiments in a cuvette. In the original submission the description of equations for blood flow and diffusion was only given in the Results, with all details of the blood flow and diffusion equations in the supplementary material, equations 24-31. Description of the extended model of cell metabolism with transport equations in tissue is now explicitly included in the Methods section of the revised manuscript with reference to the details given in the Supplementary Text. In the abstract the sentence “The model, extended with equations for glucose and O2 transport in tissue, is subsequently used to predict metabolism in tumor cells during fluctuations of tissue blood flow resulting in cycling hypoxia.” was rephrased: “The model is subsequently extended with equations for glucose and O2 transport to predict the role of metabolism during fluctuations of blood flow in tumor tissue.” 4. The whole experimental system (cells in buffer) is considered as a single compartment. I am wondering how the model prediction would alter if one considers compartmentation (e.g. extracellular, cytosol, and mitochondria as separate compartments with transport of species in and out of the compartments). Three compartments actually exist in the present model: extracellular, cytosolic and mitochondrial. All glycolytic intermediates are confined to the cytosolic compartment and their concentration changes are calculated based on the cytosolic volume. Fructose 1,6-bisphosphate and other phosphorylated glycolytic intermediates do not cross the cell membrane or the inner mitochondrial membrane, as usually assumed in biochemistry. ADP enters the mitochondrial compartment in exchange for ATP synthesized in the mitochondria. Their rate of change in the cytosol is calculated based on the cytosolic volume. Oxygen, glucose, lactate and pyruvate are assumed to have the same concentration in the extracellular and cytosolic compartment and their concentration is calculated based on the sum of extracellular and cytosolic volumes. This is of course a simplification of the real situation. The cell membrane causes little impediment for oxygen: the apparent Michaelis-Menten constant for oxygen, KO2,mit, was determined by Froese (18) to be 0.26 μM based on the O2 concentration in the environment of the cell. This shows that the oxygen concentration gradient between environment and cell interior is very small. Transport of glucose into the cell is part of the lumped equation for the head section of glycolysis, as is now emphasized in the revised manuscript. The transport of glucose across the cell membrane in EATC was measured by Saha and Coe (19). They found two transport systems: a high affinity one with Km=100 µM and Vmax = 467 µM/s and a low affinity one with Km=25 mM and Vmax = 3333 µM/s. I tested the effect of adding the glucose transport equation with these parameters to the model, which means that glucose transport was separate from the head section. The parameters of the model were then estimated again by the same procedure, keeping the glucose parameters fixed at the value measured by Saha and Coe. The fit of the model to the data was virtually indiscernible from that in Fig. 2A and the cost function value was even slightly lower. Changes in parameter values were small: for instance, the Vmax,head was 394 μM/s without the separate glucose transport equations, and 376 μM/s with the additional glucose transport equations. However, there was one exception: the Kglucose,head which represents the apparent affinity of the head section for glucose fell from 51 μM to 4.9 μM, and the calculated glucose concentration in the cytosol was decreased. Note that in the model in the present manuscript the contribution of glucose transport is integral part of the head section of glycolysis and parameters are estimated for the behavior of this entire head section, taking measured glucose uptake and FBP accumulation into account, as is now emphasized in the revised text. Lactate and pyruvate are transported via highly active transporters: lactate transport across the cell wall of EATC is characterized by a Vmax of 1866 µM/s with Km=4.7 mM and pyruvate transport is characterized by a Vmax of 1290 µM/s with Km=8.5 mM (20). Although the finite permeability of the cell membrane causes concentration differences and some delay between the extracellular and cytosolic compartment, the arguments given above suggest that these effects are limited. To keep the model as simple as possible, the transport processes across the cell membrane were therefore not explicitly included in the model. These considerations are now added to Supplementary Text, in the paragraphs following Equations 16 and 18. REFERENCES 1. A. Marin-Hernandez, J. C. Gallardo-Perez, S. Rodriguez-Enriquez, R. Encalada, R. Moreno-Sanchez, E. Saavedra, Modeling cancer glycolysis. Biochim Biophys Acta 1807, 755-767 (2011). 2. P. J. Mulquiney, P. W. Kuchel, Model of 2,3-bisphosphoglycerate metabolism in the human erythrocyte based on detailed enzyme kinetic equations: equations and parameter refinement. Biochem J 342, 581-596 (1999). 3. A. Marin Hernandez, S. Rodriguez-Enriquez, P. A. Vital-Gonzalez, F. L. Flores-Rodriguez, M. Macias-Silva, M. Sosa-Garrocho, R. Moreno-Sanchez, Determining and understanding the control of glycolysis in fast-growth tumor cells. Flux control by an over-expressed but strongly product-inhibited hexokinase. FEBS J 273, 1975-1988 (2006). 4. D. P. Kosow, I. A. Rose, Origin of the delayed feedback control of glucose utilization in ascites tumor cells. Biochem Biophys Res Commun 48, 376-383 (1972). 5. M. J. Lambeth, M. J. Kushmerick, A computational model for glycogenolysis in skeletal muscle. Annals of Biomedical Engineering 30, 808-827 (2002). 6. S. Bagui, M. Ray, S. Ray, Glyceraldehyde-3-phosphate dehydrogenase from Ehrlich ascites carcinoma cells. Its possible role in the high glycolysis of malignant cells. Eur J Biochem 262, 386-395 (1999). 7. M. S. Jurica, A. Mesecar, P. R. Heath, W. Shi, T. Nowak, B. L. Stoddard, The allosteric regulation of pyruvate kinase by fructose-1,6-bisphosphate. Structure 6, 195-210 (1998). 8. H. R. Christofk, M. G. Vander Heiden, N. Wu, J. M. Asara, L. C. Cantley, Pyruvate kinase M2 is a phosphotyrosine-binding protein. Nature 452, 181-186 (2008). 9. A. A. Shestov, X. Liu, Z. Ser, A. A. Cluntun, Y. P. Hung, L. Huang, D. Kim, A. Le, G. Yellen, J. G. Albeck, J. W. Locasale, Quantitative determinants of aerobic glycolysis identify flux through the enzyme GAPDH as a limiting step. Elife 3, (2014). 10. J. H. van Beek, Adenine nucleotide-creatine-phosphate module in myocardial metabolic system explains fast phase of dynamic regulation of oxidative phosphorylation. Am J Physiol Cell Physiol 293, C815-829 (2007). 11. H. Hettling, J. H. van Beek, Analyzing the functional properties of the creatine kinase system with multiscale 'sloppy' modeling. PLoS Comput Biol 7, e1002130 (2011). 12. M. Vendelin, O. Kongas, V. Saks, Regulation of mitochondrial respiration in heart cells analyzed by reaction-diffusion model of energy transfer. Am J Physiol Cell Physiol 278, C747-C764 (2000). 13. B. Korzeniewski, Regulation of ATP supply during muscle contraction: theoretical studies. Biochem J 330, 1189-1195 (1998). 14. J. M. Berg, J. L. Tymoczko, L. Stryer, Biochemistry. (W.H. Freeman and Company, ed. Sixth, 2007). 15. M. V. Liberti, Z. Dai, S. E. Wardell, J. A. Baccile, X. Liu, X. Gao, R. Baldi, M. Mehrmohamadi, M. O. Johnson, N. S. Madhukar, A. A. Shestov, I. I. C. Chio, O. Elemento, J. C. Rathmell, F. C. Schroeder, D. P. McDonnell, J. W. Locasale, A Predictive Model for Selective Targeting of the Warburg Effect through GAPDH Inhibition with a Natural Product. Cell Metab 26, 648-659 e648 (2017). 16. W. J. Nelson, S. J. Nelson, P. Traub, Comparison of Ehrlich ascites tumour and mouse liver cells by analytical subcellular fractionation combined with a sensitive computational method for data analysis. Hoppe-Seyler's Z. Physiol. Chem. 362, 903-918 (1981). 17. P. Vaupel, H. Günther, J. Grote, Atemgaswechsel und Glucosestoffwechsel von tumoren (DS-Carcinosarkom) in vivo. I. Experimentelle Untersuchungen der versorgungsbestimmenden parameter. Z ges exp Med 156, 283-294 (1971). 18. G. Froese, The respiration of ascites tumor cells at low oxygen concentration. Biochimica en Biophysica Acta 57, 509-519 (1962). 19. J. Saha, E. L. Coe, Evidence indicating the existence of two modes of glucose uptake in Ehrlich ascites tumor cells. Biochem Biophys Res Commun 26, 441-446 (1967). 20. T. L. Spencer, A. L. Lehninger, L-Lactate transport in Ehrlich ascites-tumour cells. Biochem J 154, 405-414 (1976)."
}
]
},
{
"id": "37981",
"date": "14 Sep 2018",
"name": "Ziwei Dai",
"expertise": [
"Reviewer Expertise metabolism",
"epigenetics",
"computational biology",
"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\nThe manuscript ‘The dynamic side of the Warburg effect: glycolytic intermediates as buffer for fluctuating glucose and O2 supply in tumor cells’ presents a coarse-grained kinetic model of glycolysis. The model, despite being extremely simplified, is able to reproduce several experimental datasets in Ehrlich ascites tumor cells after parameter calibration. The author then applies this model to predict dynamic behaviors of tumor cells under fluctuating blood flow. I like the idea of using simplified models to interpret complex biological phenomena but would like to see more discussions about the rationale of using this model instead of other models with full details of every reaction involved in glycolysis. There is space to improve presentation of the results as well. Specific comments are listed below for the author’s reference.\nSome important interactions that may affect FBP dynamics are missing in the model. For instance, it is known that FBP is an allosteric activator of pyruvate kinase (Jurica et al1; Christofk et al2). I expect that this feedback will attenuate the FBP buffering mechanism proposed in this study since lower glycolysis (i.e. 'tail' in this model) is activated by high concentration of FBP thus enhancing its consumption. Moreover, the two sections (‘head’ and ‘tail’) of glycolysis are both considered to be irreversible, thus neglecting the effects of thermodynamics on glycolytic flux. This may also lead to overestimation of FBP concentration as well because FBP is not allowed to be converted back to glucose. The main goal of developing this coarse-grained model is not clear to me. Besides the model developed by Chance and Hess, there are numerous mathematical models for glycolysis, most are much more detailed than the model presented in this study. The author claims that this model replaces the model of Chance and Hess – I feel this inappropriate since this statement doesn’t give any credit to all other glycolysis models. Most, if not all, results are presented as curves from the simulation with little information about the take-home messages. It is thus very difficult to read the key findings directly from the figures. For instance, it is my understanding that one of the two most important hypotheses drawn based on the simulation is that cells with high glycolytic capacity (which mimics ‘Warburgian’ cancer cells) consume glucose much more quickly than cells with lower glycolytic capacity after switching from low glucose to high glucose condition, thus being more competitive under conditions with frequent nutrient deprivation. To emphasize this point, I would recommend using one figure directly comparing glucose uptake fluxes in cells with different glycolytic capacities instead of the 10 figure panels currently included in Fig 4.\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? Yes",
"responses": [
{
"c_id": "4294",
"date": "28 Dec 2018",
"name": "Hans van Beek",
"role": "Author Response F1000Research Advisory Board Member",
"response": "Answer to Reviewer Ziwei Dai … There is space to improve presentation of the results as well. Specific comments are listed below for the author’s reference. Thank you for your constructive criticism. Your point is well taken and I followed your suggestions. Your remarks were very helpful to improve the presentation, as specified below. Some important interactions that may affect FBP dynamics are missing in the model. For instance, it is known that FBP is an allosteric activator of pyruvate kinase (Jurica et al ; Christofk et al ). I expect that this feedback will attenuate the FBP buffering mechanism proposed in this study since lower glycolysis (i.e. 'tail' in this model) is activated by high concentration of FBP thus enhancing its consumption. Equation 2 of the model implies that the flux in the tail section increases with FBP. This reflects not only that FBP is a substrate for the tail end of glycolysis (via aldolase), but also reflects a potential activating role of FBP on pyruvate kinase. To examine multiple roles of FBP, I tested various forms of the mathematical relation between FBP and flux in the tail section, among others by using powers of FBP (Hill coefficient), but these attempts did not improve simulations. Note that the tail flux turns out to be sensitive to extremely low levels of FBP, with KFBP,tail estimated to be 0.24 μM, a level that is reached almost immediately after reintroduction of glucose in the experiment of Fig. 2A (see Dataset 1). This agrees with activation of pyruvate kinase by FBP in the micromolar range (1). This also agrees with the very fast response of lactate production immediately after reintroduction of glucose, found experimentally and reproduced by the model (Figures 2A and 2B). The activating effect of FBP on tail flux is therefore taken into account in the model. I have added discussion of this point and the references you gave to Jurica and Christofk to the main text (Methods, seventh paragraph) and Supplementary Text (paragraph around Eq. 2). Moreover, the two sections (‘head’ and ‘tail’) of glycolysis are both considered to be irreversible, thus neglecting the effects of thermodynamics on glycolytic flux. This may also lead to overestimation of FBP concentration as well because FBP is not allowed to be converted back to glucose. Your fellow reviewer, Dr Dash gave a similar comment. I have therefore addressed this in a joint answer which is also included in the answer to Dr Dash: Glycolysis is considered by biochemists to be essentially irreversible, in particular the hexokinase, phosphofructokinase and pyruvate kinase reactions (see for instance the textbook by Berg, Tymocko, Stryer. Biochemistry, Sixth Edition, 2007, e.g. page 460, ref. (2)). In addition the glyceraldehyde 3-phosphate dehydrogenase reaction may also be an important limiting step in cells showing the Warburg effect (3, 4). To produce glucose in the reverse direction (fructose 1,6-bisphosphate to glucose) the enzyme glucose 6-phosphatase is required to bypass the irreversible hexokinase reaction. Glucose 6-phosphatase’s activity in Ehrlich ascites cells is low (5), while its activity in the liver, which can produce glucose, is high. The tail section of glycolysis is also considered to be virtually irreversible and two additional enzymes (pyruvate carboxylase and phoshoenolpyruvate carboxykinase) are required to enable conversion of pyruvate in the direction of glucose, bypassing pyruvate kinase (see (2), p. 460-463). I had implemented two existing detailed models of the glycolytic chain and used these to explore the reversibility of the reactions. The first model was based on the model of Lambeth and Kushmerick (6), extended with the equation for hexokinase given by Mulquiney and Kuchel (7). The second model was based on the model of Marin-Hernandez et al. (8) for cancer cells. Both models represent the ten enzymatic steps in the glycolytic chain and incorporates both the forward and reverse reactions for all the steps. All these equations take the Haldane relation into account, and constraints based on thermodynamics are obeyed. Simulations with both models showed that at least one reaction in the head as well as the tail section of the glycolytic chain showed reverse fluxes that were extremely small relative to the forward fluxes. The results are discussed in more detail in the revised Supplementary Text, section titled “Effect of Reverse Fluxes on the Glycolytic Model”. These simulations illustrate the biochemical concept that glycolysis is essentially irreversible. In order to reduce the model as much as possible, the reverse reactions in glycolysis were therefore neglected. However, the lactate dehydrogenase reaction is included in fully reversible form and lactate is taken up for mitochondrial metabolism if glucose is depleted in the simulations. Mitochondrial oxidative phosphorylation coupled to oxygen consumption is also modeled as being essentially irreversible, as generally accepted by biochemists. However, these considerations point to a limitation: the model is not suitable for determining the magnitude of small reversed fluxes in the glycolytic chain under conditions where glycolytic products (pyruvate, NADH, ATP) are high and glucose concentration is very low. Reverse fluxes may be revealed by isotope exchange experiments. These considerations are now included in the Methods section of the main paper (8th paragraph) and the Supplementary Text (section title “Effect of Reverse Fluxes on the Glycolytic Model”). The main goal of developing this coarse-grained model is not clear to me. Besides the model developed by Chance and Hess, there are numerous mathematical models for glycolysis, most are much more detailed than the model presented in this study. The author claims that this model replaces the model of Chance and Hess – I feel this inappropriate since this statement doesn’t give any credit to all other glycolysis models. There are indeed many mathematical models for glycolysis. I referred to several of these models in the initial submission: the model by Chance and Hess dating back to 1959 (9, 10), the model of Lambeth and Kushmerick (6) for muscle, the model of Mulquiney and Kuchel (7) for red blood cells and the model of Marin-Hernandez et al. (8) for cancer cells were all referenced. Further the model studies of Van Eunen et al. (11), Chandra et al. (12) and Van Heerden et al. (13) on yeast metabolism were referenced. I now have added additional references to model analyses by Shestov et al. (3) and Liberti et al. (4). Although this is not a comprehensive list, I reckon these references (and the references cited therein) probably give sufficient background on mathematical models of glycolysis. I had in fact originally implemented two separate detailed models of glycolysis: one developed specifically for cancer cell metabolism by Marin-Hernandez et al. (8) and another one based on the model of Lambeth and Kushmerick (6), extended with the equation for hexokinase by Mulquiney and Kuchel (7). Both glycolytic models represent all the ten enzymatic steps in the glycolytic chain and incorporate both the forward and reverse reaction for all these steps, taking the Haldane relation and thermodynamic constraints into account. Despite simulation attempts with both detailed models, not all features of the experimental response of Ehrlich ascites cells could be reproduced well. The measured overshoot of fructose 1,6-bisphosphate (FBP) after adding a high concentration of glucose to EATC was for instance not reproduced with these detailed models, even when additional regulatory loops were added. The coarse-grained model described in this manuscript performed much better and depends on far fewer unknown parameters than the detailed models which contain about 55 and 70 parameters, respectively, for the glycolytic chain alone. Detailed measurements for all separate glycolytic enzymes in the EATC are not available. In contrast, the system level response is well defined by measurements of key nodes in the system: glucose uptake (input head part of glycolysis), fructose 1,6-bisphosphate (output head part of glycolysis and input tail part of glycolysis), lactate (output tail part of glycolysis), respiration, ATP etc. making parametrization of the coarse-grained model feasible. I therefore employed the course-grained model for the purposes of this study. Because a broad range of experimental results in vitro are well represented, the model can subsequently be used to explore the role of metabolism in vivo during fluctuating nutrient and oxygen supply. I understand that the term ‘replace’ in relation to the model of Chance and Hess may be felt to be inappropriate. The Chance-Hess model is ingenious and of huge historical interest, because it probably is the first digital model of a biochemical reaction system ever published. This is emphasized in the main text (Introduction, first paragraph) and Supplementary Text (section “Short Description of the Computational Model”). Their goal was to explain the metabolic response of the same Ehrlich ascites tumor cells that were investigated in the present study. The phrase containing ‘replace’ has now been removed from the manuscript. What was meant is simply that a new model was used that represents the metabolic responses of EATC to glucose and oxygen levels well and does not contain biochemically untenable assumption. I had in fact also implemented the original model of glycolysis and oxidative phosphorylation by Chance and Hess developed to explain their data on Ehrlich ascites tumor cells, which appeared unsuitable for accurate description of the experimental data of Figure 2A. One key assumption in the Chance-Hess model was that newly synthesized ATP is retained in the mitochondria and is not quickly available for cytosolic processes. Their mitochondrial ATP synthesis mechanism did not comply with the chemiosmotic hypothesis. These assumptions in the Chance-Hess model are unfortunately incompatible with present biochemical knowledge. Nevertheless, the model of Chance and Hess was of course a very meritorious historical step in the computational analysis of biochemical systems. You will now find these issues discussed more extensively in the main text (Introduction; Methods, first paragraph and third paragraph) and Supplementary Material (first page and new section “Reasons for Employing a Coarse-grained Rather than Detailed Model of Glycolysis”). Most, if not all, results are presented as curves from the simulation with little information about the take-home messages. It is thus very difficult to read the key findings directly from the figures. For instance, it is my understanding that one of the two most important hypotheses drawn based on the simulation is that cells with high glycolytic capacity (which mimics ‘Warburgian’ cancer cells) consume glucose much more quickly than cells with lower glycolytic capacity after switching from low glucose to high glucose condition, thus being more competitive under conditions with frequent nutrient deprivation. To emphasize this point, I would recommend using one figure directly comparing glucose uptake fluxes in cells with different glycolytic capacities instead of the 10 figure panels currently included in Fig 4. I followed your excellent suggestion and added a figure comparing the glucose uptake fluxes in the two cell types (new Figure 5) to illustrate this important message. I considered it helpful to retain the original Figure 4 to show effects on FBP levels and ATP synthesis, among others because it illustrates another important take-home message regarding the role of the ATP buffering mechanism in maintaining energy homeostasis. I also added extensive clarifications about the take home messages at many points in the text of the Results section. REFERENCES 1. H. R. Christofk, M. G. Vander Heiden, N. Wu, J. M. Asara, L. C. Cantley, Pyruvate kinase M2 is a phosphotyrosine-binding protein. Nature 452, 181-186 (2008). 2. J. M. Berg, J. L. Tymoczko, L. Stryer, Biochemistry. (W.H. Freeman and Company, ed. Sixth, 2007). 3. A. A. Shestov, X. Liu, Z. Ser, A. A. Cluntun, Y. P. Hung, L. Huang, D. Kim, A. Le, G. Yellen, J. G. Albeck, J. W. Locasale, Quantitative determinants of aerobic glycolysis identify flux through the enzyme GAPDH as a limiting step. Elife 3, (2014). 4. M. V. Liberti, Z. Dai, S. E. Wardell, J. A. Baccile, X. Liu, X. Gao, R. Baldi, M. Mehrmohamadi, M. O. Johnson, N. S. Madhukar, A. A. Shestov, I. I. C. Chio, O. Elemento, J. C. Rathmell, F. C. Schroeder, D. P. McDonnell, J. W. Locasale, A Predictive Model for Selective Targeting of the Warburg Effect through GAPDH Inhibition with a Natural Product. Cell Metab 26, 648-659 e648 (2017). 5. W. J. Nelson, S. J. Nelson, P. Traub, Comparison of Ehrlich ascites tumour and mouse liver cells by analytical subcellular fractionation combined with a sensitive computational method for data analysis. Hoppe-Seyler's Z. Physiol. Chem. 362, 903-918 (1981). 6. M. J. Lambeth, M. J. Kushmerick, A computational model for glycogenolysis in skeletal muscle. Annals of Biomedical Engineering 30, 808-827 (2002). 7. P. J. Mulquiney, P. W. Kuchel, Model of 2,3-bisphosphoglycerate metabolism in the human erythrocyte based on detailed enzyme kinetic equations: equations and parameter refinement. Biochem J 342, 581-596 (1999). 8. A. Marin-Hernandez, J. C. Gallardo-Perez, S. Rodriguez-Enriquez, R. Encalada, R. Moreno-Sanchez, E. Saavedra, Modeling cancer glycolysis. Biochim Biophys Acta 1807, 755-767 (2011). 9. B. Chance, B. Hess, Spectroscopic evidence of metabolic control. Science 129, 700-708 (1959). 10. B. Chance, D. Garfinkel, J. Higgins, B. Hess, A solution for the equations representing interaction between glycolysis and respiration in ascites tumor cells. Journal of Biological Chemistry 235, 2426-2439 (1960). 11. K. van Eunen, J. A. Kiewiet, H. V. Westerhoff, B. M. Bakker, Testing biochemistry revisited: how in vivo metabolism can be understood from in vitro enzyme kinetics. PLoS Comput Biol 8, e1002483 (2012). 12. F. A. Chandra, G. Buzi, J. C. Doyle, Glycolytic oscillations and limits on robust efficiency. Science 333, 187-192 (2011). 13. J. H. van Heerden, M. T. Wortel, F. J. Bruggeman, J. J. Heijnen, Y. J. Bollen, R. Planque, J. Hulshof, T. G. O'Toole, S. A. Wahl, B. Teusink, Lost in transition: start-up of glycolysis yields subpopulations of nongrowing cells. Science 343, 1245114 (2014)."
