title stringlengths 1 1.19k | keywords stringlengths 0 668 | concept stringlengths 0 909 | paragraph stringlengths 0 61.8k | PMID stringlengths 10 11 |
|---|---|---|---|---|
ACKNOWLEDGMENTS | This article was extracted from a Master's thesis in Midwifery. Hereby, we would like to thank the Research Administration of Mazandaran University of Medical Sciences, Sari, Iran, which supported this project [Grant number: 9188]. | PMC10009422 | ||
DATA AVAILABILITY STATEMENT | Since our data contain sensitive personal information, it is forbidden to share these data with a third party without obtaining an additional written form of informed consent for information sharing. We did not obtain additional written consent for information sharing. | PMC10009422 | ||
REFERENCES | PMC10009422 | |||
Background | It is believed that negative postoperative behavioral changes (NPOBC) is associated with negative perioperative outcomes in children. The importance of development of a predictive model of NPOBC was noted. This study aims to identify potential risk factors develop a nomogram to predict NPOBC on postoperative day 3 based on a prospective cohort. | PMC10401797 | ||
Methods | REGRESSION | A prospective observational study was conducted on children(American Society of Anesthesiologists I ~ III) aged 2 ~ 12 years who underwent selective surgery under general anesthesia between September 2022 and February 2023. The patient’s clinical data were analyzed. The method of measuring NPOBC is with the The Posthospital Behaviour Questionnaire (PHBQ), and all of children remained hospitalized at the time of assessment. The enrolled patients were categorized into the NPOBC group and the non-NPOBC group according to if children developed NPOBC on postoperative day 3. Univariate and multivariate logistic regression analyses were performed to identify independent risk factors and develop the nomogram to predict NPOBC. Internal validation was performed using the parametric bootstrapping method. | PMC10401797 | |
Results | postoperative pain, cerebral desaturation | REGRESSION | One hundred ninety-two patients were enrolled in the study, 44.8% (86/192 patients) of children developed NPOBC on postoperative day 3. Univariate and multivariate logistic regression analysis demonstrated that the Pediatric Anesthesia Behavior (PAB) score (OR: 1.23, 95%CI: 1.14–1.33), cerebral desaturation (OR: 1.16, 95%CI: 1.02–1.32), and postoperative pain score (OR: 1.07, 95%CI: 1.02–1.13) were independent predictors for NPOBC on postoperative day 3 ( | PMC10401797 |
Conclusion | postoperative pain, cerebral desaturation | Based on our prospective observational study, pre-anesthesia patients with higher PAB scores, presence of cerebral desaturation, and higher postoperative pain score were more likely to develop NPOBC on postoperative day 3. We established and validated a nomogram for predicting NPOBC, which could help assess patients individually, identify high-risk groups of NPOBC and improve patient prognosis. | PMC10401797 | |
Trial registration | MAY | Chinese Clinical Trial Registry, ChiCTR‐2,200,059,776. Registered 11 May 2022. | PMC10401797 | |
Supplementary Information | The online version contains supplementary material available at 10.1186/s12871-023-02228-4. | PMC10401797 | ||
Keywords | PMC10401797 | |||
Introduction | ADVERSE EFFECTS | Based on the principle of personalized medicine, protecting children’s perioperative mental health is gradually being emphasized [In the present study, we prospectively collected patients’ data, analyzed the characteristics of patients who developed NPOBC on postoperative day 3, sought independent risk factors, and developed a predictive model. We expect the nomogram can help clinicians assess patients individually, identify high-risk patients of NPOBC, give early intervention to decrease adverse effects, and improve patient prognosis. | PMC10401797 | |
Methods | PMC10401797 | |||
Participants | neuropsychiatric disorders, skin lesions, rash | EMERGENCY, SKIN LESIONS | This prospective observational study was conducted at the National Center for Children’s Health, Beijing Children’s Hospital, China, over 6 months between September 2022 and February 2023. The study was registered at the Chinese clinical trial registry (ChiCTR‐2,200,059,776). The Ethics Committee of Beijing Children's Hospital, Capital Medical University has approved the use of fully anonymized cohort data for research (ID: 2021-E-114-Y). All children's guardians provided written informed consent.Patients between the ages of 2 and 12 years, with American Society of Anesthesiologists (ASA) physical status I to III and scheduled for elective noncardiac surgery under general anesthesia for more than 60 min, with mechanical respiratory assistance, were included in the study. The exclusion criteria were: a. Children with previous neuropsychiatric disorders; b. Emergency surgery; c. Monitoring site with skin lesions or rash; d. Children who need to be transferred to Intensive Care Unit (ICU) for further treatment after surgery, e. Refusal by the parents to participate. | PMC10401797 |
Assessment | agitation, adverse behavior, nausea, pain | The primary outcome was the presence of NPOBC on postoperative day 3. NPOBC was assessed by the PHBQ (Supplemental Table Parents completed the PHBQ on postoperative day 3 in order to avoid the influence of first two days agitation, significant pain or nausea on questionnaire response. In addition, all of children remained hospitalized at the time of assessment, and the doctor visits the ward for postoperative follow-up, which can explain to parents how to fill out the questionnaire and discover adverse behavior in hospitalized children. | PMC10401797 | |