}
]
}
] | 1
|
https://f1000research.com/articles/7-1177
|
https://f1000research.com/articles/7-1986/v1
|
28 Dec 18
|
{
"type": "Research Article",
"title": "Two parallel reporting systems for malaria surveillance in Pakistan, 2013–17: is exact burden reflected?",
"authors": [
"Hammad Habib",
"Razia Fatima",
"Abdul Baseer Achakzai",
"Ahmad Wali",
"Aashifa Yaqoob",
"Hina Najmi",
"Mahboob Ul Haq",
"Abdul Majeed",
"Razia Fatima",
"Abdul Baseer Achakzai",
"Ahmad Wali",
"Aashifa Yaqoob",
"Hina Najmi",
"Mahboob Ul Haq",
"Abdul Majeed"
],
"abstract": "Background: Pakistan is facing challenges regarding the availability of reliable data for malaria surveillance. These include lack of coordination between different reporting systems and fragmented information system. This study aimed to compare the reporting of malaria surveillance systems in Pakistan. Methods: There are two parallel reporting systems for malaria surveillance in Pakistan, the District Health Information System (DHIS) and Malaria Information System (MIS). DHIS reports on all morbidity at health facility level, while MIS is only used for malaria surveillance in the donor supported districts. A cross sectional study was conducted between July-September 2018 by using the retrospective records of DHIS and MIS data reported to the Directorate of Malaria Control (DOMC) Islamabad during 2013-17. Descriptive and inferential analysis was performed to compare the coverage, outcome and impact indicators. Results: During 2013-17, all districts (n=145, 100%) across Pakistan reported on the DHIS. The MIS reporting coverage has gradually increased from 21 (14.5%) to 72 (49.7%) districts. Reported number of suspected screened and confirmed malaria cases were compared. MIS reported twice the number of suspects screened for malaria (100.5%) and confirmed malaria cases (124.4%) as compared to the DHIS. The difference in the reported average annual blood examination rate (ABER) was 3.8, test positivity rate (TPR) was -0.9 and the annual parasite incidence (API) was 4.9/1000 population over five years between two systems. DHIS reported only half the ABER and API as compared to MIS. Conclusion: There is huge under-reporting of suspected and confirmed malaria cases in the DHIS as compared to MIS. Urgent attention is needed to address this, as it is vital to have uniform reporting of true disease burden across the country. An integrated disease surveillance system, improved data validation systems, and use of the online DHIS-2 are potential options for better integrity and coherence of reported data.",
"keywords": [
"Malaria surveillance",
"comparison",
"operational research",
"DHIS",
"Pakistan",
"reporting system"
],
"content": "Introduction\n\nMalaria ranks sixth amongst the top ten causes of deaths in low income countries of the world1. In 2016, 91 countries reported an increase in malaria cases (216 million) as compared to 2015 (211 million). Around 0.4 million deaths have also been reported during the same year. Most cases were reported by the World Health Organization (WHO) African Region (90%), followed by the South-East Asia Region (7%), and the Eastern Mediterranean Region (2%)2. The incidence of malaria cases varies from low to high in different countries. Countries with low malaria incidence are progressing towards malaria elimination, while others having a high disease burden, including Pakistan, have implemented malaria control programs3.\n\nThe World Health Organization (WHO) has emphasized the critical need for transforming malaria surveillance as a core intervention in the Global Technical Strategy for Malaria. WHO has further stressed on the importance of prioritizing investments in malaria surveillance system to ensure that reliable data is available for decision making4. Effective surveillance of malaria is essential for identifying and prioritizing the most affected areas or population groups5. Moreover, uniformity of the surveillance tools and timeliness of reporting is important in countries with low disease burden at the malaria control phase6. It has been observed that despite all the efforts and investments, malaria surveillance system has many challenges related to the timeliness, representativeness, data quality and reliability in high and low burden countries from the WHO African and Eastern Mediterranean regions7,8.\n\nPakistan is among seven countries of the WHO Eastern Mediterranean Region sharing 95% of the regional malaria burden9. An estimated 98% of Pakistan population (205 million) is at varying risk, while around 60% population (123 million) at high risk for malaria2. In Pakistan, Malaria due to Plasmodium vivax is most common (88%), followed by Plasmodium falciparum (12%)10. Epidemiologically, Pakistan is classified as a moderate malaria endemic country with the national annual parasite incidence (API) averaging at 1.16, with a high variation within different provinces of Pakistan11. The districts and agencies in the Five provinces and Federally Administered Tribal Areas (FATA) region of Pakistan were stratified into three strata (I, II and III) based on the malaria annual parasite incidence (API), and slide positivity rate (SPR) of 2011–13 under the country’s National Strategic Plan. Based on this stratification, 72 districts are placed in stratum-I (having API >5), ten in stratum-II (API 1–5) and 63 in stratum-III (API <1) for prioritizing the highest endemic districts for resource allocation12.\n\nThe malaria surveillance in the country has many issues. Major challenges include lack of coordination between different reporting systems, fragmented information systems and relying on parallel reporting for malaria cases in the highest burden sharing districts through a paper based malaria information system (MIS) and district health information system (DHIS)13. The diverse epidemiology of malaria disease in Pakistan stratifying the country into high and low burden sharing areas has further contributed to the difficulties in proper disease surveillance12. It is estimated that due to such challenges, only around 23% of malaria cases have been captured in Pakistan through various surveillance systems during 20162.\n\nLimited evidence has been found regarding the comparison of various malaria surveillance systems in Pakistan. This study aims to compare the malaria coverage, screening, cases, outcome and impact as reported through the MIS and DHIS in high burden sharing districts of Pakistan.\n\n\nMethods\n\nThis was a cross sectional retrospective record review of malaria routine surveillance data for the period 2013–17 reported through the DHIS and MIS from donor supported districts of Pakistan. The study was conducted from July to September 2018 at the Directorate of Malaria Control (DOMC), Islamabad which is an attached department of the Ministry of National Health Services, Regulations and Coordination. DOMC is primarily responsible for malaria surveillance in Pakistan in collaboration with the provincial malaria control programs. The Global Fund (donor) is supporting the malaria control interventions in the highest burden sharing stratum-I districts located mainly in the provinces of Balochistan, Sindh, Khyber Pakhtunkhwa (KP) and FATA12.\n\nMalaria surveillance data from the public health facilities is reported through two parallel systems, namely the DHIS (introduced in 2008 for all districts) and MIS (only for the Global Fund supported districts). DHIS reports on all morbidity at health facility level while MIS is used in only the donor supported districts for malaria surveillance13. Fever cases with signs and symptoms of malaria are screened as suspected cases, confirmed through the microscopy or rapid diagnostic test (RDT) kits, and then reported in the DHIS and MIS according to the identified species, i.e. P. falciparum, P. vivax or mix cases on monthly basis (Figure 1). Data of five years as reported by the districts having both DHIS and MIS simultaneously from 2013 to 2017 was used for comparing the annual blood examination rate (ABER), test positivity rate (TPR) and annual parasite incidence (API).\n\nAll malaria cases reported to the Directorate of Malaria Control (DOMC) through DHIS and MIS between 1st January 2013 and 31st December 2017 from the districts where parallel reporting on both DHIS and MIS has been used.\n\nEpidemiological records of five years are available in electronic forms at the Directorate of Malaria Control in Islamabad. The data of this particular study was extracted from the DHIS and MIS. Study variables including the suspected cases screened for malaria, and confirmed malaria cases disaggregated by species (P. falciparum, P. vivax and mix) were doubled entered, and cleaned in the EpiData Entry version 4.4.3.1.\n\nDescriptive analysis of selected variables such as reporting coverage of the two systems, annual blood examination rate, test positivity rates and reported annual malaria cases was performed in SPSS version 23. The DHIS and MIS data is regularly validated on monthly basis for the DHIS, and quarterly basis for the MIS at the district, provincial, and federal levels. Randomly, the hard copies of the reported data were matched with the entered data for validation.\n\nAs this study was conducted on two malaria surveillance systems comprising of aggregated districts level data, there was no human subject directly involved in this study. Ethical and administrative approval (Reference F.No.2-30/2018/CMU-NFR; M&E/Surveillance/SORT-IT) was taken from the Director, DOMC for using the malaria program data for this study.\n\n\nResults\n\nDuring 2013–17, all districts (n=145, 100%) across Pakistan reported on DHIS. Reporting on the MIS increased gradually from 21 (14.5%) districts in 2013 to 72 (49.7%) districts in 2017. (Figure 2) For this study, data of only those districts was analyzed which had reported simultaneously on the DHIS and MIS during these five years. (Table 1)\n\nDHIS=District Health Information System, MIS=Malaria Information System, API=Annual Parasite incidence, TGF= The Global Fund, *MIS-reported API= The cases include cumulative figures for only the Global Fund supported districts for 2013–2017, *DHIS-reported API= The cases include cumulative figures for only the Global Fund supported districts for 2013–2017.\n\nAJK= Azad Jammu Kashmir, FATA= Federally Administered Tribal Areas, KP= Khyber Pakhtunkhwa, DHIS= District Health Information System, MIS= Malaria Information System\n\nFor screening the suspected malaria cases, DHIS reported a total of 4,260,610 suspected cases screened for malaria. During the same period, MIS reported a total of 8,540,702 suspected cases screened for malaria, which was 4,280,092 (100.5%) more than the DHIS from the same districts. The reported number of total confirmed malaria cases in the DHIS was 436,273. Out of these, P. vivax was the highest reported cause of malaria with 350,892 (80.4%), followed by P. falciparum (80,230, 18.4%) and mixed infection with 4,697 (1.1%). MIS reported 979,192 confirmed malaria cases during the same period for the same districts. P. vivax was the most reported cause of malaria (769,016, 78.5%), followed by P. falciparum (150,398, 15.4%) and mixed infection with 59,778 (6.1%). MIS reported 542,919 (124.4%) confirmed malaria cases more than the DHIS. (Table 2)\n\nTGF= The Global Fund, ABER= Annual Blood Examination Rate, API= Annual Parasite Incidence, Confirmed cases = Number of tested suspects confirmed as malaria cases, DHIS= District Health Information System, MIS= Malaria Information System, Suspects screened= Number of fever cases tested for malaria, TPR= Test Positivity Rate\n\nDHIS reported an average annual blood examination rate (ABER) of 3.3, test positivity rate (TPR) of 13.6, and annual parasite incidence (API) of 4.4 for the five years. MIS reported an average ABER of 7.1, TPR of 12.7 and API of 9.3. The difference in the reported ABER was 3.8 (115.2%), TPR was -0.9 (6.6%) and API was 4.9 (111.4%). (Table 2)\n\n\nDiscussion\n\nPakistan like other low and middle-income countries (LMICs) has substantial reliance on external funding14. The Global Fund providing 50 percent of all international financing for malaria, has been supporting the Government of Pakistan for the control of TB, AIDS and malaria since 20033. This support has been mainly targeted for decreasing the burden of disease in the highest endemic districts of the country through the provision of prompt diagnostic, treatment, and preventive services for malaria13. Keeping in view the importance of surveillance, the Ministry of Health in Pakistan has taken strengthening the Health Information Systems as a major thematic pillar under the 10 years National Health Vision15. However, studies in Pakistan have highlighted issues related to the data reliability, and concrete data analysis in various health programs16.\n\nDisease reporting in Pakistan across all provinces and at the federal level is carried out through the DHIS and several other parallel surveillance systems for the diseases according to the specific needs of the programs. Some of these parallel systems include Malaria, Dengue, TB, EPI, and HIV/AIDS. DHIS is the nation-wide health information system which is being used for reporting of the district level aggregated data from all the public health facilities17.\n\nOur study shows that reporting coverage through the MIS, which is mainly used for reporting malaria in high burden districts of the country, has been increasing gradually from 21 to 72 districts from 2013–17. Balochistan and FATA provinces have the highest average number of districts reporting on both DHIS and MIS, which may be due to the highest annual parasite incidence in these provinces; hence they are supported by donor funds by the national program, and prioritized for malaria control interventions18.\n\nThis study reveals major differences in the numbers of suspected screening for malaria and confirmed malaria cases reported between the two surveillance systems. A consistent under-reporting of suspected and confirmed malaria cases was seen in the DHIS as compared to MIS. This is contradictory to the findings of another recent study conducted in Swaziland comparing three reporting systems in malaria elimination settings where the national reporting system was found to be over-reporting malaria cases19. Although both the DHIS and MIS are paper based surveillance systems, the difference in reported number of suspected cases screened and malaria confirmed cases observed from this study is very high. The reported confirmed malaria cases in MIS are 121.9% more as compared to the DHIS. Another study conducted in Malawi on data quality has shown discrepancy of 12–24% between paper-based and electronic data systems20. Ideally, there should be no variance between the two surveillance systems as the reported data is from same health facilities within the same districts.\n\nThe huge difference and under-reporting of malaria figures from DHIS may be due to several reasons. First, the start of donor support for malaria control interventions in the country has brought with it an additional reporting system to the DHIS, i.e., the MIS which is more comprehensive, and has many additional indicators as per the donor requirement. The MIS has its own data recording and reporting tools at the health facility level in parallel to already existing tools for DHIS. When there are parallel reporting systems with different recording and reporting tools in the same health facilities, the data quality is usually compromised as entering the data into separate platforms results in more errors19.\n\nThe second important factor to be explored will be that of who is responsible for data entry and reporting into DHIS and MIS at the very basic level, i.e. the health facility and district levels. Thirdly, there was considerable under-reporting of confirmed malaria cases in the DHIS. A possible reason may be that the tools in the DHIS are more specific for screening of malaria suspects with microscopy, which is the gold standard for confirmed diagnosis of malaria cases. However, DOMC with the support of the Global Fund grant, introduced RDT kits in health facilities without microscopy diagnosis facilities in the high burden districts, mostly in Basic Health Units (BHUs). This may have resulted in screening of more suspects using RDT at these health facilities and reporting of more confirmed cases. The cases screened and confirmed through RDTs may not have been reported into the DHIS by many of these centers as RDT is a relatively new diagnostic method21.\n\nThe study strengths are that the national surveillance data reported to routine program settings was used for analysis which is likely to reflect the reality on the ground. All districts reporting simultaneously on DHIS and MIS over the span of five years were included in the study across all provinces of Pakistan. The data reported in the MIS has been used as the benchmark as this database is carefully supervised and validated at the district level on a quarterly basis. The Global Fund grant has comprehensive data validation and quality assurance mechanisms in place hence we believe that the MIS data is more reliable22.\n\nThe current study is limited by the fact we did not have facility level data for analysis and we did not explore the exact reasons for observed discrepancies between the two surveillance systems. In-depth interviews of the district level supervisors and data entry operators for the DHIS and MIS along with the health facility staff actively involved in reporting on the two systems may help in better understanding of the dynamics and reasons for these discrepancies.\n\nDespite of these limitations, the findings have a number of policy and practice implications. In the context of devolution in Pakistan, DHIS has utmost importance for timely surveillance of communicable diseases at the provincial level. MIS, being donor driven, is present in only around half of the country. The major concern in regards of the under-reporting seen in the DHIS is that the true malaria burden may be under-estimated. This under-reporting of the confirmed malaria cases can delay the early detection of malaria outbreaks predisposing the population to malaria epidemic. Moreover, the data from DHIS is used for decision making, disease prioritization and resource allocation according to the reported disease burden in the provinces. Under-reporting from DHIS will result in reporting of false burden of malaria cases, i.e. fewer cases than actual. This may result in lesser disease prioritization due to false reporting of low disease burden and hence less resource allocation.\n\nFurther research is required at the health facility, district and provincial levels to assess whether there is any mechanism for data validation, combined reporting, comparison and coherence for the DHIS and MIS data before being finalized for reporting into the two parallel systems. This is important as continuous validation of the health facility data is important before reporting into various systems23. The DHIS-2 being a free and open source, web based software may be the potential option for improving the completeness and quality of surveillance data being reported for malaria surveillance24,25.\n\n\nData availability\n\nOpen Science Framework: Pakistan malaria surveillance data 2013–17 (DHIS-&-MIS), https://doi.org/10.17605/OSF.IO/NC54V26.\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\nThis research was conducted through the Structured Operational Research and Training Initiative (SORT IT), a global partnership led by the Special Programme for Research and Training in Tropical Diseases at the World Health Organization (WHO/TDR). The training model is based on a course developed jointly by the International Union Against Tuberculosis and Lung Disease (The Union, Paris, France) and Médecins Sans Frontières (MSF, Geneva, Switzerland). The specific SORT IT programme that resulted in this publication was implemented by the National Tuberculosis Control Programme of Pakistan, through the support of the Global Fund to Fight AIDS, Tuberculosis and Malaria (The Global Fund, Geneva, Switzerland). The publication fee was covered by the Special Programme for Research and Training in Tropical Diseases at the World Health Organization (WHO/TDR).\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nWorld Health Organization: The top 10 causes of death. 2018; Accessed Oct 06, 2018. Reference Source\n\nWorld Health Organization: World Malaria Report. 2017. Reference Source\n\nThe Global Fund: Malaria: The Global Fund to Fight AIDS, Tuberculosis and Malaria. 2018; Accessed Oct 06, 2018. Reference Source\n\nWorld Health Organization: Global Technical Strategy for Malaria 2016-2030. 2015; 9–11. Reference Source\n\nWorld Health Organization: Health Metrics Network: Framework and Standards for Country Health Information Systems. 2012; Accesed Oct 06, 2018. Reference Source\n\nWorld Health Organization: Disease surveillance for malaria control; An operational manual. 2012; Accesed Oct 07, 2018. Reference Source\n\nIbrahim BS, Abubakar AA, Bajoga UA, et al.: Evaluation of the Malaria Surveillance System in Kaduna State, Nigeria 2016. Online J Public Health Inform. 2017; 9(1):e177. Publisher Full Text | Free Full Text\n\nWorld Health Organization: Ninth intercountry meeting of national malaria programme managers from HANMAT and PIAM-Net countries. 2017; Accesed Oct 06, 2018. Reference Source\n\nWorld Health Organization: Regional profile: Eastern Mediterranean Region. World Malaria Report. 2017; 86–87. Reference Source\n\nKhattak AA, Venkatesan M, Nadeem MF, et al.: Prevalence and distribution of human plasmodium infection in Pakistan. Malar J. 2013; 12(1): 297. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDirectorate of Malaria Control: Malaria Annual Report. 2016. Reference Source\n\nDirectorate of Malaria Control: Strategic Plan Malaria Control Program Pakistan (2015-2020). 2015; Accesed Oct 07, 2018. Reference Source\n\nDirectorate of Malaria Control: Pakistan Malaria Programmae Review. 2016; Acessed Oct 08, 2018. Reference Source\n\nKhan MS, Meghani A, Liverani M, et al.: How do external donors influence national health policy processes? Experiences of domestic policy actors in Cambodia and Pakistan. Health Policy Plan. 2018; 33(2): 215–23. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMinistry of National Health Services Regulation and Coordination: National Health Vision Pakistan 2016-2025. 2016; 1–17. Reference Source\n\nNishtar S, Boerma T, Amjad S, et al.: Pakistan's health system: performance and prospects after the 18th Constitutional Amendment. Lancet. 2013; 381(9884): 2193–206. PubMed Abstract | Publisher Full Text\n\nSabih F, Bile KM, Buehler W, et al.: Implementing the district health system in the framework of primary health care in Pakistan: can the evolving reforms enhance the pace towards the millennium development goals? East Mediterr Health J. 2010; 16 Suppl: S132–44. PubMed Abstract | Publisher Full Text\n\nKakar Q, Khan MA, Bile KM: Malaria control in Pakistan: new tools at hand but challenging epidemiological realities. East Mediterr Health J. 2010; 16 Suppl: S54–60. PubMed Abstract | Publisher Full Text\n\nZulu Z, Kunene S, Mkhonta N, et al.: Three parallel information systems for malaria elimination in Swaziland, 2010-2015: are the numbers the same? Public Health Action. 2018; 8(Suppl 1): S13–S17. PubMed Abstract | Publisher Full Text | Free Full Text\n\nGadabu OJ, Munthali CV, Zachariah R, et al.: Is transcription of data on antiretroviral treatment from electronic to paper-based registers reliable in Malawi? Public Health Action. 2011; 1(1): 10–12. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorld Health Organization: Diagnostic testing and treatment of malaria. World Malaria Report. 2011; 39–43. Reference Source\n\nThe Global Fund: The Global Fund’s approach to monitoring and evaluation. 2016; Acessed Oct 08, 2018. Reference Source\n\nOduro AR, Maya ET, Akazili J, et al.: Monitoring malaria using health facility based surveys: challenges and limitations. BMC Public Health. 2016; 16: 354. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKiberu VM, Matovu JK, Makumbi F, et al.: Strengthening district-based health reporting through the district health management information software system: the Ugandan experience. BMC Med Inform Decis Mak. 2014; 14: 40. PubMed Abstract | Publisher Full Text | Free Full Text\n\nKaruri J, Waiganjo P, Orwaa D, et al.: DHIS2: The Tool to Improve Health Data Demand and Use in Kenya. J Health Inform Dev Ctries. 2014; 8(1): 38–60. Reference Source\n\nHabib H: Pakistan Malaria Surveillance Data 2013-17 (DHIS-&-MIS). OSF. 2018. http://www.doi.org/10.17605/OSF.IO/NC54V"
}
|
[
{
"id": "48960",
"date": "12 Jun 2019",
"name": "Lindsey Wu",
"expertise": [
"Reviewer Expertise Malaria epidemiology and surveillance"
],
"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 appreciated reading this paper by Habib and colleagues on a topic that is timely as many countries are assessing the utility of different platforms for routine malaria surveillance data and, importantly, how best to harmonise them. Overall, the authors demonstrate that the MIS data is more reliable than the DHIS system, which they attribute to the comprehensive data validation and quality assurance mechanisms embedded into Global Fund grants. They conclude that further research is required, but it is seems clear that the MIS platform is advantageous. Rather than conducting further research, what is likely to be more useful now are developing more detailed recommendations on lessons learned from the MIS system and how this could be adopted or incorporated into national platforms once donor funding ends.\nSpecific areas that would benefit from additional clarification or modifications in data presentation include:\nIntroduction:\n“Stratifying the country into high and low burden sharing areas has further contributed to the difficulties in proper disease surveillance.” – Some explanation as to why and in what ways this has caused issues would be helpful.\n\n“Limited evidence has been found regarding the comparison of various malaria surveillance systems in Pakistan.” – Is this because there have not been parallel reporting systems in the country until recently and have there been similar assessments conducted in other countries?\nMethods:\nThis study was conducted between July and September – was there any influence of the season on the survey? “Fever cases with sign and symptoms of malaria area screen as suspected cases, confirmed through microscopy and rapid diagnostic test (RDT) kits.” – Is there any prior evidence of discordance or different reporting rates between the two methods in the country and how might this be accounted for in the analysis?\nResults:\nIs the reason that MIS reporting coverage has gradually increased due to the fact that the MIS is targeted and this indicates that burden has increased in these districts? Some details on how Global Fund MIS-supported districts are selected should be included. For instance, what are the risk-strata used to select? Figure 1 - It would be helpful to include more detail to address whether there are potential structural explanations for the discordance between reporting systems. For example, are there separate forms in the districts with both systems? Figure 2 – The data presented here may be more clear as a bar chart indicating the proportion or percent of MIS-reporting districts out of the total Table 1:\nThe data in this table would benefit from being reported by risk strata or endemicity. Is there are reason for the lack of MIS data in AJK and Punjab regions? Zero reporting data can be represented with a dashed line instead.\n\nTable 2:\nDo the reporting discrepancies differ by province/region or over time, given that the number of MIS-reporting districts gradually increases over the period of the study? The methods used to calculated ABER and TPR should be detailed in the footnotes, as well. There should be further details in the methods regarding how cases are confirmed. Percent difference is a somewhat unclear way to present the differences, as their comparison between different metrics is not easily interpreted. In the Table title, “impact in donor supported districts”. It should be clarified that this study is in places with both systems. Similarly, for the text “During 2013-2017, all districts across Pakistan reported on DHIS”, it should be clarified that this is only in districts which also had the MIS system.\n\nDiscussion:\n“…there is a huge-under reporting” – “Huge” should be expressed as “significant”, with the appropriate statistical tests to justify this statement conducted. “This study reveals major differences in the numbers of suspected screening for malaria and confirmed malaria cases” – It would be good to clarify whether this underreporting is due to data not being entered and reported. “When there are parallel reporting systems with different recording and reporting tools in the same health facilities, the data quality is usually compromised as entering the data into separate platforms results in more errors.” – Some discussion by the authors about the specifics of the data forms/platforms that might result in biased reporting would be useful here. “The second important factor to be explored will be that of who is responsible for data entry and reporting into DHIS and MIS at the very basic level” – Similar to the comment above, details or examples of staffing differences that might lead to systemic differences in reporting would be useful for the reader here. “DOMC introduced RDT kits in health facilities without microscopy diagnosis facilities in the high burden districts…This may have resulted in screening of more suspects using RDT” – Can the authors clarify which year was it introduced and whether this coincides with any temporal variation between DHIS/MIS discordance?\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? No source data required\n\nAre the conclusions drawn adequately supported by the results? Partly",
"responses": []
},
{
"id": "51672",
"date": "13 Aug 2019",
"name": "Mrigendra P Singh",
"expertise": [
"Reviewer Expertise Epidemiology",
"Malaria and other vector borne 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\nRecord of malaria cases was reviewed for the period of 2013-2017 retrospectively. Therefore the statement \"a cross sectional study conducted between July - September\" is confusing and need to be deleted.\nDescriptive analysis to compare the coverage and reported suspected and confirmed malaria cases either by microscopy or by RDT was performed. No any inferential statistics was performed to analyze the relation between outcome and impact indicators. Therefore, this line should be corrected.\nAuthors did not mentioned the methods of surveillance for malaria screening in DHIS and MIS reporting system. As per our understanding DHIS is a passive and MIS is an active surveillance system. Patients with any ailments visited to the health facilities were reported in DHIS if they have febrile illness (Passive) and under MIS, specific field staff particularly engaged for malaria survey visited to the villages/households (door to door visit) for screening of suspected malaria cases (Active) and if it is true then analysis, discussion and conclusion should be revised accordingly. The previously published literature (Singh et al. 2016)1 mentioned that less cases reported in passive surveillance as compared to active surveillance system.\n\nPopulation coverage during the reported period is not mentioned whereas API and ABER indices are dependent with population coverage.\nTable 1: Zero showed that there was no cases reported instead of no survey was conducted. It should be replaced with DASH or NA (information not available).\nTable 2: A separate column for population covered in DHIS and MIS districts should be mentioned for more detail understanding.\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? No",
"responses": []
}
] | 1
|
https://f1000research.com/articles/7-1986
|
https://f1000research.com/articles/7-1985/v1
|
28 Dec 18
|
{
"type": "Research Article",
"title": "A bovine CD18 signal peptide variant with increased binding activity to Mannheimia hemolytica leukotoxin",
"authors": [
"Aspen M. Workman",
"Carol G. Chitko-McKown",
"Timothy P. L. Smith",
"Gary L. Bennett",
"Theodore S. Kalbfleisch",
"Veronica Basnayake",
"Michael P. Heaton",
"Carol G. Chitko-McKown",
"Timothy P. L. Smith",
"Gary L. Bennett",
"Theodore S. Kalbfleisch",
"Veronica Basnayake"
],
"abstract": "Background: Mannheimia haemolytica is the major bacterial infectious agent of bovine respiratory disease complex and causes severe morbidity and mortality during lung infections. M. haemolytica secretes a protein leukotoxin (Lkt) that binds to the CD18 receptor on leukocytes, initiates lysis, induces inflammation, and causes acute fibrinous bronchopneumonia. Lkt binds the 22-amino acid CD18 signal peptide domain, which remains uncleaved in ruminant species. Our aim was to identify missense variation in the bovine CD18 signal peptide and measure the effects on Lkt binding. Methods: Missense variants in the integrin beta 2 gene (ITGB2) encoding CD18 were identified by whole genome sequencing of 96 cattle from 19 breeds, and targeted Sanger sequencing of 1238 cattle from 46 breeds. The ability of different CD18 signal peptide variants to bind Lkt was evaluated by preincubating the toxin with synthetic peptides and applying the mixture to susceptible bovine cell cultures in cytotoxicity-blocking assays. Results: We identified 14 missense variants encoded on 15 predicted haplotypes, including a rare signal peptide variant with a cysteine at position 5 (C5) instead of arginine (R5). Preincubating Lkt with synthetic signal peptides with C5 blocked cytotoxicity significantly better than those with R5. The most potent synthetic peptide (C5PQLLLLAGLLA) had 30-fold more binding activity compared to that with R5. Conclusions: The results suggest that missense variants in the CD18 signal peptide affect Lkt binding, and animals carrying the C5 allele may be more susceptible to the effects of Lkt. The results also identify a potent class of non-antibiotic Lkt inhibitors that could potentially protect cattle from cytotoxic effects during acute lung infections.",
"keywords": [
"Cattle",
"CD18",
"integrin beta 2",
"missense mutation",
"signal peptide variants",
"bacterial leukotoxin",
"bovine respiratory disease",
"shipping fever",
"Mannheimia haemolytica"
],
"content": "Introduction\n\nMannheimia haemolytica is the major bacteria associated with bovine respiratory disease, a heterogeneous complex of highly infectious pathogens that are the primary cause of morbidity, mortality, and economic loss affecting beef and dairy cattle industries1,2. M. haemolytica is a commensal bacterium found in tonsillar crypts and the upper respiratory tracts of healthy cattle3,4. Exposure to environmental stresses or co-infection with other viral or bacterial pathogens can impair host defenses allowing M. haemolytica to proliferate and colonize the lungs where infection causes acute fibrinonecrotic pleuropneumonia2,5,6. This bacterium expresses a variety of virulence factors that contribute to disease pathogenesis in the lungs. However, leukotoxin (Lkt) is the primary virulence factor contributing to the clinical signs and severe lung damage observed following infection7,8. Within hours of bacterial colonization of the lung, large numbers of polymorphonuclear leukocytes (PMN) infiltrate the airways2,9. Lkt binds to the bovine CD18 subunit of the heterodimeric integrins on the surface of PMN causing cell lysis and the release of pro-inflammatory cytokines, proteolytic enzymes, and reactive oxygen intermediates that intensify local inflammation6,10–13. Experimental depletion of PMN prior to infection14, or infection with a Lkt-deletion mutant of M. haemolytica7,8, results in decreased morbidity and reduced lung lesions in calves. Thus, the interaction between the toxin and its receptor is critical to the pathogenesis of M. haemolytica infection and is a potential intervention point for the prevention of disease.\n\nA 13-amino acid sequence in the CD18 signal peptide has been identified as the site which binds to bacterial Lkt10. The 22-amino acids that comprise the CD18 signal peptide remain uncleaved in leukocytes of ruminant species due to a conserved cleavage-inhibiting glutamine residue at position 18 (Q18) of the propeptide10. In non-ruminant species, such as human and murine, leukocytes are naturally resistant to Lkt because their CD18 signal peptides undergo cleavage due to a glycine residue at position 18 (G18). However, when murine cell lines were transfected with bovine ITGB2, the gene which encodes CD18, they became susceptible to Lkt. When site-directed mutagenesis of ITGB2 was used in the same murine cell lines to change the bovine Q18 residue to G18, the bovine CD18 signal peptide was cleaved and the murine cells once again became resistant to Lkt-induced lysis10. This strategy was taken further with gene-editing, showing that leukocytes isolated from a cloned bovine fetus, homozygous for CD18 G18, were unaffected by Lkt exposure because the signal peptide was cleaved and unavailable for Lkt binding15. Thus, retention of the ruminant CD18 signal peptide appears to be the cause of Lkt sensitivity in leukocytes.\n\nNaturally occurring CD18 amino acid sequence variation can also interfere with Lkt cytotoxicity. In Holstein dairy cattle, a CD18 substitution of glycine for aspartate at polypeptide position 128 in the extracellular I-like domain of CD18 causes bovine leukocyte adhesion deficiency (BLAD) in homozygous animals16,17. These calves do not express functional CD18 on the surface of their leukocytes and have significantly reduced sensitivity to Lkt compared to control calves18,19. We hypothesized that other variation in the CD18 polypeptide sequence, if it exists, may alter the Lkt-CD18 binding interaction or cell signaling and result in differences in lymphocyte sensitivity to M. haemolytica Lkt. Thus, the goals of this study were to identify CD18 protein variants encoded by ITGB2 in U.S. cattle breeds, and evaluate the effects of signal peptide variants on Lkt binding. We report the identification of 15 predicted protein variants, including one with enhanced Lkt binding.\n\n\nMethods\n\nAll animal procedures were reviewed and approved by the U.S. Department of Agriculture, Agricultural Research Service, U.S. Meat Animal Research Center (USMARC) Institutional Animal Care and Use Committee (IACUC project number 2.2).\n\nTwo panels of DNAs were used to determine ITGB2 genotypes from U.S. cattle. The first was a previously described panel of 96 unrelated beef cattle from 19 popular U.S. beef breeds that had already been characterized by whole-genome sequencing20. This identified predicted coding changes throughout the ITGB2 gene. The second panel included a non-overlapping set of 1142 unrelated cattle from 46 breeds, on which targeted Sanger sequencing was performed to identify any predicted coding changes in the signal peptide region and the region containing the D128G variant causing BLAD (ITGB2 exons 2, 3, and 5). Briefly, the first panel of 96 beef cattle (USMARC Beef Cattle Diversity Panel version 2.9 [MBCDPv2.9]) was based on commercially-available purebred registered sires. Pedigrees were obtained from leading suppliers of U.S. beef cattle semen and analyzed to identify unrelated individuals for inclusion. The number of sires representing each breed (four, five, or six) was based on their numbers of registered progeny circa 2000: Angus (n = 6), Hereford (n = 6), Charolais (n = 6), Simmental (n = 6), Red Angus (n = 6), Limousin (n = 6), Gelbvieh (n = 6), Brangus (n = 5), Beefmaster (n = 5), Salers (n = 5), Shorthorn (n = 5), Maine-Anjou (n = 5), Brahman (n = 5), Chianina (n = 4), Texas Longhorn (n = 4), Santa Gertrudis (n = 4), Braunvieh (n = 4), Corriente (n = 4), and Tarentaise (n = 4). On the basis of the number of registered progeny, the breeds were estimated to represent greater than 99% of the germplasm used in the US beef cattle industry, contain more than 187 unshared haploid genomes, and allow a 95% probability of detecting any allele with a frequency greater than 0.01621.\n\nThe second panel of 1142 cattle consisted of samples from male and female registered purebred cattle with diverse pedigrees from 46 breeds. Samples were from semen, blood, or hair follicles, depending on gender and availability22. Where possible, animals within breed were chosen so they did not share parents or grandparents, and none were closely related to the 96 sires in the MBCDPv2.9. The breeds used in the second panel were: Angus (n = 24), Ankole-Watusi (n = 20), Ayrshire (n = 24), Beefmaster (n = 24), Belgian Blue (n = 24), Blonde d'Aquitaine (n = 24), Brahman (n = 23), Brahmousin (n = 24), Brangus (n = 24), Braunvieh (n = 24), Brown Swiss (n = 26), Charolais (n = 24), Chianina (n = 24), Corriente (n = 24), Devon (n = 23), Dexter (n = 22), Gelbvieh (n = 23), Guernsey (n = 23), Hereford (n = 24), Highland (n = 24), Holstein (n = 81), Indu-Brazil (n = 25), Jersey (n = 29), Limousin (n = 24), Maine-Anjou (n = 24), Marchigiana (n = 24), Mini-Hereford (n = 24), Mini-Zebu (n = 24), Montbeliard (n = 24), Murray Grey (n = 20), Nelore (n = 24), Piedmontese (n = 25), Pinzgauer (n = 23), Red Angus (n = 23), Red Poll (n = 24), Romagnola (n = 24), Salers (n = 24), Santa Gertrudis (n = 24), Senepol (n = 23), Shorthorn (n = 23), Simmental (n = 23), Tarentaise (n = 24), Texas Longhorn (n = 23), Texas Longhorn, Cattlemen’s Texas Longhorn Registry (CTLR, n = 19), Tuli (n = 23), and Wagyu (n = 22).\n\nUnless otherwise indicated, reagents were molecular-biology grade. DNA from whole blood samples was extracted by use of a solid-phase system incorporating either spin-columns or 96-well microtitration plates according to the manufacturer's instructions (Qiagen Inc., Germantown, MD, USA). DNA from liver, muscle, skin, or hair samples was extracted by standard procedures22. Briefly, minced tissue (35 mg) or hair follicles (100 trimmed bulbs) were suspended in 2.5 mL of a lysis solution containing 10 mM TrisCl, 400 mM NaCl, 2 mM EDTA, 1% wt/vol sodium dodecyl sulfate, RNase A (250 ug/ml; Sigma-Aldrich, St. Louis, MO, USA), pH 8.0. The solution was incubated at 37°C with gentle agitation. After 1 hour, 1 mg proteinase K was added (Sigma-Aldrich) and the solution was incubated overnight at 37°C with continued agitation. The solution was transferred to 15 ml tube containing 3 ml of a phase-separation gel (high-vacuum grease, Dow Corning Corporation, Midland, MI, USA) and extracted twice with 1 vol of phenol:chloroform:isoamyl alcohol (25:24:1), and once with 1 vol of chloroform before precipitation with 2 vol of 100% ethanol. The precipitated DNA was washed once in 70% ethanol, briefly air dried, and dissolved in a solution of 10 mM TrisCl, 1 mM EDTA (TE, pH 8.0).\n\nDNA from commercial bull semen was extracted similarly, with slight modification23. Briefly, three 0.5 ml straws of commercial semen from a single animal were pooled, and the cells were collected by centrifugation for 5 min at 1000 x g. The cell pellet was washed three times in 1 ml of a wash solution (TE with 100 mM NaCl, TNE) and suspended in 1 ml of the same solution with 1% wt/vol sodium dodecyl sulfate, 1 mg proteinase K (Sigma-Aldrich), and 40 mM dithiothreitol (DTT). This 1 ml lysis solution was incubated overnight at 37°C, transferred to a 15 ml tube containing 1.5 ml of TNE with 40 mM DTT, and 3 ml of a phase-separation gel, extracted twice with 1 vol of phenol:chloroform:isoamyl alcohol (25:24:1), and once with 1 vol of chloroform before precipitation with 0.1 vol of 3 M sodium acetate (pH 5.2) and 2 vol of 100% ethanol. The precipitated DNA was washed once in 70% ethanol, briefly air dried, and dissolved in a solution TE.\n\nPCR-amplified fragments of genomic DNA from ITGB2 exons 2, 3 and 5, encoding the CD18 signal peptide sequence and the region containing the known variant causing BLAD, were produced for Sanger sequencing in 96 or 384-well plates. A standard 25 µl amplification reaction contained 2.5 µl of genomic DNA in TE (10 ng/µl), 12.5 µl of a concentrated PCR cocktail (Maxima Hot Start, Thermo Fisher Scientific, Waltham, MA, USA), 1.25 µl each of an oligonucleotide primer stock solution (100 uM, in TE), 1 µl dimethyl sulfoxide (Sigma-Aldrich), and 6.5 µl water. The sense and antisense primer sequences for exons 2 and 3 were 5'-AGG-GAG-ACT-GAC-CTG-TGT-G-3' and 5'-CTG-GGA-AGC-AGA-GTG-ATA-GT-3', respectively (USMARC primer no. 89878 and 89880). The sense and antisense primer sequences for exon 5 were 5'-AGA-GAG-ATC-CAG-GTA-GAA-CTG-3' and 5'-GTG-CAG-AGG-TGC-AGA-GGT-G-3', respectively (USMARC primer no. 89887 and 89889). The final concentration of each primer was of 5 uM (Integrated DNA Technologies, Inc., Coralville, IA, USA). PCR was performed with either the PTC 200, the PTC 220 Dyad, or the PTC 225 Tetrad thermal cycler chassis (MJ Research, Watertown, MA, USA). Reactions were denatured at 94°C for 15 min, subjected to 45 cycles of denaturation at 94°C for 20 s, annealed at 58°C for 30 s, and extended at 72°C for 1 min. After cycling the final products were extended at 72°C for 3 min before storage at 4°C. A 5 µl portion of each amplified product was analyzed by agarose gel electrophoresis (0.8%) in buffer containing 90 mM Tris-borate (pH 8.0), 2 mM ethylenediamine tetraacetic acid, and 0.1 µg/ml ethidium bromide. A 6 µl portion of each amplified product was treated with Exonuclease I (1.4 U, New England Biolabs Inc., Ipswich, MA, USA) at 37°C for 1 hr in a 13 µl reaction volume to digest single-stranded primer oligonucleotides. The Exonuclease I was inactivated with a 65°C incubation for 20 min and the DNA was precipitated with two volumes of 100% ethanol. The plates were centrifuged at 1800 x g for 30 min, decanted, and air dried.\n\nSequencing reactions were accomplished by dissolving air-dried DNA pellets in 5 µl of sequencing reaction cocktail containing 0.25 µl of dye terminators (DYEnamic ET Dye terminators, Amersham Biosciences, Piscataway, NJ, USA), 1.75 µl dye terminator dilution buffer (Amersham Biosciences), 2 µl oligonucleotide primer (1.6 µM stock solution in water), and 1 µl water according to the manufacturer’s instructions. The final oligonucleotide concentration was 640 nM. Reactions were denatured at 96°C for 30 s, subjected to 26 cycles of denaturation at 96°C for 10 s, annealed at 50°C for 5 s, and extended at 60°C for 4 min. After cycling the final products were stored at 4°C until the DNA was precipitated with 22 µl of 70% isopropyl alcohol. The plates were centrifuged, at 1800 x g for 30 min, decanted, and the samples were washed with 22 µl of 70% ethanol. The plates were centrifuged again, decanted, air dried, sealed with foil, and stored at -20°C until use. Sequencing reactions were resolved by capillary electrophoresis as described by the manufacturer (3730xl DNA Analyzer, Applied Biosystems, Foster City, CA, USA). Animal sequences from both strands were analyzed with polyphred software version 6.1824 in conjunction with the phred/phrap/consed software version 2925–27.\n\nWhole genome sequencing of BL3 cells (a bovine lymphoma cell line; kindly provided by Dr. Subramaniam Srikumaran) was accomplished with methods as described elsewhere20. Briefly, genomic DNA was used to make a 500 bp paired-end library and sequenced with a massively parallel sequencing machine and high-output kits (NextSeq500, two by 150 paired-end reads, Illumina, San Diego, CA, USA) until a minimum of 40 GB of data with greater than Q20 quality, was collected. After sequencing, the raw reads were filtered to remove adaptor sequences, contaminating dimer sequences, and low-quality reads. The DNA sequence alignment process was similar to that previously reported20. FASTQ files were aggregated for each sample and DNA sequences were aligned individually to the bovine reference assembly UMD3.128 with the Burrows-Wheeler aligner (BWA) aln algorithm version 0.7.1229, then merged and collated with bwa sampe. The resulting sequence alignment map (SAM) files were converted to binary alignment map (BAM) files, and subsequently sorted via SAMtools version 0.1.1830. Potential PCR duplicates were marked in the BAM files using the Genome Analysis Toolkit (GATK) version 1.5-32-g2761da931. Regions in the mapped dataset that would benefit from realignment due to small indels were identified with the GATK module RealignerTargetCreator, and realigned using the module IndelRealigner. The BAM files produced at each of these steps were indexed using SAMtools. The resulting indexed BAM files were made available via the USMARC WGS browser and the raw reads for the BL3 cell line were deposited at NCBI BioProject PRJNA325058, BioSample number SAMN05217649. Mapped datasets for each sample were individually genotyped with the GATK UnifiedGenotyper with arguments “--alleles” set to the VCF file (Extended Data File S1)32, “--genotyping_mode” set to “GENOTYPE_GIVEN_ALLELES”, and “--output_mode” set to “EMIT_ALL_SITES”. Lastly, some SNP variants were identified manually by inspecting the sequence with IGV software version 2.1.2833,34 (described in the Methods section entitled ‘Identifying protein variants encoded by ITGB2’). In these cases, read depth, allele count, allele position in the read, and quality score were considered when the manual genotype determination was made.\n\nAligned WGS data from 96 sires of MBCDPv2.9 were visually analyzed in the ITGB2 coding region to identify potential CD18 protein variants. Viewing the aligned sequences and detecting variants was accomplished with the IGV software and a browser developed for this purpose. Briefly, public internet sites at the USDA, ARS, USMARC were used in combination with open source software installed on a laptop computer and recorded manually in a spreadsheet as previously described20. A Java Runtime Environment (Oracle Corporation, Redwood Shores, CA, USA) was first installed on the computer. When links to the data were selected by the user, IGV software33,34 was loaded from a third-party site (University of Louisville, Louisville, KY, USA) and aligned DNA sequence reads were displayed in the context of the bovine UMD3.1 reference genome assembly. For viewing ITGB2 gene variants, WGS from a set of eight animals of different breeds was loaded, and the IGV browser was directed to the appropriate genome region by entering “ITGB2” in the search field. The IGV zoom function was used to view the first exon at nucleotide resolution with the [show translation] option selected in IGV. An example of the alignment view for ITGB2 codon 27 with eight animals is shown in Extended data, Figure S135.\n\nThe exon sequences were visually scanned for polymorphisms predicted to alter amino acid sequences, including missense, nonsense, frameshift, splice site, and insertion/deletion mutations. An in silico analysis of other potential splice-affecting variants was not performed as there are no consensus guidelines on the selection of programs or protocols to interpret the predicted results in cattle. Once identified, the variant nucleotide position was viewed and recorded for all 96 animals. The codons affected by SNP alleles were translated into their corresponding amino acids with IGV, codon tables, and knowledge of the CD18 protein sequence (NP_786975). Haplotype phases of predicted polypeptide variants were unambiguously assigned with homozygous individuals, and those with only one variant amino acid. A maximum parsimony phylogenetic tree was manually constructed from the unambiguously phased protein variants and used to infer phases in the remaining variants with maximum parsimony assumptions.\n\nA single multiplex assay was designed for the 14 ITGB2 missense SNPs with software provided by the manufacturer (Agena Biosciences, San Diego, CA, USA). The oligonucleotide sequences and assay conditions are provided in Table S1. After design and validation with bovine control DNAs for each SNP, the DNA from the 96 bulls in the MBCDPv2.9 diversity panel were tested in a blinded experiment. Assay design and genotyping was performed at GeneSeek (Lincoln, NE, USA) with the MassARRAY platform and iPLEX Gold chemistry according to the manufacturer’s instructions (Agena Biosciences).\n\nM. haemolytica strains for toxin production were isolated from cattle with severe fibrinous pleuropneumonia in feedlot environments and had complete closed whole genome sequence assemblies available at NCBI. M. haemolytica strain 89010807 N serotype A1 (lktA+) has been widely used for Lkt production for in vitro cytotoxicity assays and has the added advantage of being the parent strain of an isogenic leukotoxin deletion mutant (lktA-)36,37. A second strain, M. haemolytica strain USDA-ARS-USMARC-183 serotype A1, was isolated from an animal that was part of a high-mortality respiratory disease outbreak in a Kansas feedlot in 1991 and represents the first strain with a complete closed genome assembly38; however, it had not previously been used in in vitro assays.\n\nIsolates were maintained on Brain Heart Infusion (BHI) agar (Sigma-Aldrich) frozen stocks were kept in BHI broth with 20% glycerol at -80°C. RPMI 1640 medium (without Phenol Red and L-glutamine, Sigma-Aldrich) and semi-defined medium 2 (SDM2) were used for batch culture production of Lkt. SDM2 is an amino acid-limited culture medium supplemented with cysteine, glutamine, ferric iron, and manganese and was previously shown to greatly improve Lkt production in aerobic batch culture39. For Lkt production, a single, 24-hour colony isolate from BHI agar was inoculated into 5 ml BHI broth in a 10 ml culture tube and incubated overnight at 37°C in 5% CO2 without shaking. The following morning, 1 ml of BHI liquid culture was inoculated into 100 ml of fresh culture medium in a 300 ml Delong-style Erlenmeyer flask with baffles (Corning, Inc., Corning, NY, USA) and incubated at 37°C, 250 rpm, in 5% CO2. At intervals, 14 ml samples were removed and centrifuged at 13,100 x g for 10 min at 4°C. The clarified supernatant was decanted, flash frozen in liquid nitrogen and stored at -80°C until use.\n\nLkt and other proteins secreted into the growth media by M. haemolytica were analyzed by SDS-PAGE. Clarified supernatants were precipitated with one volume of acetone on ice for 30 min, followed by centrifugation at 20,800 x g for 5 minutes at room temperature, and air dried 30 min. Sedimented proteins were dissolved in a commercial sample buffer with lithium dodecyl sulfate and dithiothreitol and used per the manufacture instructions (Thermo Fisher Scientific). Samples were heated to 70°C for 10 minutes and loaded on 4–12% precast polyacrylamide Bis-Tris gels (Thermo Fisher Scientific) at run 160 volts for approximately 35 min in 2-[N-morpholino]ethanesulfonic acid (MES) SDS running buffer (Thermo Fisher Scientific). Proteins sorted by SDS PAGE were stained with coomassie-dye reagent (GelCode Blue, Thermo Fisher Scientific) and destained in water. Prestained protein standards (Novex Sharp, Thermo Fisher Scientific) were used to estimate molecular weights of M. haemolytica proteins from clarified supernatants. ImageJ software (version 1.52A) was used to estimate relative proportions of protein bands on coomassie-stained page gel40.\n\nFor protein immunoblots (western blots), proteins from SDS-PAGE gels were electrophoretically transferred to 0.2 µm polyvinylidene difluoride membranes (PVDF, Invitrolon, Thermo Fisher Scientific) with a Mini Blot Module (Thermo Fisher Scientific) per the manufacturer's instructions. PVDF membranes were wetted in 100% methanol prior to equilibrating in transfer buffer (Bolt transfer buffer, Thermo Fisher Scientific) and assembling in the blotting apparatus. Proteins were transferred to membranes for 60 min at a constant voltage of 20 V. Blots were removed from the apparatus and incubated in a blocking reagent (StartingBlock(PBS), Thermo Fisher Scientific), for 60 min at room temperature with gentle agitation. This solution was replaced with a fresh blocking reagent that had a rabbit polyclonal Ltk antibody at a concentration of 1 µg/ml (M. haemolytica Lkt Antibody, catalog number LS-C369014, LifeSpan, BioSciences, Inc, Seattle, WA, USA) and incubated as above for 60 min. The primary Lkt antibody was washed three times in for 10 min each in 25 mM Tris, 0.15 M NaCl, 0.05% Tween-20, pH 7.5 (TBS Tween-20, Thermo Fisher Scientific). After washing, a goat anti-rabbit IgG antibody conjugated to horseradish peroxidase (HRP) (catalog number ab97040, Abcam, Cambridge, MA, USA) was added at a concentration of 1 µg/ml in blocking reagent and incubated for 60 min at room temperature with gentle agitation. This secondary anti-rabbit antibody was washed three times for 10 min each in TBS Tween-20 prior to detection with chemiluminescent substrate (Pierce ECL Western Blotting substrate, Thermo Fisher Scientific). The immunoblot was incubated in chemiluminescent substrate for 1 min and imaged for approximately 5 min (ChemiDoc, Bio-Rad Laboratories, Inc. Hercules, CA, USA).\n\nBL3 cells were propagated in RPMI 1640 medium (Gibco, Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (Atlas Biologicals, Fort Collins, CO, USA),1x antibiotic/antimycotic (Gibco, Thermo Fisher Scientific) and 2 mM L-glutamine (Gibco, Thermo Fisher Scientific). Custom bovine CD18 signal peptides (Figure 4–Figure 7) were commercially synthesized, (Thermo Fisher Scientific) purified by preparative high-performance liquid chromatography, and lyophilized. Peptides were dissolved in dimethysulfoxide at a concentration of 10 mg/ml, aliquoted, and stored at −20°C.\n\nPrimary cells were collected from two mixed breed animals (kept as part of the USMARC cattle population) that were each homozygous for the most common ITGB2 haplotype (variant “1”). For isolation of primary bovine cells, 50 ml of blood were collected by jugular puncture using 16-guage needles into syringes containing EDTA as an anticoagulant. PMN were isolated using a standard hypotonic lysis procedure. Briefly, blood was spun for 25 min at 1000 x g at 4°C. Plasma and buffy coat layers were removed and discarded. Sterile water was added to the red blood cell (RBC) layer to lyse RBC followed by addition of 10X PBS to restore tonicity. PMN were isolated by centrifugation for 10 min at 250 x g at 4°C. The PMN cell pellet was washed three times with 1x PBS and the final cell pellet was resuspended in RPMI 1640 medium.\n\nPeripheral blood mononuclear cells (PBMC) were isolated essentially as described41. Briefly, PBMC were isolated over Ficoll-Paque Plus (GE Healthcare Bio-Sciences AB, Uppsala, Sweden), as per the manufacturer’s instructions, with modification. Briefly, 15 ml of whole blood mixed 1:1 with PBS was underlayed beneath 14 ml of the density gradient in a 50 ml conical tube. The tubes were then centrifuged for 45 min at 900 x g at room temp and with no brake. The PBMC layer was carefully removed and brought up to 45 ml in PBS in a new 50 ml conical tube followed by centrifugation for 15 min at 400 x g at 4°C with high brake. Erythrocytes were removed using RBC lysing buffer (Sigma-Aldrich). The PBMC pellet was further washed three times with 1x PBS and the final pellet was resuspended in RPMI 1640 medium.\n\nThe ability of different CD18 signal peptide variants to bind Lkt and inhibit Lkt-induced cytolysis was measured with the MTT dye-reduction cytotoxicity assay10. The BL3 cell line was selected because it is the most well-studied, readily available, immortalized cell line susceptible to Ltk-induced cytolysis. CD18 signal peptides were tested at concentrations ranging from 50 μM to 0.195 μM. CD18 signal peptides were diluted using serial 2-fold dilutions in 96-well round bottom plates containing 50 µl/well of Lkt at a 50% toxicity end point titer in colorless RPMI 1640 medium without phenol red (Sigma-Aldrich). Synthetic signal peptides and Lkt preparations were pre-incubated for 1 hr on ice prior to the addition of 5 × 105 cells/wells. These cells were added as a 50 ul suspension with a density of 1 × 107 cells/ml in colorless RPMI. Cells were incubated with synthetic signal peptides and Lkt for 1 hr at 37°C in 5% CO2 and subsequently centrifuged at 600 x g for 7 min at 4°C and the supernatant was removed and discarded. Cells were resuspended in 100 µl of colorless RPMI and 20 µl 0.5% MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide; Sigma-Aldrich) and incubated for 20 min at 37°C in 5% CO2, centrifuged as before, and the supernatant was removed and discarded. The purple formazan precipitate was then dissolved in 100 µl of acid isopropanol (0.04N HCl in isopropanol). Following a 5 min incubation, cellular debris was pelleted and the supernatant was transferred to a new 96-well plate. The optical density (OD) of the samples was measured at 570 nm and 690 nm. The background measurement at 690 nm was then subtracted from the 570 nm measurement to give the background adjusted OD. The percent cytotoxicity was calculated as follows: (1-(OD of toxin treated cells/OD of cells without toxin)) x 100. The percent inhibition of cytotoxicity in the presence of synthetic CD18 peptides was calculated as follows: ((percent cytotoxicity in the absence of peptide – percent cytotoxicity in the presence of peptide)/percent cytotoxicity in the absence of peptide) x 100. Given normal variability in the assay, values were occasionally obtained outside the range of 0 to 100% inhibition. Negative values were replaced with 0 and values greater than 100 were replaced with 100 for graphical presentation.\n\nFresh bovine PBMC or PMN were used for comparison to results obtained using immortalized BL3 cells. Cells were suspended in 50 µl colorless RPMI medium and 10 µl Biolog Redox Dye MB, containing 500 μM water-soluble tetrazolium (Biolog, Hayward, CA, USA). This dye was used because it was found to be more sensitive to changes in cellular respiration when assaying primary cells (data not shown). Cells were incubated for 3 hr at 37°C in 5% CO2, centrifuged as before, and the supernatant was transferred to a clean 96 well plate prior to the OD being measured at 590 nm and 750 nm. Calculations for percent cytotoxicity and percent inhibition were calculated using background adjusted OD values as described above.\n\nThe half maximal inhibitory concentration (IC50) was estimated for each signal peptide variant to compare the concentration of peptide needed to block 50% of Lkt-induced cytotoxicity of susceptible bovine cells. Univariate nonlinear regression (NLIN procedure, SAS 9.4, SAS Institute Inc., Cary, NC, USA) was used to fit data from three replicates to the following equation: Yijk = T + (B - T)/(1 + (Cj/IC50i)Si) + e ijk, where Yijk is the measured Lkt inhibition for the kth replicate of the ith peptide at the jth concentration (Cj) of peptide. T is the estimated maximum, B is the estimated minimum of the fitted nonlinear relationship between peptide concentration and Lkt inhibition. Si is the slope of the relationship at IC50i, the estimated concentration that the ith peptide inhibits Lkt-induced cytotoxicity by 50%. e ijk is random variation. Standard errors of IC50i were used for comparisons of estimates.\n\n\nResults\n\nBovine ITGB2 consists of 16 exons spanning 29.1 kb of genomic DNA and encodes a 769 amino acid protein with multiple functional domains (Figure 1A,B). Using software to view the aligned genome sequences from 96 diverse bulls, 13 codons with 14 missense variants were identified (Extended Data Table S2)42. There were no frameshifts, splice sites, or indel polymorphisms observed that would affect the predicted amino acid sequence. Two ITGB2 regions of interest were further selected for additional Sanger sequencing in 1142 purebred cattle from 46 breeds: the signal peptide region (exon 2), and the region containing the D128G variant causing BLAD (exon 5). DNA sequence analyses revealed no additional polymorphisms that were predicted to alter the polypeptide sequence, except the previously described D128G variant in Holstein cattle (data not shown). The genotypes were independently verified with a single, multiplexed, MALDI-TOF MS assay for 14 SNPs in the 96 diverse bulls, and in 1142 of 1168 cattle from 46 breeds (Extended Data Table S3)43. Thus, in total, there were 14 missense variants identified in 13 codons (Table 1).\n\n(A) Genomic DNA map of ITGB2: blue arrows, coding regions of exons; orange arrows, untranslated regions of exons; grey horizontal lines, intron regions. (B) Map of CD18 domains in relationship to missense mutations found in cattle. (C) Rooted maximum parsimony phylogenetic tree of CD18 protein variants from 1142 cattle from 46 breeds with connections to closely-related species. The most frequent CD18 isoform (“variant 1”) was used as the reference amino acid sequence for the trees. For “variants 1” through “14”, each node in the tree represents a different isoform of CD18 that varies by one amino acid compared to adjacent nodes. The hash marks represent nodes that were inferred but not observed. The areas of the circles are proportional to the overall variant frequency in the cattle tested. Yak CD18 was identical to bovine CD18 “variant 3” and denoted as the most likely root of the phylogenetic tree based on the relationship of CD18 sequences in yak, gaur, banteng, and bison. SP, signal peptide; PSI, plexin-semaphorin-integrin domain; SH, spacer-hybrid domain; I-L, I-like domain; MR, mid-region hybrid domain; EGF, epidermal growth factor-like domain; BT, beta tail domain; TM, transmembrane domain; CP, cytoplasmic domain.\n\naAll sequences presented are from the sense strand bovine ITGB2 gene. However, in the UMD3.1 reference assembly, ITGB2 oriented in the antisense direction. bCD18 protein domain abbreviations: SP, signal peptide; PSI, plexin-semaphorin-integrin domain; SH, spacer-hybrid domain; IL, I-like domain; and BT, beta-tail domain. cIUPAC/IUBMB ambiguity codes used for nucleotides: R= a/g, Y = c/t, M= a/c, K = g/t, S = c/g, W = a/t46. dThe major allele is listed first. eMinor allele frequency in the beef cattle diversity panel MBCDPv2.9 (n = 96). fMinor allele frequency in the extended purebred cattle panel with 46 breeds (n = 1142). gAllele not detected in the indicated group of cattle. hMissense mutation associated with BLAD17.\n\nIdentifying haplotypes that encode distinct combinations of missense variants on the CD18 polypeptide is important for evaluating their potential function. A total of 15 ITGB2 haplotypes were identified that, when translated, were predicted to encode different CD18 proteins (Table 2). These 15 predicted polypeptide sequences were placed in the context of a maximum parsimony phylogenetic tree (Figure 1C). Haplotypes encoding CD18 protein variants “1 to 7”, “9”, “13”, and “14” were confirmed by their presence in homozygous animals. Haplotypes encoding CD18 protein variants “8”, “10”, and “15” were unambiguously confirmed in animals with only one heterozygous site. However, haplotype phase was ambiguous when the distance between two heterozygous sites exceeds the length of the DNA sequence read (150 bp in these WGS data sets). Thus, haplotypes for the remaining CD18 protein variants “11” and “12” were tentatively inferred from additional breed-level frequency information. For example, the inferred phase for variant “12” (P155L656) was only observed in two of 27 Braunvieh cattle that were each heterozygous for both missense variants. However, all 25 of the other Braunvieh cattle sequenced were homozygous for the variant “1” (Table 3 and Extended Data Table S3)43. Thus, it was reasonable to infer that P155 (exon 5), and L656 (exon 14) are present on one rare haplotype in two animals, rather than on two rare haplotypes in each animal. Similarly, the inferred phase for variant “11” (C5L656) was only observed in six of 22 Wagyu cattle and each were heterozygous at positions 5 and 656. Since the 22 Wagyu cattle have a variant “1” frequency of 0.8 it seems likely that C5 and L656 variants are present on the same chromosome (i.e. diplotype “1,11”) rather than split across two chromosomes (i.e., diplotype “5,15”). In spite of the potential for ambiguous haplotype phases with rare variants, the phylogenetic tree of predicted CD18 proteins provides a solid framework for further evaluation.\n\naCD18 protein variant allele definitions are shown in Figure 4C. bThe red bold residues are those differing from “variant 1”. cThe coefficient of determination for these frequencies (r2) was 99.7. dAllele not detected in the indicated group of cattle.\n\naCD18 protein variant allele definitions are shown in Table 2. bThe CD18 C5 missense variant appears in protein variants 9, 11, 13, and 15. cThe CD18 G128 missense variant causes BLAD in Holstein cattle. dAllele not detected in the indicated group of cattle. eTexas Longhorn, Cattlemen’s Texas Longhorn Registry.\n\nDetermining the most likely phylogenetic root of the CD18 tree is important for establishing the likely order of mutational events. Comparing cattle CD18 precursor protein variants to those from closely related species in the Bos genus indicated that variant “3” (K27) was the most likely root of the phylogenetic tree (Figure 1C). Cattle are predicted to share a common ancestor with other species in the Bos genus approximately 5 million years ago. In addition, the K27 variant was associated with indicine cattle breeds, while the N27 variant was associated with taurine breeds. For example, the N27 frequency in 840 cattle from 33 taurine breeds was 0.998, while the K27 frequency in 70 Brahman, Indu Brazil, and Nelore cattle was 0.95. Thus, the structure of the rooted tree suggests that CD18 protein variants “1” and “3” are the ancestral polypeptide sequence of taurine and indicine breeds, respectively. The rooted tree also suggests that the distal nodes represent CD18 variants that have arisen sometime after the split between taurines and indicines, approximately 500,000 years ago.\n\nThe conservation of amino acid residues throughout vertebrate species is a measure of their potential impact on protein function. Highly conserved residues are more likely to be indispensable for function and thus, variation at these positions is assumed to be deleterious. The 769 amino acid sequence of cattle CD18 is highly similar to those from yak, sheep, whale, and humans (99, 95, 90, and 83% identity, respectively). The cattle CD18 sequence is also remarkably similar to chicken, frog, fish, lamprey, and fruit fly (63, 57, 49, 42, and 19% identity, respectively), and has some polypeptide regions that are invariant throughout the Bilateria (Extended Data Table S4)44. The 13 variant amino acid positions in the bovine CD18 precursor protein were compared to those in 35 representative Bilateria species. The aspartate residue at position 128 (D128) is invariant throughout the Bilateria, except in cattle wherein G128 causes the complete loss of CD18 function and results in BLAD in homozygous individuals of the Holstein breed (Figure 2). CD18 positions 65, 104, and 155 are also conserved and the variant allele in cattle is rare. Based only on the degree of residue conservation across species, the proposed predicted order of negative impact on CD18 function in cattle was: G128 > G65 = M104 > P155. The CD18 K27 indicine variant was conserved throughout the jawed vertebrates and thus, the major N27 taurine variant was not observed in any other species. However, based on its high frequency in most taurines, the N27 variant does not appear to be deleterious.\n\nAligned and gapped protein sequences from a representative set of 56 bilateria species were compared (Extended Data Table S4). At variant sites in cattle, the residues were summarized for a representative subset of 35 species. TMRCA, estimated time to most recent common ancestor in millions of years47; letters, IUPAC/IUBMB codes for amino acids; dot, amino acid residues identical to those in cattle “variant 1”; dash, not enough sequence similarity for comparison or missing residue in that peptide region.\n\nThe conservation of arginine at positions 3 (R3) and 5 (R5) in CD18 signal peptides was of particular interest because this region binds bacterial Lkt in ruminant species, which includes the Bovids, Cervids, Giraffids, musk deer, chevrotains, and pronghorns. The R3 residue was conserved throughout the Bovinae, but not in sheep and goat, which have proline at that position (Figure 2). The histidine residue at position 3 (H3) was rare in cattle, not observed in other ruminants, and on a distal node of the phylogenetic tree (CD18 polypeptide variant “7“, Figure 1C). Two animals from the Brahman breed group were identified with the H3 and one was homozygous, indicating that H3 is not a lethal recessive variant (H3,H3, Extended Data Table S245, and R3,H3, Extended Data Table S343). Unlike R3, the R5 variant was not conserved in Bovinae, since eland have cysteine at this site (C5). In addition, the C5 variant was present on four distinct putative CD18 polypeptide variants: “9“, “11”, “13”, and “15” (Figure 1C), including both taurine- and indicine-influenced breeds. The presence of the C5 variant on multiple but infrequent haplotypes indicates recombination has occurred between this and other CD18 missense variants.\n\nThe discovery of missense variants in the CD18 signal peptide provided the opportunity to test these variants using in vitro cell assays with bacterial Lkt. However, producing sufficient and consistent batches of biologically-active Lkt from reference bacterial strains was a challenge with traditional cell culture medium (RPMI). In an effort to overcome this barrier, Lkt production in RPMI was compared with that in a semi-defined bacterial culture medium (SDM2, Methods) with two wild-type M. haemolytica strains. The total biological activity of Lkt excreted into SDM2 culture was up to 80-fold greater at its peak than that in RPMI for a given reference strain of M. haemolytica (e.g. Strain 183, Figure 3A). In both media, Lkt activity was induced in late log phase as batch cultures made the transition to stationary phase. Although the biological activity quickly diminished as the culture progressed to stationary phase, the total Lkt protein measured by SDS PAGE continued to increase and the level was stable for hours in the culture supernatant (Figure 3B). The Lkt of clarified SDM2 culture supernatants with the highest cytolytic activity (Strain 183, fraction C, Figure 3B) was estimated to be 90% pure based on gel densitometry imaging of coomassie-stained SDS-PAGE gels. Preparations of similar quality were used for in vitro assays. The cytolytic activity of these toxin preparations was stable at -80°C for more than 2 years.\n\n(A) Growth curves of M. haemolytica strains (black lines) and the cytolytic activity of their secreted Lkt (red lines). The strain abbreviations: 183 (ltkA+), USDA-ARS-USMARC-183 (ltkA+); N (ltkA+), 89010807N (lktA+); and N (ltkA-), 89010807N (lktA-). Clarified culture supernatants were collected at time points indicated on the x-axis as A through E. (B) Coomassie-stained SDS-PAGE of proteins from 50 µl each of clarified culture supernatants. (C) Western blot of SDS-PAGE from the same clarified SDM2 culture supernatants that were used in panel (B). The black arrow indicates the expected position of Lkt on SDS-PAGE.\n\nCD18 signal peptide variants were tested for their ability to bind Lkt and inhibit Lkt-induced cytolysis of immortalized bovine BL3 cells using a MTT dye reduction assay. CD18 peptide variants representing amino acids 1 to 22 (A) or 5 to 17 (B) were tested using 2-fold serial dilutions at concentrations ranging from 50 μM to 0.195 μM. Half-maximal inhibitory concentration (IC50) for each peptide was determined using non-linear regression analyses. Data are expressed as the mean with standard deviation (n=3).\n\nThe 13-mer synthetic signal peptides (from Figure 4B) containing cattle CD18 residues 5 to 17 were tested for their ability to inhibit Lkt-induced cytolysis of primary bovine peripheral blood mononuclear cells (PBMC) (A) and polymorphonuclear leukocytes (PMN) (B). Donor animals were genotyped at the CD18 locus and were homozygous for CD18 variant “1”. Peptides were tested using 2-fold dilutions at concentrations ranging from 50 μM to 0.098 μM. The half-maximal inhibitory concentration (IC50) for each peptide was determined using non-linear regression analyses. Data are expressed as the mean with standard deviation (n = 3).\n\nAnalysis of cattle ITGB2 haplotypes showed that three distinct polypeptide sequences are encoded in the 22-amino acid signal peptide region: the common variant with arginine at positions 3 and 5 (R3R5), a rare variant with histidine at position three (H3R5), and second rare variant with cysteine at position 5 (R3C5). Synthetic 22-mer peptides representing these three, full-length, signal peptides were tested for their ability to bind Lkt. The synthetic peptides were pre-incubated with M. haemolytica Lkt preparations to allow binding, and applied to Lkt-sensitive BL3 cell cultures in vitro (Figure 4A). The common R3R5 signal peptide was used as a reference since it has a frequency of approximately 0.98 in U.S. cattle, and is predicted to be present on 10 of the full-length CD18 sequences. The synthetic R3R5 signal peptide had a IC50 of 17.9 μM, which represents the concentration of peptide needed to block 50% of Lkt-induced cytolysis of BL3 cells. The rare H3R5 signal peptide variant, which is only found on one full-length CD18 variant (Figure 1C, variant 7), was similar to the reference, with an IC50 of 21.6 μM. In contrast, the rare R3C5 signal peptide found on full-length CD18 variants 9, 11, 13, and 15 had an IC50 of 5.9 μM, which was 3-fold lower that the reference, indicating an increased affinity for Lkt (Figure 4A).\n\nThe optimum blocking of Ltk-induced cytotoxicity has been previously reported to occur with the 13-mer peptide corresponding to CD18 signal peptide residues 5 to 1710. Thus, we tested the effect of cysteine at position 5 (C5) in this shorter peptide. The synthetic 13-mer C5 peptide was 3.6-fold more effective at blocking Lkt toxicity compared to the 22-mer R3C5 peptide (IC50 of 1.6 and 5.9 μM, respectively). When compared to the 13-mer reference peptide with arginine at position five (R5), the 13-mer C5 peptide was 8-fold better at blocking Lkt-induced cytotoxicity (IC50 13.0 and 1.6 μM, respectively; Figure 4B). A negative control 13-mer peptide containing randomly assorted amino acids from the reference peptide sequence failed to inhibit Lkt-induced cytolysis, even at the highest concentration tested (50 μM), indicating that inhibition of cytolysis was sequence specific (Extended Data File S2)48. Together, these results suggest that peptide sequence and length affect Lkt binding.\n\nThe 13-mer C5 synthetic signal peptide containing CD18 residues 5 to 17 were also tested for their ability to inhibit Lkt-induced cytolysis of primary cells isolated from cattle. The purpose was to demonstrate that the reduction in cytotoxicity observed with C5 signal peptides was comparable between the immortalized cell line and freshly isolated leukocytes from cattle. Like the BL3 cell line, the animals used as donor were homozygous for CD18 variant “1”, and thus have the reference (R3R5) signal peptide. With primary PBMC (Figure 5A) or PMN (Figure 5B), the 13-mer C5 synthetic signal peptide was significantly better at blocking Lkt-induced cytotoxicity compared to the R5 signal peptide (7- and 14-fold respectively) and was similar to that for BL3 cells (8-fold).\n\nSome bovid species have CD18 signal peptide sequences that are slightly different from those in cattle (Figure 6A and Table S3). In water buffalo, the 13-mer peptide sequence corresponding to CD18 signal peptide amino acids 5 to 17 differs at position 16 (S16), while the same region in sheep differs at positions 10 and 12 (F10S12). However, synthetic peptides corresponding to the water buffalo and sheep sequences were similar to the R5 reference cattle signal peptide when tested for their ability to bind Lkt and inhibit cytolysis of BL3 cells in vitro. In contrast, the eland signal peptide was variant at three positions in this region compared to cattle (C5V8G16), and its 13-mer synthetic signal peptide had a 24-fold decrease in IC50 compared to the reference cattle R5 signal peptide and a similar IC50 compared to the reference cattle C5 signal peptide (Figure 6A). However, if the synthetic peptide is expanded to include the eland residue at position 4, the results were different. Eland have an arginine residue at position 4 (R4) where other ruminants have a glutamine (Q4). A synthetic 14-mer signal peptide with R4 (i.e., Eland R4C5V8G16) caused a 66-fold reduction in the ability of the eland signal peptide to bind Lkt compared to the 13-mer (C5V8G16) as measured by IC50 (Figure 6B). When R4 was replaced with the Q4 normally found in cattle and other ruminants (i.e. Eland Q4C5V8G16 versus R4C5V8G16), Lkt binding was restored. Similarly, when Q4 in the cattle signal peptide (Q4C5) was replaced with R4 (R4C5) there was a significant reduction in the ability of this peptide to bind Lkt (5.6-fold increase in IC50). These results suggest that the amino acid position 4 in the signal peptide can affect Lkt binding and that R4 amino acid sequence in eland may disrupt the enhanced Lkt binding conferred by the C5 variant residue.\n\nToxin inhibitors represent a potentially potent class of therapeutics that could protect animals during acute lung infection. Since truncated synthetic C5 signal peptides showed increase affinity for Lkt, various peptide lengths were tested to identify those with maximum Lkt binding. Removing the first four N-terminal amino acids (MLRQ) resulted in no significant difference in Lkt binding for the C5 or reference R5 signal peptides (Extended Data Figure S2)42. In contrast, stepwise deletions of C-terminal residues of the C5 signal peptide had a significant impact on Lkt binding with the highest affinity being a 12-mer C5 signal peptide with amino acids 5 to 16. The IC50 of this 12-mer C5 peptide (CPQLLLLAGLLA) was 23-fold lower than the 13-mer reference R5 signal peptide (residues 5 to 17) and 30-fold lower than the 12-mer R5 signal peptide. Thus, the 12-mer C5 peptide (CPQLLLLAGLLA) represents the minimal naturally-occurring peptide sequence with maximal inhibition of Lkt-induced BL3 cell lysis (Figure 7).\n\n\nDiscussion\n\nThe present report describes bovine CD18 amino acid sequence differences encoded by ITGB2 in 46 breeds of beef and dairy cattle. All of the protein coding variants discovered were missense mutations and their haplotypes were predicted to encode 15 distinct polypeptide sequences. A C5 variant in the CD18 signal peptide region was shown to cause increased binding to M. haemolytica Lkt, a secreted toxin that causes cell lysis and acute inflammation leading to lung injury characteristic of bovine respiratory disease. The C5 signal peptide variant increased the affinity for Lkt, and this effect was influenced by variation at adjacent residues. The increased Lkt binding and protection from cytotoxic effects were observed in the immortalized BL3 cell line, and freshly isolated PBMC and PMN from beef cattle.\n\nThe identification of naturally-occurring CD18 variants with increased binding to Lkt has important potential implications for animal health. For example, cattle with the CD18 C5 signal peptide variant may be at increased risk for toxin-related respiratory disease. However, despite the fact that the C5 variant was found in four predicted protein variants in taurine and indicine cattle, its overall frequency in the U.S. cattle population in still very low (0.01). Thus, identifying available homozygous cattle for testing their leukocytes ex vivo for altered Lkt sensitivity will be challenging. Determining whether this altered binding phenotype contributes to differences in lung lesion severity or disease outcome following M. haemolytica infection will also be difficult. Together, these factors suggest that identifying and removing animals with the CD18 C5 signal peptide would be premature and unwarranted for most cattle operations at this time.\n\nRecent examples of gene editing in animals have shown that this can be a successful strategy for creating novel host genetic resistance. Groundbreaking work with porcine reproductive and respiratory syndrome virus (PRRSV) has shown that gene editing of a critical entry factor (CD163) confers complete resistance to infection in pigs49,50. Similarly, genetic resistance to M. haemolytica Lkt has been demonstrated in leukocytes isolated from a homozygous, gene-edited, bovine fetus expressing a cleavable CD18 signal peptide15. However, the uncleaved CD18 signal peptide is universally conserved in ruminants and thus, its removal may have deleterious effects on the animal. CD18 forms heterodimers with distinct, but structurally homologous alpha integrin subunits (e.g., CD11a/CD18, CD11b/CD18, and CD11c/CD18), and thus the effect of a cleaved signal peptide may have unknown, but far-reaching effects on biological functions. To date there are no reports of a healthy, live calf expressing a cleavable CD18 signal peptide. Thus, modifying the amino acid sequence of an uncleaved ruminant CD18 signal peptide may be useful as an alternative strategy to reduce Lkt binding, while preserving its normal evolutionarily conserved cellular function.\n\n(A) Synthetic 13-mer signal peptides representing amino acids 5 to 17 of sheep, water buffalo, and eland were compared to cattle variants R5 and C5 for their ability to inhibit Lkt-induced cytolysis of bovine BL3 cells. Peptides were tested using 2-fold dilutions at concentrations ranging from 50 μM to 0.195 μM. (B) Synthetic 14-mer signal peptides representing CD18 amino acids 4 to 17 from eland and cattle variant C5 were tested for their ability to bind Lkt. In addition, eland CD18 signal peptides were synthesized where the amino acid at position 4 in eland (arginine, R) was replaced with the amino acid normally found in cattle at this position (glutamine, Q; Eland Q4C5V8G16). Similarly, cattle variant C5 peptides were synthesized where the amino acid at position 4 in cattle was replaced with the amino acid naturally encoded in eland (Cattle R4C5). The half-maximal inhibitory concentration (IC50) for each peptide was determined using non-linear regression analyses. Data are expressed as the mean with standard deviation (n=3 or 4).\n\nSynthetic CD18 signal peptides were synthesized with single amino acid C-terminal deletions. These peptides were tested for their ability to inhibit Lkt-induced cytolysis of bovine BL3 cells. Peptides were tested at concentrations ranging from 50 μM to 0.195 μM. The half-maximal inhibitory concentration (IC50) for each peptide was determined using non-linear regression analyses. Data are expressed as the mean with standard deviation (n=3).\n\nIdentifying the naturally-ocurring CD18 signal peptide residues that influence Lkt binding provides a guide for more extensive analyses that may inform gene edits. Previously, synthetic signal peptides containing the R5 residue were used to identify a 13-amino acid minimum CD18 binding site for Lkt (spanning residues 5 through 17,10). Using synthetic peptides representing cattle Q4C5 and eland R4C5, we showed that residues in position 4 can drastically affect Lkt binding and that the net charge of the signal peptide is important. The positively charged side chain of R4 apparently disrupts the enhanced binding of C5 signal peptides compared to the neutral side chain of Q4. In addition, the most common CD18 22-mer signal peptide (R3R5) and the rare H3R5 signal peptide have a net charge of +2 and reduced Lkt binding compared the R3C5 peptide (net charge of +1). It would be of interest to determine whether increasing the signal peptide net charge to +3 by introducing another positively charged residue could further reduce Lkt binding. However, there are numerous combinatorial peptide possibilities that could lead to a significantly reduced affinity of CD18 for Lkt, while only a few may preserve the basic biological functions of the signal peptide. Although large-scale screening of candidate peptides in vitro and testing them in vivo is beyond the scope of this study, our results suggest this as a possible future avenue of research.\n\nAn alternative strategy for reducing the impact of Lkt would be to block its activity in vivo with synthetic CD18 “decoy” peptides such as the 12-mer C5 signal peptide identified here. A similar strategy has been used to neutralize anthrax toxin from Bacillus anthracis (reviewed in 51). A synthetic 12-mer peptide antitoxin, attached to liposome scaffolds in multiple copies, protected host cells from cytotoxicity in vitro and protected rats from becoming moribund in vivo52. Although the feasibility of delivering decoy peptides is unknown in cattle, it could be used to neutralize Lkt, and thus protect calves from the major virulence factor associated with lung pathology in bovine respiratory disease complex. One can further imagine combining decoy peptide technology with gene editing to have alveolar leukocytes secrete decoy peptide inhibitors of Lkt at the sites of M. haemolytica infection. Although the theoretical possibilities of using antitoxin and gene editing strategies may be vast, the feasibility of these technologies are still relatively unknown. A better understanding of the underlying molecular mechanisms involved, together with significant improvements in livestock gene editing and decoy peptide technologies, are needed to move the field forward.\n\n\nConclusion\n\nThere are more than a dozen missense variants in the CD18 polypeptide, including a C5 variant in the signal peptide that affects Lkt binding. Results in vitro suggest that animals carrying the C5 allele may be more susceptible to the effects of Lkt. These results also identify a potentially potent class of non-antibiotic Lkt inhibitors that could protect cattle from cytotoxic effects during acute lung infections.\n\n\nData availability\n\nWhole genome sequence files (FASTQ) for the BL3 cell line are available in the NCBI SRA under accession number SRX4645762.\n\nThe BL3 sequence data have also been deposited with links to BioProject accession number PRJNA325058 (BioSample SAMN05217649) in the NCBI BioProject database.\n\nIn addition, access to the aligned sequences is available via the USDA internet site: https://www.ars.usda.gov/plains-area/clay-center-ne/marc/wgs/celllines/ as described in the Methods.\n\nTable S1. MALDI-TOF MS assay design for 14 ITGB2 missense SNPs. DOI: https://doi.org/10.6084/m9.figshare.7449374.v153.\n\nTable S2. ITGB2 genotypes recorded manually from WGS reads mapped to UMDv3.1 assembly for the USMARC Beef Cattle Diversity Panel v2.9. DOI: https://doi.org/10.6084/m9.figshare.7449989.v145.\n\nTable S3. Haplotype-phased genotypes (diplotypes) for ITGB2 from MALDI-TOF MS assays for 1142 cattle. DOI: https://doi.org/10.6084/m9.figshare.7450673.v143.\n\nTable S4. Alignment of CD18 sequences from Bilateria species. DOI: https://doi.org/10.6084/m9.figshare.7450796.v144.\n\nFigure S1. Screen image of Integrated Genome Viewer software displaying ITGB2 N27KI genotype data for eight bulls. DOI: https://doi.org/10.6084/m9.figshare.7450814.v135.\n\nFigure S2. Effect of N-terminal truncations of synthetic CD18 signal peptides on Lkt binding. Synthetic CD18 signal peptides were synthesized with four amino acids removed from N-terminus (MLRQ). The common R3R5 (A) or the rare R3C5 variant (B) signal peptides were tested for their ability to inhibit leukotoxin-induced cytolysis of bovine BL3 cells using a MTT cytotoxicity assay. Peptides were tested at concentrations ranging from 50 μM to 0.195 μM. Data are expressed as the mean with standard deviation (n=3). DOI: https://doi.org/10.6084/m9.figshare.7450823.v142.\n\nFile S1. VCF file with SNPs from BL3 WGS aligned to the bovine UMD3.1 reference assembly. https://doi.org/10.6084/m9.figshare.7450826.v132.\n\nFile S2. MTT dye-reduction cytotoxicity assay results. Percent inhibition of cytotoxicity was calculated as described in the Methods. Shown are the results used for statistical analyses of data and for generating Figure 1–Figure 3, Figure 7. DOI: https://doi.org/10.6084/m9.figshare.745103948.",
"appendix": "Grant information\n\nFunding for this research was provided by the USDA, ARS appropriated projects 5438-32000-029-00D (CGCM) and 5438-31320-012-00D (TPLS).\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 thank Brad Sharp, Stacy Bierman, and Jacky Carnahan for technical support and Stephanie Schmidt for secretarial support. The use of product and company names is necessary to accurately report the methods and results; however, the United States Department of Agriculture (USDA) neither guarantees nor warrants the standard of the products. The use of names by the USDA implies no approval of the product to the exclusion of others that may also be suitable. The USDA is an equal opportunity provider and employer.\n\n\nReferences\n\nGriffin D: Bovine pasteurellosis and other bacterial infections of the respiratory tract. Vet Clin North Am Food Anim Pract. 2010; 26(1): 57–71, table of contents. PubMed Abstract | Publisher Full Text\n\nWhiteley LO, Maheswaran SK, Weiss DJ, et al.: Pasteurella haemolytica A1 and bovine respiratory disease: pathogenesis. J Vet Intern Med. 1992; 6(1): 11–22. PubMed Abstract | Publisher Full Text\n\nFrank GH, Smith PC: Prevalence of Pasteurella haemolytica in transported calves. Am J Vet Res. 1983; 44(6): 981–5. PubMed Abstract\n\nAngen O, Thomsen J, Larsen LE, et al.: Respiratory disease in calves: microbiological investigations on trans-tracheally aspirated bronchoalveolar fluid and acute phase protein response. Vet Microbiol. 2009; 137(1–2): 165–71. PubMed Abstract | Publisher Full Text\n\nGriffin D, Chengappa MM, Kuszak J, et al.: Bacterial pathogens of the bovine respiratory disease complex. Vet Clin North Am Food Anim Pract. 2010; 26(2): 381–94. PubMed Abstract | Publisher Full Text\n\nSingh K, Ritchey JW, Confer AW: Mannheimia haemolytica: bacterial-host interactions in bovine pneumonia. Vet Pathol. 2011; 48(2): 338–48. PubMed Abstract | Publisher Full Text\n\nHighlander SK, Fedorova ND, Dusek DM, et al.: Inactivation of Pasteurella (Mannheimia) haemolytica leukotoxin causes partial attenuation of virulence in a calf challenge model. Infect Immun. 2000; 68(7): 3916–22. PubMed Abstract | Publisher Full Text | Free Full Text\n\nTatum FM, Briggs RE, Sreevatsan SS, et al.: Construction of an isogenic leukotoxin deletion mutant of Pasteurella haemolytica serotype 1: characterization and virulence. Microb Pathog. 1998; 24(1): 37–46. PubMed Abstract | Publisher Full Text\n\nAckermann MR, Brogden KA: Response of the ruminant respiratory tract to Mannheimia (Pasteurella) haemolytica. Microbes Infect. 2000; 2(9): 1079–88. PubMed Abstract | Publisher Full Text\n\nShanthalingam S, Srikumaran S: Intact signal peptide of CD18, the beta-subunit of beta2-integrins, renders ruminants susceptible to Mannheimia haemolytica leukotoxin. Proc Natl Acad Sci U S A. 2009; 106(36): 15448–53. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDassanayake RP, Maheswaran SK, Srikumaran S: Monomeric expression of bovine beta2-integrin subunits reveals their role in Mannheimia haemolytica leukotoxin-induced biological effects. Infect Immun. 2007; 75(10): 5004–10. PubMed Abstract | Publisher Full Text | Free Full Text\n\nDeshpande MS, Ambagala TC, Ambagala AP, et al.: Bovine CD18 is necessary and sufficient to mediate Mannheimia (Pasteurella) haemolytica leukotoxin-induced cytolysis. Infect Immun. 2002; 70(9): 5058–64. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJeyaseelan S, Sreevatsan S, Maheswaran SK: Role of Mannheimia haemolytica leukotoxin in the pathogenesis of bovine pneumonic pasteurellosis. Anim Health Res Rev. 2002; 3(2): 69–82. PubMed Abstract | Publisher Full Text\n\nSlocombe RF, Malark J, Ingersoll R, et al.: Importance of neutrophils in the pathogenesis of acute pneumonic pasteurellosis in calves. Am J Vet Res. 1985; 46(11): 2253–8. PubMed Abstract\n\nShanthalingam S, Tibary A, Beever JE, et al.: Precise gene editing paves the way for derivation of Mannheimia haemolytica leukotoxin-resistant cattle. Proc Natl Acad Sci U S A. 2016; 113(46): 13186–13190. PubMed Abstract | Publisher Full Text | Free Full Text\n\nShuster DE, Bosworth BT, Kehrli ME Jr: Sequence of the bovine CD18-encoding cDNA: comparison with the human and murine glycoproteins. Gene. 1992; 114(2): 267–71. PubMed Abstract | Publisher Full Text\n\nShuster DE, Kehrli ME Jr, Ackermann MR, et al.: Identification and prevalence of a genetic defect that causes leukocyte adhesion deficiency in Holstein cattle. Proc Natl Acad Sci U S A. 1992; 89(19): 9225–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nJeyaseelan S, Hsuan SL, Kannan MS, et al.: Lymphocyte function-associated antigen 1 is a receptor for Pasteurella haemolytica leukotoxin in bovine leukocytes. Infect Immun. 2000; 68(1): 72–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAckermann MR, Kehrli ME Jr, Hawkins HK, et al.: Identification of beta2 integrins in bovine neutrophils by scanning electron microscopy in the backscatter mode and transmission electron microscopy. Vet Pathol. 1993; 30(3): 296–8. PubMed Abstract | Publisher Full Text\n\nHeaton MP, Smith TP, Carnahan JK, et al.: Using diverse U.S. beef cattle genomes to identify missense mutations in EPAS1, a gene associated with pulmonary hypertension [version 2; referees: 2 approved]. F1000Res. 2016; 5: 2003. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHeaton MP, Chitko-McKnown CG, Grosse WM, et al.: Interleukin-8 haplotype structure from nucleotide sequence variation in commercial populations of U.S. beef cattle. Mamm Genome. 2001; 12(3): 219–26. PubMed Abstract | Publisher Full Text\n\nHeaton MP, Keele JW, Harhay GP, et al.: Prevalence of the prion protein gene E211K variant in U.S. cattle. BMC Vet Res. 2008; 4: 25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHeaton MP, Grosse WM, Kappes SM, et al.: Estimation of DNA sequence diversity in bovine cytokine genes. Mamm Genome. 2001; 12(1): 32–7. PubMed Abstract | Publisher Full Text\n\nNickerson DA, Tobe VO, Taylor SL: PolyPhred: automating the detection and genotyping of single nucleotide substitutions using fluorescence-based resequencing. Nucleic Acids Res. 1997; 25(14): 2745–51. PubMed Abstract | Publisher Full Text | Free Full Text\n\nEwing B, Green P: Base-calling of automated sequencer traces using phred. II. Error probabilities. Genome Res. 1998; 8(3): 186–94. PubMed Abstract | Publisher Full Text\n\nEwing B, Hillier L, Wendl MC, et al.: Base-calling of automated sequencer traces using phred. I. Accuracy assessment. Genome Res. 1998; 8(3): 175–85. PubMed Abstract | Publisher Full Text\n\nGordon D, Abajian C, Green P: Consed: a graphical tool for sequence finishing. Genome Res. 1998; 8(3): 195–202. PubMed Abstract | Publisher Full Text\n\nZimin AV, Delcher AL, Florea L, et al.: A whole-genome assembly of the domestic cow, Bos taurus. Genome Biol. 2009; 10(4): R42. 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\nLi H, Handsaker B, Wysoker A, et al.: The Sequence Alignment/Map format and SAMtools. Bioinformatics. 2009; 25(16): 2078–9. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMcKenna A, Hanna M, Banks E, et al.: The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 2010; 20(9): 1297–303. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorkman A, Heaton M: File S1. VCF file with SNPs from BL3 WGS aligned to the bovine UMD3.1 reference assembly. figshare. Fileset. 2018.\n\nRobinson JT, Thorvaldsdottir H, Winckler W, et al.: Integrative genomics viewer. Nat Biotechnol. 2011; 29(1): 24–6. PubMed Abstract | Publisher Full Text | Free Full Text\n\nThorvaldsdottir H, Robinson JT, Mesirov JP: Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Brief Bioinform. 2013; 14(2): 178–92. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorkman A, Heaton M: Figure S1. Screen image of Integrated Genome Viewer (IGV) software displaying ITGB2 N27KI genotype data for eight bulls. figshare. Figure. 2018.\n\nHeaton MP, Harhay GP, Smith TP, et al.: Complete Closed Genome Sequences of a Mannheimia haemolytica Serotype A1 Leukotoxin Deletion Mutant and Its Wild-Type Parent Strain. Genome Announc. 2015; 3(3): pii: e00417-15. PubMed Abstract | Publisher Full Text | Free Full Text\n\nMurphy GL, Whitworth LC, Clinkenbeard KD, et al.: Hemolytic activity of the Pasteurella haemolytica leukotoxin. Infect Immun. 1995; 63(8): 3209–12. PubMed Abstract | Free Full Text\n\nHarhay GP, Koren S, Phillippy AM, et al.: Complete Closed Genome Sequences of Mannheimia haemolytica Serotypes A1 and A6, Isolated from Cattle. Genome Announc. 2013; 1(3): pii: e00188-13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nvan Rensburg E, du Preez JC: Effect of pH, temperature and nutrient limitations on growth and leukotoxin production by Mannheimia haemolytica in batch and continuous culture. J Appl Microbiol. 2007; 102(5): 1273–82. PubMed Abstract | Publisher Full Text\n\nSchneider CA, Rasband WS, Eliceiri KW: NIH Image to ImageJ: 25 years of image analysis. Nat Methods. 2012; 9(7): 671–5. PubMed Abstract | Publisher Full Text | Free Full Text\n\nChitko-McKown CG, Fox JM, Miller LC, et al.: Gene expression profiling of bovine macrophages in response to Escherichia coli O157:H7 lipopolysaccharide. Dev Comp Immunol. 2004; 28(6): 635–45. PubMed Abstract | Publisher Full Text\n\nWorkman A, Heaton M: Figure S2. Effect of N-terminal truncations of synthetic CD18 signal peptides on Lkt binding. figshare. Figure. 2018.\n\nWorkman A, Heaton M: Table S3. Haplotype-phased genotypes (diplotypes) for ITGB2 from MALDI-TOF MS assays for 1142 cattle. figshare. Dataset. 2018.\n\nWorkman A, Heaton M: Table S4. Alignment of CD18 sequences from Bilateria species. figshare. Dataset. 2018.\n\nWorkman A, Heaton M: Table S2. ITGB2 genotypes recorded manually from WGS reads mapped to UMDv3.1 assembly for the USMARC Beef Cattle Diversity Panel v2.9. figshare. Dataset. 2018.\n\nNomenclature for incompletely specified bases in nucleic acid sequences. Recommendations 1984. Nomenclature Committee of the International Union of Biochemistry (NC-IUB). Proc Natl Acad Sci U S A. 1986; 83(1): 4–8. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHedges SB, Marin J, Suleski M, et al.: Tree of life reveals clock-like speciation and diversification. Mol Biol Evol. 2015; 32(4): 835–45. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorkman A, Heaton M: File S1. VCF file with SNPs from BL3 WGS aligned to the bovine UMD3.1 reference assembly. figshare. Fileset. 2018.\n\nBurkard C, Lillico SG, Reid E, et al.: Precision engineering for PRRSV resistance in pigs: Macrophages from genome edited pigs lacking CD163 SRCR5 domain are fully resistant to both PRRSV genotypes while maintaining biological function. PLoS Pathog. 2017; 13(2): 1006206. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWhitworth KM, Rowland RR, Ewen CL, et al.: Gene-edited pigs are protected from porcine reproductive and respiratory syndrome virus. Nat Biotechnol. 2016; 34(1): 20–2. PubMed Abstract | Publisher Full Text\n\nNestorovich EM, Bezrukov SM: Designing inhibitors of anthrax toxin. Expert Opin Drug Discov. 2014; 9(3): 299–318. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBasha S, Rai P, Poon V, et al.: Polyvalent inhibitors of anthrax toxin that target host receptors. Proc Natl Acad Sci U S A. 2006; 103(36): 13509–13. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWorkman A, Heaton M: Table S1. MALDI-TOF MS assay design for 14 ITGB2 missense SNPs. figshare. Dataset. 2018."
}
|
[
{
"id": "42342",
"date": "08 Jan 2019",
"name": "Alan L. Archibald",
"expertise": [
"Reviewer Expertise molecular genetics",
"gene editing"
],
"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 a clear and thorough description of their analysis of sequence variation in the bovine ITGB2 gene that encodes CD18. They have not only identified and characterised sequence variants, particularly in two exons of interest, but also tested the effect of these variants in a functional assay in respect of binding to Mannheimia hemolytica leukotoxin. The methods used are described in exemplary detail.\nThere would be value in noting that some of the missense variants that are a key focus of the analyses have been reported by others previously and citing the relevant rs-identifiers: R5C rs517776076 N27K rs521222417 R120W rs109222573 R120Q rs522421330 D128G rs445709131 Q392K rs469652717 P656L rs379267066\n\nThe authors should be encouraged to submit their new variants to the European Variation Archive that now hosts sequence variation data for non-human species now that dbSNP has ceased this service. The authors should not only submit the sequence variants but also the genotype data and frequencies. Two of the variants listed above do not appear to have survived the migration from dbSNP to EVA and may need to be resubmitted (i.e. R120W / rs109222573; Q392K / rs469652717).\nThere a few (very few) minor typographical errors: 5 uM rather than 5 μM on page 4, column 2 line 11 a missing gap after the comma on page 6, column 2, line 1 50 ul rather than 50 μl on page 6, column 2, 5 line from the bottom It may be appropriate to mention the contribution of Mannheimia hemolytica to \"shipping fever\" and note the cost of this disease and pathogen to the US cattle industry.\nThe authors discuss their results in the context of a report of gene editing the bovine ITGB2 gene. The paper by Shanthalingam et al. 2016 (cited as reference 15 in this manuscript) report that gene editing a single codon in the ITGB2 gene and generation of a fetus carrying the edited yielded cells that were resistant to Mannheimia hemeolytica in an in vitro assay. The current authors correctly note that there are no reports of live born calves carrying this edit in homozygous form and thus the viability of this gene edit for resistance to Mannheimia hemolytica remains unproven. However, none of the natural ITBG2 variants characterised by the authors delivers resistance. As the authors search for variation was thorough and should have identified almost all variants of useful frequency, then designing variant signal peptide sequences that are not cleaved but do not bind the leukotoxin may be the way forward. The authors' discussion of these points could be more explicit.\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": "44372",
"date": "06 Mar 2019",
"name": "Holly L. Neibergs",
"expertise": [
"Reviewer Expertise Genetic susceptibility to 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\nThe authors have presented a well written article describing a nicely designed and thorough set of experiments to determine if variants in the CD18 signal peptide gene (ITGB2) effect Lkt binding in leukocytes. Lkt is a protein secreted by M. haemolytica that activates an inflammatory response leading to fibrinous bronchopneumonia in cattle. M. haemolytica is an important bacterial pathogen associated with bovine respiratory disease that causes the death of over one million cattle annually in the U.S. alone. This article provides a better understanding of the molecular mechanisms involved in M. haemolytica infection that may lead to more effective means to prevent and treat this common disease in cattle.\nSuggested minor edits were very few:\nPage 3, line 1 of paragraph 1; insert “one of” after “Manheimia haemolytica is\" Page 6, line 10 of paragraph 2; replace “at run” with “and run at” after (Thermo Fisher Scientific) Page 6, line 17 of paragraph 3; remove “in” after “times” Page 9, line 9 of paragraph 2; I am not sure I would agree that a conserved sequence with less than 20% identity is remarkable\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": "44373",
"date": "09 Apr 2019",
"name": "Jonathan E. Beever",
"expertise": [
"Reviewer Expertise Molecular genetics of livestock species."
],
"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 excellent manuscript. The work is very well done and the conclusions drawn are well thought out. This manuscript will be well-sited from a scientific standpoint and the results are very translatable.\nComments:\nIn the Abstract -- 'Background' perhaps add bovine as in \" ...binds to the CD18 receptor on bovine leukocytes, ...\" followed by \"Binding of Lkt to bovine CD18 results from an intact 22-amino acid signal peptide which remains uncleaved in all ruminant species.\" Although these are minor changes, they reflect the knowledge that this binding phenomenon is described specifically in cattle and is now inferred in other susceptible ruminant species based on the Lkt/SD18 signal peptide binding studies in cattle.\nA similar argument could also be applied to the beginning of the second paragraph of the Introduction. However, this paragraph is better in its logical progression.\nIt is completely unclear as to why the whole-genome sequencing of the BL3 cell line was performed. Suggest removing that section from the M&M.\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/7-1985
|
https://f1000research.com/articles/5-1408/v1
|
17 Jun 16
|
{
"type": "Software Tool Article",
"title": "RNA-seq analysis is easy as 1-2-3 with limma, Glimma and edgeR",
"authors": [
"Charity W. Law",
"Monther Alhamdoosh",
"Shian Su",
"Gordon K. Smyth",
"Matthew E. Ritchie",
"Charity W. Law",
"Monther Alhamdoosh",
"Shian Su"
],
"abstract": "The ability to easily and efficiently analyse RNA-sequencing data is a key strength of the Bioconductor project. Starting with counts summarised at the gene-level, a typical analysis involves pre-processing, exploratory data analysis, differential expression testing and pathway analysis with the results obtained informing future experiments and validation studies. In this workflow article, we analyse RNA-sequencing data from the mouse mammary gland, demonstrating use of the popular edgeR package to import, organise, filter and normalise the data, followed by the limma package with its voom method, linear modelling and empirical Bayes moderation to assess differential expression and perform gene set testing. This pipeline is further enhanced by the Glimma package which enables interactive exploration of the results so that individual samples and genes can be examined by the user. The complete analysis offered by these three packages highlights the ease with which researchers can turn the raw counts from an RNA-sequencing experiment into biological insights using Bioconductor.",
"keywords": [
"RNA sequencing",
"data analysis",
"gene expression"
],
"content": "Introduction\n\nRNA-sequencing (RNA-seq) has become the primary technology for gene expression profiling and the detection of differentially expressed genes between two or more conditions is one of the most commonly asked questions using this approach. The edgeR1 and limma packages2 available from the Bioconductor project3 offer a well-developed suite of statistical methods for dealing with this type of analysis. In this article, we describe an edgeR - limma workflow for analysing RNA-seq data that takes gene-level counts as its input, and moves through pre-processing and exploratory data analysis before obtaining lists of differentially expressed (DE) genes and gene signatures. This analysis is enhanced through the use of interactive graphics from the Glimma package4, that allows for a more detailed exploration of the data at both the sample and gene-level than is possible using static R plots.\n\nThe experiment analysed in this workflow is from Sheridan et al. (2015)5 and consists of three cell populations (basal, luminal progenitor (LP) and mature luminal (ML)) sorted from the mammary glands of female virgin mice, each profiled in triplicate. RNA samples were sequenced across three batches on an Illumina HiSeq 2000 to obtain 100 base-pair single-end reads. The analysis outlined in this article assumes that reads obtained from an RNA-seq experiment have been aligned to an appropriate reference genome and summarised into counts associated with gene-specific regions. In this instance, reads were aligned to the mouse reference genome (mm10) using the R based pipeline available in the Rsubread package (specifically the align function6 followed by featureCounts7 for gene-level summarisation based on the in-built mm10 RefSeq-based annotation). Count data for these samples can be downloaded from the Gene Expression Omnibus (GEO) http://www.ncbi.nlm.nih.gov/geo/ using GEO Series accession number GSE63310. Further information on experimental design and sample preparation is also available from GEO under this accession number.\n\n\nData packaging\n\nTo get started with this analysis, download the file GSE63310_RAW.tar available online from http://www.ncbi.nlm.nih.gov/geo/download/?acc=GSE63310&format=file, and extract the relevant files from this archive. Each of these text files contains the raw gene-level counts for a given sample. Note that our analysis only includes the basal, LP and ML samples from this experiment (see associated file names below).\n\n\n\n\n\nWhilst each of the nine text files can be read into R separately and combined into a matrix of counts, edgeR offers a convenient way to do this in one step using the readDGE function. The resulting DGEList-object contains a matrix of counts with 27,179 rows associated with unique Entrez gene identifiers (IDs) and nine columns associated with the individual samples in the experiment.\n\n\n\n\n\n\n\n\n\nIf the counts from all samples were stored in a single file, the data can be read into R and then converted into a DGEList-object using the DGEList function.\n\nFor downstream analysis, sample-level information related to the experimental design needs to be associated with the columns of the counts matrix. This should include experimental variables, both biological and technical, that could have an effect on expression levels. Examples include cell type (basal, LP and ML in this experiment), genotype (wild-type, knock-out), phenotype (disease status, sex, age), sample treatment (drug, control) and batch information (date experiment was performed if samples were collected and analysed at distinct time points) to name just a few.\n\nOur DGEList-object contains a samples data frame that stores both cell type (or group) and batch (sequencing lane) information, each of which consists of three distinct levels. Note that within x$samples, library sizes are automatically calculated for each sample and normalisation factors are set to 1. For simplicity, we remove the GEO sample IDs (GSM*) from the column names of our DGEList-object x.\n\n\n\n\n\n\n\n\n\n\n\nA second data frame named genes in the DGEList-object is used to store gene-level information associated with rows of the counts matrix. This information can be retrieved using organism specific packages such as Mus.musculus8 for mouse (or Homo.sapiens9 for human) or the biomaRt package10,11 which interfaces the Ensembl genome databases in order to perform gene annotation. The type of information that can be retrieved includes gene symbols, gene names, chromosome names and locations, Entrez gene IDs, Refseq gene IDs and Ensembl gene IDs to name just a few. biomaRt primarily works off Ensembl gene IDs, whereas Mus.musculus packages information from various sources and allows users to choose between many different gene IDs as the key. The Entrez gene IDs available in our dataset were annotated using the Mus.musculus package to retrieve associated gene symbols and chromosome information.\n\n\n\n\n\nAs with any gene ID, Entrez gene IDs may not map one-to-one to the gene information of interest. It is important to check for duplicated gene IDs and resolve them to ensure the gene order is consistent between our annotation and the counts in the DGEList-object.\n\n\n\n\n\n\n\n\n\nThe match-ing process above takes care of the duplication by keeping the first occurrence of each gene ID. Now that non-unique gene IDs have been resolved, the data frame of gene annotations is added to the data object:\n\n\n\nNow our data is packaged neatly in a DGEList-object containing raw count data with associated sample information and gene annotations.\n\n\n\n\n\n\nData pre-processing\n\nFor differential expression and related analyses, gene expression is rarely considered at the level of raw counts since libraries sequenced at a greater depth will result in higher counts. Rather, it is common practice to transform raw counts onto a scale that accounts for such library size differences. Popular transformations include counts per million (CPM), log2-counts per million (log-CPM), reads per kilobase of transcript per million (RPKM), and fragments per kilobase of transcript per million (FPKM).\n\nIn our analyses, CPM and log-CPM transformations are used regularly although they do not account for feature length differences which RPKM and FPKM values do. Whilst RPKM and FPKM values can just as well be used, CPM and log-CPM values can be calculated using a counts matrix alone and will suffice for the type of comparisons we are interested in. Differential expression analyses look at gene expression differences between conditions, rather than comparing expression across multiple genes or drawing conclusions on absolute levels of expression. In other words, gene lengths remain constant for comparisons of interest and any observed differences are a result of changes in condition rather than changes in gene length.\n\nHere raw counts are converted to CPM and log-CPM values using the cpm function in edgeR, where log-transformations use a prior count of 0.25 to avoid taking the log of zero. RPKM values are just as easily calculated as CPM values using the rpkm function in edgeR if gene lengths are available.\n\n\n\nAll datasets will include a mix of genes that are expressed and those that are not expressed. Whilst it is of interest to examine genes that are expressed in one condition but not in another, some genes are unexpressed throughout all samples. In fact, 19% of genes in this dataset have zero counts across all nine samples.\n\n\n\n\n\nGenes that are not expressed at a biologically meaningful level in any condition should be discarded to reduce the subset of genes to those that are of interest, and to reduce the number of tests carried out downstream when looking at differential expression. Upon examination of log-CPM values, it can be seen that a large proportion of genes within each sample is unexpressed or lowly-expressed (Figure 1A). Using a nominal CPM value of 1 (which is equivalent to a log-CPM value of 0) a gene is deemed to be expressed in a given sample if its transformed count is above this threshold, and unexpressed otherwise. Genes must be expressed in at least one group (or roughly any three samples across the entire experiment) to be kept for downstream analysis.\n\nThe density of log-CPM values for raw pre-filtered data (A) and post-filtered data (B) are shown for each sample. Dotted vertical lines mark the log-CPM of zero threshold (equivalent to a CPM value of 1) used in the filtering step.\n\n\n\n\n\nUsing this criterion, the number of genes is reduced to approximately half the number that we started with (14,165 genes, Figure 1B). Note that subsetting the entire DGEList-object removes both the counts as well as the associated gene information. Code to produce Figure 1 is given below.\n\n\n\nDuring the sample preparation or sequencing process, external factors that are not of biological interest can affect the gene expression levels of individual samples. For example, samples processed in the first batch of an experiment may have been sequenced more deeply than samples processed in a second batch. It is assumed that all samples should have a similar range and distribution of expression values. Normalisation is generally required to ensure that the expression distributions of each sample are similar across the entire experiment.\n\nAny plot showing the per sample expression distributions, such as a density or boxplot, is useful in determining whether any samples are dissimilar to others. Distributions of log-CPM values are similar throughout all samples within this dataset (Figure 1B).\n\nNonetheless, normalisation by the method of trimmed mean of M-values (TMM)12 is performed using the calcNormFactors function in edgeR. The normalisation factors calculated here are used as a scaling factor for the library sizes. When working with DGEList-objects, these normalisation factors are automatically stored in x$samples$norm.factors. For this dataset the effect of TMM-normalisation is mild, as evident in the magnitude of the scaling factors, which are all relatively close to 1.\n\n\n\n\n\nTo give a better visual representation of the effects of normalisation, the data was duplicated then adjusted so that the counts of the first sample are reduced to 5% of their original values, and in the second sample they are inflated to be 5-times larger.\n\n\n\nFigure 2 shows the expression distribution of samples for unnormalised and normalised data, where distributions are noticeably different pre-normalisation and are similar post-normalisation. Here the first sample has a small TMM scaling factor of 0.05, whereas the second sample has a large scaling factor of 6.13 – neither values are close to 1.\n\n\n\n\n\nIn our opinion, one of the most important exploratory plots to examine for gene expression analyses is the multidimensional scaling (MDS) plot, or similar. The plot shows similarities and dissimilarities between samples in an unsupervised manner so that one can have an idea of the extent to which differential expression can be detected before carrying out formal tests. Ideally, samples would cluster well within the primary condition of interest, and any sample straying far from its group could be identified and followed up for sources of error or extra variation. If present, technical replicates should lie very close to one another.\n\nExample data: Boxplots of log-CPM values showing expression distributions for unnormalised data (A) and normalised data (B) for each sample in the modified dataset where the counts in samples 1 and 2 have been scaled to 5% and 500% of their original values respectively.\n\nSuch a plot can be made in limma using the plotMDS function. The first dimension represents the leading-fold-change that best separates samples and explains the largest proportion of variation in the data, with subsequent dimensions having a smaller effect and being orthogonal to the ones before it. When experimental design involves multiple factors, it is recommended that each factor is examined over several dimensions. If samples cluster by a given factor in any of these dimensions, it suggests that the factor contributes to expression differences and is worth including in the linear modelling. On the other hand, factors that show little or no effect may be left out of downstream analysis.\n\nIn this dataset, samples can be seen to cluster well within experimental groups over dimension 1 and 2, and then separate by sequencing lane (sample batch) over dimension 3 (Figure 3). Keeping in mind that the first dimension explains the largest proportion of variation in the data, notice that the range of values over the dimensions become smaller as we move to higher dimensions. Whilst all samples cluster by groups, the largest transcriptional difference is observed between basal and LP, and basal and ML over dimension 1. For this reason, it is expected that pairwise comparisons between cell populations will result in a greater number of DE genes for comparisons involving basal samples, and relatively small numbers of DE genes when comparing ML to LP. In other datasets, samples that do not cluster by their groups of interest may also show little or no evidence of differential expression in the downstream analysis.\n\nTo create the MDS plots, different colour groupings are assigned to factors of interest. Dimensions 1 and 2 are examined using the color grouping defined by cell types.\n\n\n\nDimensions 3 and 4 are examined using the colour grouping defined by sequencing lanes (batch).\n\n\n\nMDS plots of log-CPM values over dimensions 1 and 2 with samples coloured and labeled by sample groups (A) and over dimensions 3 and 4 with samples coloured and labeled by sequencing lane (B). Distances on the plot correspond to the leading fold-change, which is the average (root-mean-square) log2-fold-change for the 500 genes most divergent between each pair of samples by default.\n\nAlternatively, the Glimma package offers the convenience of an interactive MDS plot where multiple dimensions can be explored. The glMDSPlot function generates an html page (that is opened in a browser if launch=TRUE) with an MDS plot in the left panel and a barplot showing the proportion of variation explained by each dimension in the right panel. Clicking on the bars of the bar plot changes the pair of dimensions plotted in the MDS plot, and hovering over the individual points reveals the sample label. The colour scheme can be changed as well to highlight cell population or sequencing lane (batch). An interactive MDS plot of this dataset can be found at http://bioinf.wehi.edu.au/folders/limmaWorkflow/glimma-plots/MDS-Plot.html.\n\n\n\n\nDifferential expression analysis\n\nIn this study, it is of interest to see which genes are expressed at different levels between the three cell populations profiled. In our analysis, linear models are fitted to the data with the assumption that the underlying data is normally distributed. To get started, a design matrix is set up with both the cell population and sequencing lane (batch) information.\n\n\n\n\n\nFor a given experiment, there are usually several equivalent ways to set up an appropriate design matrix. For example, ~0+group+lane removes the intercept from the first factor, group, but an intercept remains in the second factor lane. Alternatively, ~group+lane could be used to keep the intercepts in both group and lane. Understanding how to interpret the coefficients estimated in a given model is key here. We choose the first model for our analysis, as setting up model contrasts is more straight forward in the absence of an intercept for group. Contrasts for pairwise comparisons between cell populations are set up in limma using the makeContrasts function.\n\n\n\n\n\nA key strength of limma’s linear modelling approach, is the ability accommodate arbitrary experimental complexity. Simple designs, such as the one in this workflow, with cell type and batch, through to more complicated factorial designs and models with interaction terms can be handled relatively easily. Where experimental or technical effects can be modelled using a random effect, another possibility in limma is to estimate correlations using duplicateCorrelation by specifying a block argument for both this function and in the lmFit linear modelling step.\n\nIt has been shown that for RNA-seq count data, the variance is not independent of the mean13 – this is true of raw counts or when transformed to log-CPM values. Methods that model counts using a Negative Binomial distribution assume a quadratic mean-variance relationship. In limma, linear modelling is carried out on the log-CPM values which are assumed to be normally distributed and the mean-variance relationship is accommodated using precision weights calculated by the voom function.\n\nWhen operating on a DGEList-object, voom converts raw counts to log-CPM values by automatically extracting library sizes and normalisation factors stored in the object. For a matrix of counts, the method of normalisation can be specified within voom using the normalize.method (by default no normalisation is performed).\n\nThe mean-variance relationship of log-CPM values for this dataset is shown in Figure 4A. Typically, the “voom-plot” shows a decreasing trend between the means and variances resulting from a combination of technical variation in the sequencing experiment and biological variation amongst the replicate samples from different cell populations. Experiments with high biological variation usually result in flatter trends, where variance values plateau at high expression values. Experiments with low biological variation tend to result in sharp decreasing trends.\n\nMoreover, the voom-plot provides a visual check on the level of filtering performed upstream. If filtering of lowly-expressed genes is insufficient, a drop in variance levels can be observed at the low end of the expression scale due to very small counts. If this is observed, one should return to the earlier filtering step and increase the expression threshold applied to the dataset.\n\nMeans (x-axis) and variances (y-axis) of each gene are plotted to show the dependence between the two before voom is applied to the data (A) and how the trend is removed after voom precision weights are applied to the data (B). The plot on the left is created within the voom function which extracts residual variances from fitting linear models to log-CPM transformed data. Variances are then rescaled to quarter-root variances (or square-root of standard deviations) and plotted against the mean expression of each gene. The means are log2-transformed mean-counts with an offset of 0.5. The plot on the right is created using plotSA which plots log2 residual standard deviations against mean log-CPM values. The average log2 residual standard deviation is marked by a horizontal blue line. In both plots, each black dot represents a gene and a red curve is fitted to these points.\n\nWhere sample-level variation is evident from earlier inspections of the MDS plot, the voomWithQualityWeights function can be used to simultaneously incorporate sample-level weights together with the abundance dependent weights estimated by voom14. For an example of this, see Liu et al. (2016)15.\n\n\n\n\n\nNote that the other data frames stored within the DGEList-object that contain gene- and sample-level information, are retained in the EList-object v created by voom. The v$genes data frame is equivalent to x$genes, v$targets is equivalent to x$samples, and the expression values stored in v$E is analogous to x$counts, albeit on a transformed scale. In addition to this, the voom EList-object has a matrix of precision weights v$weights and stores the design matrix in v$design.\n\nLinear modelling in limma is carried out using the lmFit and contrasts.fit functions which can be used for both microarray and RNA-seq data and fit a separate model to the expression values for each gene. Next, empirical Bayes moderation is carried out by the eBayes function which borrows information across all the genes to obtain more precise estimates of gene-wise variability16. The model’s residual variances are plotted against average expression values in Figure 4B. It can be seen from this plot that the variance is no longer dependent on the mean expression level.\n\n\n\nFor a quick look at differential expression levels, the number of significantly up- and down-regulated genes can be summarised in a table. Significance is defined using an adjusted p-value cutoff that is set at 5% by default. For the comparison between expression levels in basal and LP, 4,127 genes are found to be down-regulated in basal relative to LP and 4,298 genes are up-regulated in basal relative to LP – a total of 8,425 DE genes. A total of 8,510 DE genes are found between basal and ML (4,338 down- and 4,172 up-regulated genes), and a total of 5,340 DE genes are found between LP and ML (2,895 down- and 2,445 up-regulated). The larger numbers of DE genes observed for comparisons involving the basal population are consistent with our observations from the MDS plots.\n\n\n\n\n\nSome studies require more than an adjusted p-value cut-off. For a stricter definition on significance, one may require log-fold-changes (log-FCs) to be above a minimum value. The treat method17 can be used to calculate p-values from empirical Bayes moderated t-statistics with a minimum log-FC requirement. The number of differentially expressed genes are reduced to a total of 3,135 DE genes for basal versus LP, 3,270 DE genes for basal versus ML, and 385 DE genes for LP versus ML when testing requires genes to have a log-FC that is significantly greater than 1 (equivalent to a 2-fold difference between cell types on the original scale).\n\n\n\n\n\nGenes that are DE in multiple comparisons can be extracted using the results from decideTests, where 0s represent genes that are not DE, 1s represent genes that are up-regulated, and -1s represent genes that are down-regulated. A total of 2,409 genes are DE in both basal versus LP and basal versus ML (Figure 5), twenty of which are listed below. The write.fit function can be used to extract and write results for all three comparisons to a single output file.\n\n\n\n\n\n\n\n\n\n\n\nThe top DE genes can be listed using topTreat for results using treat (or topTable for results using eBayes). By default topTreat arranges genes from smallest to largest adjusted p-value with associated gene information, log-FC, average log-CPM, moderated t-statistic, raw and adjusted p-value for each gene. The number of top genes displayed can be specified, where n=Inf includes all genes. Genes Cldn7 and Rasef are amongst the top DE genes for both basal versus LP and basal versus ML.\n\nThe number of genes that are not DE in either comparison are marked in the bottom-right.\n\n\n\n\n\n\n\n\n\nTo summarise results for all genes visually, mean-difference plots, which display log-FCs from the linear model fit against the average log-CPM values can be generated using the plotMD function, with the differentially expressed genes highlighted.\n\n\n\nGlimma extends this functionality by providing an interactive mean-difference plot via the glMDPlot function. The output of this function is an html page, with summarised results in the left panel (similar to what is output by plotMD), and the log-CPM values from individual samples in the right panel, with a table of results below the plots (Figure 6). This interactive display allows the user to search for particular genes based on their Gene symbol, which is not possible in a static R plot. The glMDPlot function is not limited to mean-difference plots, with a default version allowing a data frame to be passed with the user able to select the columns of interest to plot in the left panel.\n\nSummary data (log-FCs versus log-CPM values) are shown in the left panel which is linked to the individual values per sample for a selected gene in the right panel. A table of results is also displayed below these figures, along with a search bar to allow users to look up a particular gene using its Gene symbol identifier, e.g. Clu.\n\n\n\nThe mean-difference plot generated by the command above is available online (see http://bioinf.wehi.edu.au/folders/limmaWorkflow/glimma-plots/MD-Plot.html). The interactivity provided by the Glimma package allows additional information to be presented in a single graphical window. Glimma is implemented in R and Javascript, with the R code generating the data which is converted into graphics using the Javascript library D3 (https://d3js.org), with the Bootstrap library handling layouts and Datatables generating the interactive searchable tables. This allows plots to be viewed in any modern browser, which is convenient for including them as linked files from an Rmarkdown report of the analysis.\n\nPlots shown previously include either all of the genes that are expressed in any one condition (such as the Venn diagram of common DE genes or mean-difference plot) or look at genes individually (log-CPM values shown in right panel of the interactive mean-difference plot). Heatmaps allow users to look at the expression of a subset of genes. This can give useful insight into the expression of individual groups and samples without losing perspective of the overall study when focusing on individual genes, or losing resolution when examining patterns averaged over thousands of genes at the same time.\n\nA heatmap is created for the top 100 DE genes (as ranked by adjusted p-value) from the basal versus LP contrast using the heatmap.2 function from the gplots package (Figure 7). The heatmap correctly clusters samples into cell type and rearranges the order of genes to form blocks of similar expression. From the heatmap, we observe that the expression of ML and LP samples are very similar for the top 100 DE genes between basal and LP.\n\n\n\nExpression across each gene (or row) have been scaled so that mean expression is zero and standard deviation is one. Samples with relatively high expression within a gene are marked in red, samples with relatively low expression are marked in blue. Lighter shades and white represent genes with intermediate expression levels. Samples and genes have been reordered by the method of hierarchical clustering. A dendrogram is shown for the clustering of samples.\n\n\n\n\nGene set testing with camera\n\nWe finish this analysis with gene set testing by applying the camera method18 to the c2 gene signatures from the Broad Institute's MSigDB c2 collection19 that have been adapted for mouse and are available as Rdata objects from http://bioinf.wehi.edu.au/software/MSigDB/.\n\n\n\n\n\n\n\n\n\n\n\n\n\nThe camera function performs a competitive test to assess whether the genes in a given set are highly ranked in terms of differential expression relative to genes that are not in the set. It uses limma’s linear model framework, taking both the design matrix and contrast matrix (if present) and accommodates the observational-level weights from voom in the testing procedure. After adjusting the variance of the resulting gene set test statistic by a variance inflation factor, that depends on the gene-wise correlation (which can be estimated from the data or specified by the user) and the size of the set, a p-value is returned and adjusted for multiple testing. Here we set inter.gene.cor to 0.01 for all gene sets which produces a less conservative test that performs well in many practical situations.\n\nThis experiment is the RNA-seq equivalent of Lim et al. (2010)20, who used Illumina microarrays to profile the same sorted cell populations, so it is reassuring to see the gene signatures from this earlier publication coming up at the top of the list for each contrast. We make a barcodeplot of the Lim et al. (2010) Mature Luminal gene sets (Up and Down) in the LP versus ML contrast. Note that these sets go in the opposite direction in our dataset due to our parameterization which compares LP against ML rather than the other way around (if the contrast were reversed, the directions would be consistent).\n\n\n\n\nSoftware availability\n\nThis RNA-seq workflow makes use of various packages available from version 3.4 of the Bioconductor project, running on R21 version 3.3.0 or higher. Besides the packages highlighted in this article (limma, Glimma and edgeR) it requires a number of other software packages, including gplots22 and RColorBrewer and the gene annotation package Mus.musculus. This document was compiled using knitr23–25. Version numbers for all packages used are shown below. Code to perform this analysis is available from the Supplementary Materials website at http://bioinf.wehi.edu.au/folders/limmaWorkflow/ and as a Bioconductor workflow from http://www.bioconductor.org/help/workflows/.\n\nThe experiment of Lim et al. (2010) is very similar to the current one, with the same sorting strategy used to obtain the different cell populations, except that microarrays were used instead of RNA-seq to profile gene expression. Note that the inverse correlation (the up gene set is down and the down gene set is up) is a result of the way the contrast has been set up (LP versus ML) – if reversed, the directionality would agree.\n\n\n\n",
"appendix": "Author contributions\n\n\n\nAll authors were involved in writing and contributing code for the article.\n\n\nCompeting interests\n\n\n\nThe authors declare that they have no competing interests.\n\n\nGrant information\n\nThis worked was funded by the National Health and Medical Research Council (NHMRC) (Fellowship GNT1058892 and Program GNT1054618 to GKS, Project GNT1050661 to MER and GKS and Fellowship GNT1104924 to MER), Victorian State Government Operational Infrastructure Support and Australian Government NHMRC IRIISS.\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\nWe thank Dr Julie Sheridan for generating this dataset and for advice on its analysis.\n\n\nReferences\n\nRobinson MD, McCarthy DJ, Smyth GK: edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010; 26(1): 139–140. 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\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\nSu S, Ritchie ME: Glimma: Interactive HTML graphics for RNA-seq data. 2016; R package version 1.1.1.\n\nSheridan JM, Ritchie ME, Best SA, et al.: A pooled shRNA screen for regulators of primary mammary stem and progenitor cells identifies roles for Asap1 and Prox1. BMC Cancer. 2015; 15(1): 221. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiao Y, Smyth GK, Shi W: The Subread aligner: fast, accurate and scalable read mapping by seed-and-vote. Nucleic Acids Res. 2013; 41(10): e108. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiao Y, Smyth GK, Shi W: featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014; 30(7): 923–30. PubMed Abstract | Publisher Full Text\n\nBioconductor Core Team: Mus.musculus: Annotation package for the Mus.musculus object. 2016; R package version 1.3.1. Reference Source\n\nBioconductor Core Team: Homo.sapiens: Annotation package for the Homo.sapiens object. 2016; R package version 1.3.1. Reference Source\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\nDurinck S, Spellman PT, Birney E, et al.: Mapping identifiers for the integration of genomic datasets with the R/Bioconductor package biomaRt. Nat Protoc. 2009; 4(8): 1184–91. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRobinson MD, Oshlack A: A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 2010; 11(3): R25. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLaw CW, Chen Y, Shi W, et al.: Voom: Precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 2014; 15(2): R29. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu R, Holik AZ, Su S, et al.: Why weight? Modelling sample and observational level variability improves power in RNA-seq analyses. Nucleic Acids Res. 2015; 43(15): e97. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLiu R, Chen K, Jansz N, et al.: Transcriptional profiling of the epigenetic regulator Smchd1. Genom Data. 2015; 7: 144–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSmyth GK: Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol. 2004; 3(1): Article3. PubMed Abstract | Publisher Full Text\n\nMcCarthy DJ, Smyth GK: Testing significance relative to a fold-change threshold is a TREAT. Bioinformatics. 2009; 25(6): 765–71. PubMed Abstract | Publisher Full Text | Free Full Text\n\nWu D, Smyth GK: Camera: a competitive gene set test accounting for inter-gene correlation. Nucleic Acids Res. 2012; 40(17): e133. PubMed Abstract | Publisher Full Text | Free Full Text\n\nSubramanian A, Tamayo P, Mootha VK, et al.: Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005; 102(43): 15545–50. PubMed Abstract | Publisher Full Text | Free Full Text\n\nLim E, Wu D, Pal B, et al.: Transcriptome analyses of mouse and human mammary cell subpopulations reveal multiple conserved genes and pathways. Breast Cancer Res. 2010; 12(2): R21. PubMed Abstract | Publisher Full Text | Free Full Text\n\nR Development Core Team: R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, 2016. Reference Source\n\nWarnes GR, Bolker B, Bonebakker L, et al.: gplots: Various R Programming Tools for Plotting Data. 2016; R package version 3.0.1. Reference Source\n\nXie Y: knitr: A comprehensive tool for reproducible research in R. In V. Stodden, F. Leisch, and R. D. Peng, editors, Implementing Reproducible Computational Research. Chapman and Hall/CRC, 2014; ISBN 9781466561595. Reference Source\n\nXie Y: Dynamic Documents with R and knitr. Chapman and Hall/CRC, Boca Raton, Florida, 2nd edition, 2015; ISBN 9781498716963. Reference Source\n\nXie Y: knitr: A General-Purpose Package for Dynamic Report Generation in R. 2016; R package version 1.12.3. Reference Source"
}
|
[
{
"id": "14436",
"date": "23 Jun 2016",
"name": "James W. MacDonald",
"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\nIn this manuscript the authors perform a step-by-step analysis of a public RNA-Seq data set, using the Bioconductor packages edgeR, limma, and Glimma. The analysis proceeds through four stages (data preparation, preprocessing, univariate analyses, and gene set testing), with each stage broken up into several discrete steps. Each step is clearly explained and is accompanied by R code and output, so the reader can follow along if desired.\nThe target audience for this paper appears to be someone with passing familiarity with both R and statistics, without requiring expertise in either. Most of the code should be easy to understand, and any complex sections are clearly explained in the text. Similarly, the authors outline the analytic choices they make (e.g., filtering out unexpressed genes, specifying model coefficients, etc) at a more accessible level.\nWhile both edgeR and limma are well known, popular Bioconductor packages, Glimma is a new package that was released in April 2016. This package uses the d3.js JavaScript library to generate interactive HTML documents that can be viewed locally, rather than needing to be accessed from a server running R (as, say a shiny app requires). This is an exciting development, and is unfortunately not as compelling as it could be, if the Glimma plots were part of the HTML version of the manuscript rather than provided as links.\nThis is a well written paper, and is a useful contribution to the literature; while each package has either an extensive user's guide or vignette, by necessity these documents pertain only to the package at hand. Most RNA-Seq analyses require a combination of multiple packages to complete, and this paper provides a clear example.\nMajor comments\nNone.\n\nMinor comments\nIn the section 'Organising gene annotations', the code used to subset the one-to-many mappings is needlessly complex. Simply doing something like\ngenes <- genes[!duplicated(genes[,1]),]\nwill accomplish the same thing, in a more straightforward way.\nIn the section 'Removing heteroscedasticity from count data', the authors state:\n\"When operating on a DGEList-object, voom converts raw counts to log-CPM values by automatically extracting library sizes and normalisation factors stored in the object. For a matrix of counts, the method of normalisation can be specified within voom using the normalize.method (by default no normalisation is performed).\"\nThis is confusing; the authors already showed in an earlier section ('Normalising gene expression distributions') that converting to log-CPM using TMM normalization factors will do a shift-normalization (which is what voom will do in this instance). The normalize.method argument to voom specifies additional normalization methods that can be applied to the matrix of TMM normalized log-CPM values.",
"responses": [
{
"c_id": "2336",
"date": "30 Nov 2016",
"name": "Charity Law",
"role": "Author Response",
"response": "Dear James, thank-you for your comments. We have considered your suggestions and made changes to simplify the section on ‘Organising gene annotations’ and to clarify that voom offers additional normalisation ontop of the TMM-normsalisation that is carried out."
}
]
},
{
"id": "14439",
"date": "06 Jul 2016",
"name": "Maria A. Doyle",
"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\nEditorial Note from F1000Research - 7th July 2016:\nthe status of this review was changed after publication from \"Approved with Reservations\" to \"Approved\". The referee's reservation was caused by using the current release (Bioconductor 3.3) as opposed to the development release (Bioconductor 3.4) as used in the article.\n\nThe following text has been removed:\nMy only reservation to approving the article is that I encountered these two errors when running the Glimma code in the article. I used R 3.3.0 and the most recent version of Glimma from Bioconductor (Glimma 1.0.0) but the sessionInfo() in the article shows Glimma 1.1.1 was used which may account for the errors:\n\n1) With this line of code no plot was produced and it gave the error below: glMDSPlot(lcpm, labels=paste(group, lane, sep=\"_\"), +\n\ngroups=x$samples[,c(2,5)], launch=TRUE) Error in glScatter.default(points, xval = \"dim1\", yval = \"dim2\", xlab = \"Dimension 1\", :\n\ngroup does not correspond to a column\n\n2) With this line of code the table (shown in Fig 6.) didn't appear, just the 2 plots: glMDPlot(tfit, coef=1, status=dt[,1], main=colnames(tfit)[1],\n\ncounts=x$counts, samples=colnames(x), anno=x$genes, groups=group,\n\nid.column=\"ENTREZID\", display.columns=c(\"SYMBOL\", \"ENTREZID\"),\n\nsearch.by=\"SYMBOL\", launch=FALSE)\nThis article is a very nice description (nicely written, nicely explained) of an RNA-seq workflow, starting from gene counts and proceeding through a standard differential expression analysis, using 3 complementary Bioconductor packages edgeR, limma & Glimma. It explains clearly how to perform the steps of a differential expression analysis and also how to generate useful plots for visualising the data e.g. MDSplot, MDplot, barcodeplot, heatmap etc.\nTwo of the packages, edgeR and limma, are well established while Glimma is the new kid on the block. It's a tool along the lines of the web-based Degust (http://www.vicbioinformatics.com/degust/), in that it enables interactive exploration of RNA-seq data. It would seem to be a very useful addition to R RNA-seq workflows as the interactive html plots it can generate (MDS, MD etc) make exploring the results easier (especially by non-R savvy collaborators) and it could help save an analyst's time and effort through not having to reproduce static plots e.g. to highlight different genes. With the Glimma MDplot I like that you can search for a gene and see a plot of the log-cpm counts for that gene in the samples.\nIt was great to be able to try out the workflow really easily by downloading the data file linked to in the paper (from GEO), with no processing required other than unzipping the files, and the code in the article all worked, giving the same results shown in the paper with the exception of two errors described below.\nMinor comments:\nThe article says the workflow is available from Bioconductor here http://www.bioconductor.org/help/workflows/ but it doesn't seem to be there, however it is available at the other location mentioned http://bioinf.wehi.edu.au/folders/limmaWorkflow/\nThe section title \"Removing genes that are not expressed\" suggests only genes with no expression are removed, however the paragraph then explains that genes lowly expressed are also removed so for greater clarity maybe that could be changed to something like \"Removing genes that are not sufficiently expressed\".\nIn the section \"Organising gene annotations\" I think it's good the authors point out to check for duplicated genes, however in this case the duplication appears to be due to extracting the TXCHROM column (as some genes are reported as being present on more than one chromosome) but the TXCHROM information is not used in this workflow and if the TXCHROM column is omitted then there are no duplicates so it might be worth mentioning that.\nWith this bit \"Differential expression analyses look at gene expression differences between conditions, rather than comparing expression across multiple genes or drawing conclusions on absolute levels of expression. In other words, gene lengths remain constant for comparisons of interest and any observed differences are a result of changes in condition rather than changes in gene length.\" Wouldn't one caveat to this be if there was a significant change in the length of the isoform(s) expressed from a gene (e.g. from expression of a long isoform to a short isoform) as then the assumption of no change in gene length would no longer be valid.\nIn the online version Fig.7 is in the middle of a code block could it be moved below.\nFigs 5-8 have bold headings but Figs 1-4 don't is that an error.\nCould perhaps modify title to reflect the order in which the tools are used - edger, limma & Glimma.\nIt would be nice for consistency if the colours in the Glimma interactive MDS plots matched the colours used in the static MDS plots in Fig 3.",
"responses": [
{
"c_id": "2058",
"date": "07 Jul 2016",
"name": "Maria A. Doyle",
"role": "Reviewer Response",
"response": "I've just requested that the status of this article be changed to Approved as the code works fine with the development release of Bioconductor 3.4 but could it be made clearer in the article that 3.4 must be used for the code to work as shown."
},
{
"c_id": "2335",
"date": "30 Nov 2016",
"name": "Charity Law",
"role": "Author Response",
"response": "Dear Maria, we have made changes according to your suggestions, including a change in the section title to \"Removing genes that are lowly expressed\", and we have added comments to point out that our analysis assumes no differential isoform usage. We have also simplified our example in the \"Organising gene annotations\" section. Thanks for pointing out the inconsistencies in the figure captions. They have now been addressed. Indeed, it would be nice to make the colours consistent between the interactive and static MDS plots. We like the colours used in the static plot and can have the interactive version of the plot updated once color specification becomes available to the MDS plotting function in Glimma."
}
]
},
{
"id": "14437",
"date": "06 Jul 2016",
"name": "Jovana Maksimovic",
"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\nThis workflow outlines a step-by-step differential expression analysis of a publicly available RNA-seq dataset using several well-established Bioconductor packages such as limma and edgeR and the novel Glimma package. The workflow describes in detail the key steps that are generally performed as part of most standard, count-based RNAseq analyses: pre-processing, exploratory data analysis, differential expression testing and pathway analysis. Each of the steps is carefully explained and broken down into a series of logical substeps, which also include R code and example output. Furthermore, this workflow is a good example of how multiple Bioconductor packages can be applied in a real-world RNAseq analysis.\n\nOverall, I think this is a very clearly written and well-explained workflow that not only highlights the utility of combining Bioconductor packages for RNAseq analysis but also demystifies the process for those who may be interested in learning what an RNAseq analysis involves. Although some familiarity with R and statistical concepts would be advantageous, they are not necessary to be able to follow the logic behind the steps and what each step is trying to achieve.\nI have tested the R code under R 3.3.0 and Bioconductor 3.0 and it ran without errors.\nMinor comments:\nAs this workflow will be of particular use for those who are very new to RNAseq analysis, I think it should be explained why 1 CPM is chosen as the cutoff for lowly expressed genes. In addition, it may be useful to show the relationship between CPM and raw counts and how the selected cutoff relates to numbers of reads. In the gene set testing section, it is worth mentioning that the Hallmark gene sets (also available from the Broad MSigDB) may be a good starting point for pathway analysis as they “summarise and represent specific well-defined, states or processes”. Also, for those who are new to RNAseq analysis, it may be worth mentioning that there are other gene set testing methods available in limma e.g. roast etc. and a very brief explanation of when the different methods are appropriate.\n\nGeneral comment regarding Glimma plots: it would be handy to see the gene symbol (if available) as well as EntrezID in the title of the scatter plot next to the MD plot.",
"responses": [
{
"c_id": "2334",
"date": "30 Nov 2016",
"name": "Charity Law",
"role": "Author Response",
"response": "Dear Jovana, thank-you for the helpful suggestions. We have added extra text to explain why a CPM value of 1 was chosen and how this value relates to raw counts. We have also given more detail on c2 and Hallmark gene sets, as well as mentioning roast and when one might chose to use roast over camera."
}
]
}
] | 1
|
https://f1000research.com/articles/5-1408
|
https://f1000research.com/articles/7-1978/v1
|
24 Dec 18
|
{
"type": "Research Article",
"title": "Association of JJ stent insertion and sexual function: A cohort study",
"authors": [
"Rizky Fawzi",
"Zulfikar Ali",
"Zulfikar Ali"
],
"abstract": "Background: Indonesia is known as one of the world’s stone belt areas in Asia. Management of urolithiasis cannot be separated with the role of JJ stent insertion. However, a limited number of prior studies show that a patient with JJ stent is at risk for sexual function disorder. This study aims to evaluate the association of JJ stent insertion with sexual function, both in men and women. Methods: This is a cohort study and the subjects were patients who had undergone JJ stent insertion in July - November 2017 at Kardinah Regional Hospital, Tegal, Central Java. This study was approved by the Research Ethic Committee of Kardinah Hospital (#445/3840/2017). Data were taken using standardized self-administered questionnaires before and after insertion of the JJ stent. Male sexual function was assessed using the International Index of Erectile Function (IIEF) score, while female sexual function was assessed using Female Sexual Function Index (FSFI). Result: 60 male patients, with a mean of age 51.1 + 10.6 years, and 33 female patients, with mean of age 49.6 + 10.6 years old, underwent JJ stent insertion. A significant association was found in women before and after JJ stent insertion (p<0.05), with FSFI score 23.62 + 0.64 before insertion and 16.7 + 0.52 after insertion. A similar result was also found in men with total IIEF score 49.55 + 2.3 before JJ stent insertion and 38.4 + 1.7 after insertion. Conclusion: This study confirms that JJ stent insertion may cause a disturbance of sexual function. However, the mechanism is not clear yet.",
"keywords": [
"sexual function",
"FSFI",
"IIEF",
"JJ stent",
"Tegal"
],
"content": "Introduction\n\nSexual function plays an important role in the quality of life both in men and women. Sexual function is influenced by the complex interactions between the nervous, vascular, endocrine, and psychological systems. Sexual function in men is closely related to erectile dysfunction. Erectile dysfunction is defined as the inability to achieve and maintain an erection in order to achieve sexual satisfaction. The prevalence of erectile dysfunction varies between 4.5 and 53.5% depending on various studies according to clinical, methodological, demographic and aetiologic factors1. In women the assessment of sexual disorders itself is quite difficult, unlike in men.\n\nRisk factors for erectile dysfunction include drugs, diabetes atherosclerosis, depression, anxiety disorders, hormones, radiotherapy, nervous system disorders, chronic renal failure, smoking, and surgical intervention. Major pelvic surgical procedures such as radical prostatectomy, trans-urethral resection of prostate, as well as other rectum or urethral surgery may cause sexual function impairment1. In recent years, there have been reports of sexual dysfunction in other endo-urological procedures.\n\nUrinary tract stones are the most common urological case in Indonesia. In the general population, the risk of formation of stones in a lifetime is about 10.2% with peak incidence occurring at the age of 20–40 years2. The management of urinary stones varies depending on the position, size and number of stones, the patient's anatomical condition, and the surgeon's own experience. Percutaneous nephrolithotomy (PCNL), ureteroscopy (URS), laparoscopy or open ureter are the main procedures of the kidney and ureter for the removal of stones.\n\nProcedures to remove urinary stones must be accompanied by JJ stent insertion. Insertion of a JJ stent is a routine procedure to ensure that the flow of urine from the kidneys can get to the bladder, preventing extravasation in the area of operation during urinary stone removal procedures. Although the usage is safe, JJ stents can cause quality of life disorders by causing pain, discomfort, urinary tract infection, haematuria, irritation, and reduction in sexual function, which in the long term can cause psychological disorders such as insomnia, depression, and anxiety3–7. In a study conducted by Giannarini et al. there was a strong association between the location of stent distal loop with respect to midline to disturbance of sexual function after stent placement8.\n\nSymptoms caused by the insertion of JJ stents may be managed with administration of alpha blockers, anticholinergics, or a combination of both9–13. In a previous study, the drug tadalafil is said to be better for treatment than tamsulosin, as it has an advantage in managing sexual complaints14. The use of pregabalin has also been studied to manage post-stent insertion complaints, but it does not provide an improvement to the sexual disorders that occur15.\n\nThere have been many studies that examine JJ stent insertion association with irritating and painful complaints, but studies that evaluate the correlation with sexual disorders are remain few. Therefore, the present study sought to investigate the association of JJ stent insertion with sexual function in men and women.\n\n\nMethods\n\nThis was a cohort study in which the subjects were all patients who received JJ stent installation at Kardinah General Hospital, Jakarta between July and November 2017.\n\nThis study was approved by the Research Ethic Committee of Kardinah Hospital (#445/3840/2017). The study was conducted in accordance with the principles of the Declaration of Helsinki, and all patients provided written informed consent to participate.\n\nParticipants were recruited during consultation about insertion of JJ stent.\n\nThe inclusion criteria in the study were men and women >18 years old, who were candidates for JJ stent insertion procedure and had regular intercourse over the last 4 weeks before interviews. Exclusion criteria in this study were patients with prior history of sexual disorders, pelvic surgery, pelvic radiotherapy, renal failure, and neurogenic bladder.\n\nData were taken one day before JJ stent insertion procedure using standardized self-administered questionnaire and the questionnaire was repeated perioperatively before JJ stent removal procedure. Sexual function was assessed using the International Index of Erectile Function (IIEF) questionnaire for men and the Female Sexual Function Index (FSFI) for women. Higher questionnaire scores indicate better sexual function. Each question describes a specific domain and each domain has a \"factor\" for counting the number of scores. In IIEF, the score may vary from 5–75. Erectile function domain score interpretation are distinguished into 5 categories: severe erectile dysfunction (1–10); moderate erectile dysfunction (11–16); mild-to-moderate dysfunction (17–21); mild dysfunction (22–25); and no dysfunction (26–30). While, for the other 4 domains, a higher score indicates less dysfunction. Meanwhile in FSFI, the score may vary from 2–36, with a normal score ≥ 26. FSFI scores < 26 are assumed to indicate sexual dysfunction16.\n\nStatistical analysis was performed using SPSS 22. Data analysis was done descriptively and quantitatively. Descriptive data: age, operating procedure, and duration of JJ stent. Quantitative data: IIEF and FSFI scores at the time before installation and after insertion. Statistical analysis was performed using a paired T test and Wilcoxon signed-rank test.\n\n\nResults\n\nA total of 93 subjects from July to November 2017 were included in this study; 60 men and 33 women. The average age of men was 51.1 ± 10.6 years and women 49.6 ± 10.6 years. In both men and women the most frequent stone location was a proximal ureteral stone, with a higher stone burden seen in women. A total of 6.7% of the men underwent bilateral JJ stent insertion, while no women underwent this procedure. The length of stenting between groups of men and women was not different. Demographic data of participants can be seen in Table 1.\n\nTable 2 shows the comparison of IIEF scores in the male group. There was a significant difference (p <0.05) in the IIEF total score results before (49.55 ± 2.3) and after (38.4 ± 1.7) JJ stent insertion. This is also seen in the scores of each subdomain of the IIEF questionnaire, where in the five subdomains showed a significant difference before and after stent placement.\n\n*Wilcoxon signed-rank test; *T paired test\n\nTable 3 shows the comparison of FSFI score in women. There was a significant difference (p <0.05) in the total FSFI score before (23.62 ± 0.64) and after (16.7 ± 0.52) the insertion of the JJ stent. This is also seen in the scores of each subdomain, where each of the six subdomains showed a significant difference.\n\n*Wilcoxon signed-rank test; *T paired test\n\n\nDiscussion\n\nUrinary tract stones are rare in urology. The prevalence of urinary tract stones in general is 5–12% in men and 4–7% in women worldwide1. The upper urinary tract procedure is inseparable from the use of JJ stents. URS, PCNL, laparoscopy or open ureterolithotomy surgery are all procedures that often uses the JJ stent. Different complications can occur with the use of JJ stents, such as haematuria, dysuria, frequency, flank and suprapubic pain, to major complications, such as vesico-ureteric reflux, migration, malposition, encrustation, stent fracture, UTI, pyuria, incontinence, inadequate relief of obstruction, ureteric erosion or fistulation, a ‘forgotten stent’, necrosis and uretero-arterial fistula17. In addition, it should also be considered as having a possible effect on sexual function. Sexual dysfunction can cause physical and psychological health disorders that will interfere with quality of life.\n\nThe aim of this study was to assess the effect of stent insertion on sexual function. The study used the IIEF and FSFI questionnaires to be able to assess more specifically disturbed sexual function. The IIEF score in men and FSFI score in women decreased significantly in post-JJ stent insertion. This shows that sexual function has interference after the installation of the JJ stent in both men and women. However, a clear mechanism for how JJ stent may interfere with sexual function remains unclear.\n\nThe results in this study were not much different when compared with other studies. Bolat et al. made a study comparing sexual function using IIEF in patients undergoing URS procedures1. Follow-ups were made at one month and three months’ post-procedure. The study found no correlation between URS and sexual function disorder. Therefore, the sexual function disorder that occured may be caused by the use of the JJ stent. In the study of Bolat et al. sexual dysfunction began in the first month of post-installation and improved 3 months’ post-op.\n\nAkdeniz and Bolat performed a study comparing the FSFI score in patients undergoing URS procedures with JJ stent2. Follow -p was done at one month and three months’ post procedure with the average of long time of installation of JJ stent 15.7 ± 2.4 days. The mean FSFI score in the that study was perioperative 14.5 ± 9.6, one-month post-op 12.8 ± 6.8, and post-operative 17.7 + 5.4. FSFI score worsened in the first month and improved after the third month.\n\nEryildirim et al. did a study comparing the IIEF score in men and FSFI in women undergoing URS procedures with a JJ stent3. In the study, the assessment was done before and one-month post-JJ stent installation. There was a significant decrease in both IIEF and FSFI scores before and after the installation of JJ stents. Similar results were also obtained by Sighinolfi et al. but with different assessment times, i.e. before and 45–60 days’ post-installation4.\n\n\nConclusion\n\nThe installation of the JJ stent is a common practice in urological procedures. However, the installation of the JJ stent may cause sexual function impairment. Patients undergoing JJ stent installation procedures should be advised of this as it may impair their quality of life. The use of JJ stents needs to be reviewed, and if it is necessary to use the JJ stent, the length of installation should be as short as possible.\n\n\nData availability\n\nF1000Research: Dataset 1. Answers pre and perioperatively to the International Index of Erectile Function (IIEF) and Female Sexual Function Index (FSFI) questionnaires for men and women, respectively, who underwent JJ stent insertion., https://doi.org/10.5256/f1000research.16608.d22725418.",
"appendix": "Grant information\n\nThe author(s) declared that no grants were involved in supporting this work.\n\n\nReferences\n\nBolat MS, Akdeniz E, Asci R, et al.: Ureterorenoscopy with stenting and its effect on male sexual function: A controlled randomised prospective study. Andrologia. 2017; 49(9): e12746. PubMed Abstract | Publisher Full Text\n\nAkdeniz E, Bolat MS: Ureterorenoscopy with Stenting and Its Effect on Female Sexual Function. Urol J. 2017; 14(3): 3059–3063. PubMed Abstract | Publisher Full Text\n\nEryildirim B, Tuncer M, Kuyumcuoglu U, et al.: Do ureteral catheterisation procedures affect sexual functions? A controlled prospective study. Andrologia. 2012; 44 Suppl 1: 419–423. PubMed Abstract | Publisher Full Text\n\nSighinolfi MC, Micali S, De Stefani SD, et al.: Indwelling ureteral stents and sexual health: a prospective, multivariate analysis. J Urol. 2007; 178(1): 229–231. PubMed Abstract | Publisher Full Text\n\nJoshi HB, Stainthorpe A, Keely FX Jr, et al.: Indwelling ureteral stents: evaluation of quality of life to aid outcome analysis. J Endourol. 2001; 15(2): 151–154. PubMed Abstract | Publisher Full Text\n\nJoshi HB, Stainthorpe A, MacDonagh RP, et al.: Indwelling ureteral stents: evaluation of symptoms, quality of life and utility. J Urol. 2003; 169(3): 1065–1069; discussion 1069. PubMed Abstract | Publisher Full Text\n\nLeibovici D, Cooper A, Lindner A, et al.: Ureteral stents: morbidity and impact on quality of life. Isr Med Assoc J. 2005; 7(8): 491–494. PubMed Abstract\n\nGiannarini G, Keeley FX Jr, Valent F, et al.: Predictors of morbidity in patients with indwelling ureteric stents: results of a prospective study using the validated Ureteric Stent Symptoms Questionnaire. BJU Int. 2011; 107(4): 648–654. PubMed Abstract | Publisher Full Text\n\nEl-Nahas AR, Tharwat M, Elsaadany M, et al.: A randomized controlled trial comparing alpha blocker (tamsulosin) and anticholinergic (solifenacin) in treatment of ureteral stent-related symptoms. World J Urol. 2016; 34(7): 963–8. PubMed Abstract | Publisher Full Text\n\nMaldonado-Avila M, Garduno-Arteaga L, Jungfermann-Guzman R, et al.: Efficacy of Tamsulosin, Oxybutynin, and their combination in the control of double-j stent-related lower urinary tract symptoms. Int Braz J Urol. 2016; 42(3): 487–93. PubMed Abstract | Publisher Full Text | Free Full Text\n\nHe F, Man LB, Li GZ, et al.: Efficacy of α-blocker in improving ureteral stent-related symptoms: a meta-analysis of both direct and indirect comparison. Drug Des Devel Ther. 2016; 10: 1783–1793. PubMed Abstract | Publisher Full Text | Free Full Text\n\nYakoubi R, Lemdani M, Monga M, et al.: Is there a role for α-blockers in ureteral stent related symptoms? A systematic review and meta-analysis. J Urol. 2011; 186(3): 928–934. PubMed Abstract | Publisher Full Text\n\nLamb AD, Vowler SL, Johnston R, et al.: Meta-analysis showing the beneficial effect of α-blockers on ureteric stent discomfort. BJU Int. 2011; 108(11): 1894–1902. PubMed Abstract | Publisher Full Text\n\nAggarwal SP, Priyadarshi S, Tomar V, et al.: A Randomized Controlled Trial to Compare the Safety and Efficacy of Tadalafil and Tamsulosin in Relieving Double J Stent Related Symptoms. Adv Urol. 2015; 2015: 592175. PubMed Abstract | Publisher Full Text | Free Full Text\n\nRagab M, Soliman MG, Tawfik A, et al.: The role of pregabalin in relieving ureteral stent-related symptoms: a randomized controlled clinical trial. Int Urol Nephrol. 2017; 49(6): 961–966. PubMed Abstract | Publisher Full Text\n\nRosen R, Brown C, Heiman J, et al.: The Female Sexual Function Index (FSFI): a multidimensional self-report instrument for the assessment of female sexual function. J Sex Marital Ther. 2000; 26(2): 191–208. PubMed Abstract | Publisher Full Text\n\nAl-Marhoon MS, Shareef O, Venkiteswaran KP: Complications and outcomes of JJ stenting of the ureter in urological practice: A single-centre experience. Arab J Urol. 2012; 10(4): 372–7. PubMed Abstract | Publisher Full Text | Free Full Text\n\nFawzi R, Ali Z: Dataset 1 in: Association of JJ stent insertion and sexual function: A cohort study. F1000Research. 2018. http://www.doi.org/10.5256/f1000research.16608.d227254"
}
|
[
{
"id": "43820",
"date": "04 Feb 2019",
"name": "Ekrem Akdeniz",
"expertise": [
"Reviewer Expertise Urologist"
],
"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 prospective study about effect on sexual functions of JJ stent. The manuscript presents basic results. This was a useful and interesting article. Your manuscript is clear and understandable without discussion section. A few issues, however, need to be addressed:\nAbstract\nBackground section: Please erase this sentence: \"Indonesia is known as one of the world’s stone belt areas in Asia.\" Method section: Please erase this sentence: \"This study was approved by the Research EthicCommittee of Kardinah Hospital (#445/3840/2017).\" Result section: Please, define it. Are there any differences of sexual functions male patients before and after procedure? Keywords: Complete change without sexual function.\n\nIntroduction Generally well written.\nFirst and second paragraph: First reference, Bolat et al. made a study comparing sexual function about URS procedures in males1. But you refer to this article about the prevalence of erectile dysfunction (ED) and cause of ED. Bolat et al. did not study prevalence of ED. So you have to change this reference.\n\nThird paragraph: Second reference, Akdeniz and Bolat worked on female sexual functions in their article2. But you refer to this article about the prevalence of urolithiasis. Akdeniz did not study prevalence of urolithiasis. You cannot use these articles about prevalence of urolithiasis and ED. So you have to change this reference too.\nThird paragraph: Please add shock wave lithotripsy, and reference this sentence: \"Percutaneous nephrolithotomy (PCNL), ureteroscopy (URS), laparoscopy or open ureter are the main procedures of the kidney and ureter for the removal of stones.\"\n\nFourth paragraph: You wrote: \"Procedures to remove urinary stones must be accompanied by JJ stent insertion.\" Why? Routine stenting after uncomplicated URS is not necessary (EAU guideline 2018). How did you come to this opinion? What is your reference for this claim?\n\nAn introduction section should include precise, accurate and clear information. Please change the first and second references. You should choose suitable new references about prevalence of urolithiasis and ED.\nMethods Generally well written.\nWhat was the JJ stent insertion reason? Only urolithiasis? Please define it. When was the JJ stent removed for each operations? The authors did not mention about surgical procedure especially URS.\nResults Generally well written.\nWhat was the educational status of the patients? Were patients predominantly urban based or rural based? What was the body mass index of the patients? Were patients evaluated for depression? (For example, Beck depression scale). Did the patients get medication for stent complications (urgency, dysuria, etc)? (Alpha blockers, anticholinergics, combined, etc..). Have patients developed any complications? When was the JJ stent withdrawn for each procedure? Was the same JJ stent used in the patients? URS is more painless and more effortless procedure than PCNL, pyeloilithotomy and ureterolithotomy. Pain is the cause of sexual dysfunction. So these operations cannot be evaluated at the same level. In my opinion, the authors should evaluate only same operations not all the stone surgery (only URS or only PCNL or only open surgery patients).\nDiscussion Generally well written.\nMore details about similar studies should be provided. Please clearly explain, what is the difference between this study and others?\nConclusion Please clearly explain what are the strengths and weaknesses of your manuscript?\nTables Generally well written.\nTable 1. Written PCNL and URS. It is necessary to write the abbreviation explicitly under the table. Table 1. Authors wrote stone burden were 1.48 ± 1.58 mm2 (men) and 3.7 ± 5.7 mm2 (women). Is there a mistake here? I think, operation of such small stones is unnecessary. Table 3. Written FSFI. It is necessary to write the abbreviation explicitly under the table.\nStrengths of the manuscript:\n1. Article is appropriately organized and the headings are indicative of content. 2. Article clearly written. Simple, plain and legible. 3. References are satisfactory.\nWeaknesses of the manuscript:\n1. This article not contain new ideas or useful synthesis of existing material. 2. The subject matter is suitable for the intended audience but the results need to be developed. 3. The article does not include a single operation.\nI regret to say that the article needs to be improved. The amount of patients data is very small, and not satisfactory. The data (especially demographic and patients) should be developed (addition new data), give more details about surgical procedure, results need to be improved, and discussion should give more details about similar studies. Briefly, the manuscript should be re-written. The manuscript should be revised by authors and reconsidered for possible indexing.\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?\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/7-1978
|
https://f1000research.com/articles/7-1975/v1
|
24 Dec 18
|
{
"type": "Method Article",
"title": "Rapid identification of novel protein families using similarity searches",
"authors": [
"Matt Jeffryes",
"Alex Bateman",
"Matt Jeffryes"
],
"abstract": "Protein family databases are an important tool for biologists trying to dissect the function of proteins. Comparing potential new families to the thousands of existing entries is an important task when operating a protein family database. This comparison helps to understand whether a collection of protein regions forms a novel family or has overlaps with existing families of proteins. In this paper, we describe a method for performing this analysis with an adjustable level of accuracy, depending on the desired speed, enabling interactive comparisons. This method is based upon the MinHash algorithm, which we have further extended to calculate the Jaccard containment rather than the Jaccard index of the original MinHash technique. Testing this method with the Pfam protein family database, we are able to compare potential new families to the over 17,000 existing families in Pfam in less than a second, with little loss in accuracy.",
"keywords": [
"bioinformatics",
"protein families",
"locality sensitive hashing",
"minhash"
],
"content": "Introduction\n\nProtein family databases are an important resource for biologists seeking to characterise the function of proteins. The domains, motifs and other features found in a protein form an important organisational structure that can be used to design and interpret experiments on the protein of interest. Operating a protein family database, such as Pfam or InterPro, requires the identification of new families, and the ability to compare them with families already in the database. In this paper, we will describe a computationally efficient method for performing this comparison.\n\nProtein family databases generally describe a particular family using a sequence profile, often in the form of a hidden Markov model (HMM)1. The profile HMM is a representation of the multiple sequence alignment of a number of representatives of a family. The likelihood that a given sequence is a member of a family (that is, it has homology with the other members of the family) is thus estimated by the probability of its alignment to this profile HMM.\n\nA protein family database should cover as much of sequence space as possible, while reducing overlapping sequence profiles. This is illustrated by the idealised view of sequence space shown in Figure 1.\n\nNo sequence is contained in more than one family.\n\nAn overlap occurs when a particular region in a protein sequence is a significant match for more than one sequence profile. In this case there are two possibilities. Either the region of the protein sequence in which the overlapping matches lie is a false positive for one or both of sequence profiles, or the sequence profiles represent families which are homologous, and are in fact a single family. Maximising coverage of sequence space increases the chance of overlap. Each sequence profile added to increase coverage may overlap with the existing profile HMMs in the database. Such overlaps are a result of the fact that HMM sequence profiles are an imperfect model of the underlying homology of the families that they represent.\n\nPfam is a database of protein families first released in 1997 with models for 175 families. The current release (32.0) contains 17,929 entries2,3. Each of these entries is described by a seed alignment, which is used to generate a profile HMM, using the HMMER software package. This model is then used to query the protein sequence database UniProtKB, and significant matches for the model are recorded. In Pfam, no region should be matched by more than one model4. Prior to each release of the database, overlap analysis is performed to determine if any residue in UniProtKB is matched by more than one model. This overlap criteria is an important quality control mechanism2,4. The most recent releases of Pfam have relaxed the overlap criteria slightly, to allow short areas of overlap which do not affect a high proportion of the family members3.\n\nWhen the authors of Pfam check for overlaps prior to a release, they carry out a complete comparison of all the regions found in every family. This comparison scales with the square of the number of regions in the database. This full comparison is only carried out once per release, and therefore speed is not a significant issue. However, we envision a use case where speed of this overlap test is critical. This use case is the identification of novel families from user sequence similarity searches.\n\nProtein sequence similarity search is used to identify proteins with a similar sequence to a query protein. When two proteins have a similar sequence, it may be inferred that they are evolutionarily related. Thus, the results for a sequence similarity search form a set of potential homologues to a query sequence, which may be thought of as a protein family. The critical question is whether this protein family is novel. If overlaps could be checked quickly enough then the user could be alerted to the fact that the family was novel and be prompted to submit it to Pfam. This overlap check would need to be fast, on the order of a second so that it presented as an interactive element of the search result.\n\nIn addition to sequence similarity searches, which find entirely novel groupings of proteins, it can occur that a search matches all members of an existing family along with further proteins which are as yet unclassified in the protein family database. If these proteins are truly homologous with the existing members of the family, then they ought to be members of the family. Therefore, the search may encode a superior model for the existing family.\n\nBy analysing the search for overlaps with the families in Pfam, using the method discussed in this paper, searches which could improve Pfam can quickly be identified. This could help curators more rapidly identify novel families, but also open the way for identification of novel user inferred families from sequence search submission.\n\n\nMethods\n\nSupposing that we wish to assess a potential new family, for addition to Pfam, we can imagine three scenarios:\n\n1. The search alignment does not overlap with any existing families. It is entirely new residue coverage.\n\n2. The search alignment overlaps with only one existing family.\n\n3. The search alignment overlaps with multiple existing families.\n\nThe first case is the clearest case where we may wish to add the family to Pfam. But in the other two cases we may also wish to. In the second case, the search may offer increased residue coverage compared to the existing family. That is, it identifies more members of the family. In the third case, if the multiple families are all of the same Pfam clan, the search may either be a superior model for an existing clan member, or it could be a novel member of the clan. In the second and third case, we require some way of relating an arbitrary search to the families which already exist in Pfam.\n\nThe Jaccard index of a pair of sets is a measure of their similarity. It is calculated as follows.\n\n\n\nThis is the fraction of all members of A and B which are found in both A and B. Since we are interested in finding search result sets which are supersets of Pfam families, we might also ask what fraction of A is found in both A and B. That is, how close is A to being a subset of B. This measure is known as containment. We can calculate theb Jaccard containment as follows5,6.\n\n\n\nTo calculate either of these statistics requires the use of set intersection between A and B, which is a computationally relatively expensive operation. Assuming A and B are stored in hash maps, the average case time complexity is O(min(|A|, |B|))7.\n\nLocality sensitive hashing is a technique for quickly identifying similar sets, faster than is possible by calculating the set intersection for each particular pair. Min-wise independent permutations or MinHash is a locality sensitive hash algorithm which estimates the Jaccard index for a pair of sets. Specifically, it estimates the set intersection and union. This additionally allows us to estimate the Jaccard containment. It was was introduced by Broder6, with the original application being the elimination of identical web pages from the index of the Alta Vista search engine.\n\nThe algorithm has been applied to a number of biological problems, such as metagenomic clustering, genome assembly, and sequence database search8–12. MinHash and the Jaccard index and containment for sets A and B are estimated as follows.\n\nFor set S, define MINn(S) as\n\n\n\nLet h(x) be some hash function. Define set A′ as\n\n\n\nand B′ analogously.\n\nWe can then see that\n\n\n\nis a sample of at most n elements from A′ ∪ B′, and that\n\n\n\nis a sample of at most n elements from A′ ∩ B′ which are also contained in the sample from A′ ∪ B′. We have ensured that these samples are random by hashing the elements of A and B. Hence,\n\n\n\n\n\nis an estimate of the Jaccard index. To compute this estimate, only the smallest n hashed elements of the sets to be compared is required.\n\nWe can take a similar approach for estimating the Jaccard containment. We can find\n\n\n\nas elements from B′ which are also contained in the sample from A′ and hence\n\n\n\nis an estimate of the Jaccard containment.\n\nFor Pfam, we wish to determine whether a user’s search overlaps with an existing family. This comparison is on the basis of residues. Hence, the elements of the sets to be compared can be represented uniquely as a combination of a protein’s identifier and residue position within the protein sequence. We can compute the hashes for every family in Pfam. When a user search is performed, we can compute its hash, and estimate whether it falls into one of the desirable categories above (overlapping no families, or covering a single family or clan, with increased residue coverage).\n\nIn order to use this method interactively, the time taken to calculate the hash for the user’s search must be taken into account. The hashes for existing Pfam families need only be calculated once, ahead of time, but the hash for a user’s search must be calculated while they are waiting for their search results. For a user search with a large number of results, the number of unique residues within protein sequences which the search aligns to could be in the tens of millions. Each of these residues is an element of the set which must be hashed and sorted in order to produce the hash for the search. The number of set elements can be reduced by sacrificing accuracy. Residues can be grouped together in arbitrary sized chunks. As the size of these chunks increases, the number of elements is reduced, but the risk increases that a search which does not overlap with a Pfam family is misidentified as overlapping with the family. The chunk which a residue with coordinate i should be assigned to is computed as ⌊i/w⌋, where w is the window size of each chunk.\n\n\nResults\n\nWe implemented MinHash in the Python programming language to validate its theoretical gain in performance over exact calculations. We generated hashes of every family in Pfam 29.03. We chose 50 random families from Pfam, and for each of these we timed the calculation of the Jaccard index between the family and every family in Pfam, and the MinHash estimate for the Jaccard index with n values of 25, 50, 100, and 200. For each method, the calculation was repeated three times, and the minimum of the three used. The results are shown in Figure 2. For any family size, MinHash is faster. Also clear is the linear relationship between family size and calculation time for the Jaccard index. In Figure 3 the linear relationship between n and calculation time for MinHash is shown, and so is the constant time to estimate the Jaccard index as family size varies. In Figure 4, calculation of the Jaccard containment grows with log(n). However, for the family sizes tested, calculating the Jaccard containment was faster than the Jaccard index. This is due to the sort operation required to estimate the Jaccard index.\n\nA linear least squares best fit line for the two methods is shown.\n\nThe time taken for the Jaccard index between the family and the rest of the families to be estimated is shown. For cases where the family size is less than n, the results were excluded from the plot. (a) Time taken in seconds to estimate the Jaccard index using MinHash with n values of 25, 50, 100, and 200, with w value of 1 (that is, no chunking of residues). A linear least squares best fit is shown. Note that the data points are jittered on the x axis to better show their distribution. (b) Time taken in seconds to estimate Jaccard index with n values of 25, 50, 100, and 200, against family size on a logarithmic scale. A linear least squares best fit for each n is shown.\n\nThe time taken for the Jaccard index between the family and the rest of the families to be estimated is shown. For cases where the family size is less than n, the results were excluded from the plot. (a) Time taken in seconds to estimate the Jaccard containment using MinHash with n values of 25, 50, 100, and 200, with w value of 1 (that is, no chunking of residues). A linear least squares best fit is shown. Note that the data points are jittered on the x axis to better show their distribution. (b) Time taken in seconds to estimate Jaccard containment with n values of 25, 50, 100, and 200, against family size on a logarithmic scale. A linear least squares best fit for each n is shown.\n\nCalculation time of the order of seconds, even with high values of n, will enable fast estimation of the relationship between a potential new family and the rest of Pfam. In Figure 5, the concordance between Jaccard index and containment and their MinHash estimates, between the same sample as above and the rest of Pfam are shown. Even with n = 25 the discrepancy is not great. Thus, searches which may not overlap any existing Pfam family, and families which may be improvements over existing families can be identified in less than a second.\n\nThe diagonal lines show the position that a perfect estimate would fall. (a) Jaccard index concordance. (b) Jaccard containment concordance.\n\nIncreasing the value of w also reduces accuracy, but reduces the time to compute the hash of the potential new family. In Figure 6, the time taken to compute hashes for different values of w is shown. With a w value of 1 (that is, without chunking residues), it takes over 10 seconds to compute the hash of the largest family. Increasing w enables this time to be reduced to under a second. For a production system, regular waits of over 10 seconds would be unacceptable, so w should be set to at least 4. On the other hand, high values of w will result in more frequent errors in multidomain proteins: In cases where the domains have fewer residues separating them than w + 1, there is the possibility that the profiles for the two domains could be wrongly identified as overlapping. Therefore, w of greater than 16 could be detrimental.\n\n\nConclusions\n\nWe have developed a method for quickly comparing the search results produced by querying a profile HMM against a protein sequence database, to a protein family database. We have adapted a method for estimating the Jaccard index of a pair of sets to calculate the Jaccard containment. This allows the rapid evaluation of the relationship between a pair of multiple sequence alignments. That is, does one alignment contain a superset or subset of the regions in the other.\n\nThis method is intended to enable an automated quality control for protein family profile HMMs. The MinHash derived comparison method for protein families is a critical component of an automated pipeline for identifying families from sequence similarity search which are candidates for integration with Pfam. This method can be adjusted to meet the required speed for an interactive protein sequence similarity search by slightly reducing the accuracy of the estimate.\n\nWe foresee multiple applications for these methods. They could be used to filter user submitted family profile HMMs, enabling crowdsourcing of the Pfam database. We also see the hash-based set relationship comparison methods as useful not just for protein families, but for other types of sequence data, such as RNA families. In addition, the calculation of Jaccard containment can be used in the hierarchical classifications such as InterPro or SUPERFAMILY to help identify subfamily relationship13,14.\n\n\nData availability\n\nAn implementation of the method discussed in this paper is available in the Search-Sifter package. DOI: http://doi.org/10.5281/zenodo.156065915.\n\n\nSoftware availability\n\nThe Search-Sifter package is available at: https://github.com/bateman-research/search-sifter.\n\nArchived source code at time of publication: http://doi.org/10.5281/zenodo.156065915.\n\nLicense: MIT License.",
"appendix": "Grant information\n\nThis work was funded by the European Molecular Biology Laboratory.\n\nThe funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.\n\n\nReferences\n\nEddy SR: Profile hidden Markov models. Bioinformatics. 1998; 14(9): 755–763. PubMed Abstract | Publisher Full Text\n\nSonnhammer EL, Eddy SR, Durbin R: Pfam: a comprehensive database of protein domain families based on seed alignments. Proteins. 1997; 28(3): 405–420. PubMed Abstract | Publisher Full Text\n\nFinn RD, Coggill P, Eberhardt RY, et al.: The Pfam protein families database: towards a more sustainable future. Nucleic Acids Res. 2016; 44(D1): D279–D285. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBateman A, Birney E, Durbin R, et al.: Pfam 3.1: 1313 multiple alignments and profile HMMs match the majority of proteins. Nucleic Acids Res. 1999; 27(1): 260–262. PubMed Abstract | Publisher Full Text | Free Full Text\n\nAgrawal P, Arasu A, Kaushik R: On indexing error-tolerant set containment. In Proceedings of the 2010 international conference on management of data. New York, USA; ACM Press. 2010; 927. Publisher Full Text\n\nBroder AZ: On the resemblance and containment of documents. In Compression and complexity of sequences. Selerno, IT, IEEE. 1997; 21–29. Reference Source\n\nRivest RL, Stein C, Cormen TH, et al.: Introduction to Algorithms. 3rd Edition. MIT Press, 2009. Reference Source\n\nOndov BD, Treangen TJ, Melsted P, et al.: Mash: fast genome and metagenome distance estimation using MinHash. Genome Biol. 2016; 17(1): 132. PubMed Abstract | Publisher Full Text | Free Full Text\n\nBerlin K, Koren S, Chin CS, et al.: Assembling large genomes with single-molecule sequencing and locality-sensitive hashing. Nat Biotechnol. 2015; 33(6): 623–630. PubMed Abstract | Publisher Full Text\n\nRasheed Z, Rangwala H: MC-MinH: Metagenome Clustering using Minwise based Hashing. In Proceedings of the 2013 SIAM International Conference on Data Mining. Society for Industrial and Applied Mathematics, Philadelphia, PA. 2013; 677–685. Publisher Full Text\n\nPopic V, Kuleshov V, Snyder M, et al.: Fast Metagenomic Binning via Hashing and Bayesian Clustering. J Comput Biol. 2018; 25(7): 677–688. PubMed Abstract | Publisher Full Text\n\nKoslicki D, Zabeti H: Improving Min Hash via the Containment Index with applications to Metagenomic Analysis. bioRxiv. 2017. Publisher Full Text\n\nPandurangan AP, Stahlhacke J, Oates ME, et al.: The SUPERFAMILY 2.0 database: a significant proteome update and a new webserver. Nucleic Acids Res. 2018; gky1130. PubMed Abstract | Publisher Full Text\n\nMitchell AL, Attwood TK, Babbitt PC, et al.: InterPro in 2019: improving coverage, classification and access to protein sequence annotations. Nucleic Acids Res. 2018; gky1100. PubMed Abstract | Publisher Full Text\n\nJeffryes M: bateman-research/search-sifter: Initial public release (version v1.0). Zenodo. 2018. http://www.doi.org/10.5281/zenodo.1560659"
}
|
[
{
"id": "44111",
"date": "11 Feb 2019",
"name": "Daniel J. Rigden",
"expertise": [
"Reviewer Expertise protein 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\nThis concise paper nicely describes a problem - how to rapidly compare the results of a database search to a set of protein families - and provides a solution via an efficient hashing algorithm. I have just a few suggestions.\nTo me a title like \"Rapid identification of novel protein families in the results of similarity searches\" would better convey the content. As is, it reads as if the paper is more about innovations in the similarity searches themselves. The authors should cite UniProtKB in the Intro. Typo 'theb Jaccard containment'. Maybe a brief explanation of what hashing is - the paper rather assumes the reader knows what it means. Are the funny partial square brackets round i/w on p.5 correct? Maybe 'In Figure 4, the time taken for calculation of the...' Maybe on pg. 5 spell out what the consequences of under- or over-prediction of the two Jaccard measurements would be for placing database search results into the three categories defined at the star of Methods. On p. 5 the authors say that 'Increasing the value of w also reduces accuracy' Visually, that seems clear for n=64 but the differences for n=1, 4 or 16 are much less obvious. Maybe calculate Rsquared or some other statistical measure to quantify this tendency.\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": "46291",
"date": "28 Mar 2019",
"name": "Desmond G Higgins",
"expertise": [
"Reviewer Expertise bioinformatics",
"sequence alignment methods."
],
"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 describes a method for very quickly comparing lists of sequence parts (e.g. from a similarity search) to PFAM families. It is extremely well described. I have a few typos and queries but these are very minor. The method looks useful and, as the authors point out, can be used as the basis for a variety of production or interactive tools.\nI have no major changes to request.\nMinor points:\npage 3. The word \"regions\" is mentioned twice. What is a region? page 4 \"But in the other 2 cases we may also wish to\". Wish to what? Also \"We can calculate theb\" page 5. The minimum of the three is used. Minimum time? \"chunking\"? Is this a word? page 3 \"This overlap criterION is an important ...\" page 6 \"...identifying families from A sequence similarity search.\"\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": []
}
] | 1
|
https://f1000research.com/articles/7-1975
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.