Anesthesia management | anxiety | All patients will be hospitalized the day before surgery and undergo routine preoperative examinations. One day preoperatively, patients were identified through hospital surgical schedules. The routine preoperative checkup was performed, and written informed consent was obtained from the guardians of the patients. The standardized anesthesia protocol was provided by professional pediatric anesthesiologists. All children are fasted from solid food for six hours and from water two hours before surgery, and intravenous access was established in the ward.The preoperative anxiety of children was measured upon entrance into the operation room and during induction by a trained pediatric anesthesiologist using the PAB Score. The medium sensor (FORE-SIGHT ELITE Cerebral Oxygen Saturation Monitor; NIRS, CAS Medical Systems Inc, Branford, CT) was attached to the clean and dry forehead above the eyebrows of the child before induction of anesthesia to get the baseline value of rScOAfter surgery, the children were transferred to the post anesthesia care unit (PACU). The child was scored by the same trained observer using the PAED form and was escorted back to the ward for further observation after the child’s status stabilized. | PMC10401797 | |
Data collection | hypospadias, fracture, scoliosis | HYDRONEPHROSIS, ABDOMINAL TUMOR, SCOLIOSIS | Data were routinely collected using a standardized electronic anesthesia system (Docare, MedicalSystem Company). Information collected included demographic data including age, sex, weight, operation time, and operation type (scoliosis, fracture, abdominal tumor, hypospadias, hydronephrosis, and biliary tract). Preoperative data including the PAB score, pre-anesthesia HR, pre-anesthesia SBP, pre-anesthesia diastolic blood pressure (DBP), pre-anesthesia MAP, and pre-anesthesia rScO | PMC10401797 |
Construction and validation of the nomogram | REGRESSION | The primary outcome of the study is the presence of NPOBC on postoperative day 3. The risk factors for the presence of NPOBC on postoperative day 3 were identified using univariate logistic regression, and varibles with P < 0.05 were included in the multivariate logistic analysis. A total score was calculated by analyzing the scores corresponding to each predictor variable in the nomogram, and a probability of NPOBC was calculated. Different methods were used to evaluate and validate the poformance of the prediction model. The calibration curve were drawn to reveal the reliability of the prediction model [ | PMC10401797 | |
Statistical analyses | The sample size was estimated using PASS software (version 15.0). We assumed three or four independent risk factors would be used to predict NPOBC, and each factor requires at least 10 to 15 cases of primary outcomes to ensure the reliability of the estimation. Based on our previous results, assuming the incidence of NPOBC is between 22 and 52%. We accepted an α error of 0.05 and a β error of 0.2 in a bilateral contrast, and the sample size of 136 (30/0.22) was targeted. Assuming a 15% of possible withdrawals and loss of follow‐up, the calculated sample size was 170.Statistical analyses in the present study were performed using R software (version 4.0.3, | PMC10401797 | ||
Discussion | desaturation, anxiety, cerebral desaturation | Children in critical stages of development are particularly vulnerable. Surgical stress response, intraoperative blood transfusion, a decrease in rScONPOBC is associated with multiple risk factors, including young age, type of surgery, preoperative and induction anxiety in the preoperative periods, and method of inhalation anesthetic agent used [NPOBC has previously been associated with preoperative anxiety in children [Another independent risk factor for NPOBC on postoperative day 3 is the presence of cerebral desaturation (OR: 1.16, 95%CI: 1.02–1.32). There is no standardized, general absolute to define pathological brain region desaturation in children, perhaps partially attributable to the wide baseline variation of rScOOur nomogram included all three independent risk factors described above. To the best of our knowledge, this is the first study to build the nomogram with preoperative factors based on a prospective cohort of 192 patients and quantify the probability of NPONC individually. Parametric bootstrapping internally validated the established nomogram, which demonstrated good generalization and predictive performance. According to the calibration plot, the predicted and observed consequences were in agreement. ROC analysis revealed excellent discrimination between the nomogram and the NPOBC on postoperative day 3 (AUC = 0.762, 95%CI: 0.691—0.833). As the incidence of NPOBC in the early postoperative period reported in previous studies ranged from 24 to 80% [The prospective design of this study may exclude some confounding factors. However, it is important to note that our study has some limitations. First, PHBQ is designed for assessing children’s posthospitalization and postoperative new onset behavioral changes and has a good profile of reliability and validity [ | PMC10401797 | |
Conclusion | postoperative pain, cerebral desaturation | In our prospective observational study, 44.8% of children (86/192 patients) developed NPOBC on postoperative day 3. Patients with higher PAB scores, presence of cerebral desaturation, and higher postoperative pain score were more likely to develop NPOBC on postoperative day 3. We innovatively established and validated a nomogram for predicting NPOBC. By using the nomogram, clinicians can make customized assessments of patients, identify patients at higher risk of NPOBC early on, and provide them with the care and support they need, thereby minimizing the likelihood of negative outcomes after surgery. | PMC10401797 | |
Acknowledgements | We would like to thank the reviewers for their careful review and critique, which helped us improve the manuscript considerably. | PMC10401797 | ||
Authors’ contributions | This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.All authors made a significant contribution to the work reported and agree to be accountable for all aspects of the work. Lijing Li designed and drafted the manuscript. Jianmin Zhang participated in designing and coordinating the study. Jiayi Li helped with the data analyses and interpretation. Yi Ren and Zhengzheng Gao helped with the study design, manuscript review. Jia Gao, Fuzhou Zhang, Fang Wang and Tiehua Zheng performed the clinical part of the study and acquired data. All authors have read and approved the final manuscript. | PMC10401797 | ||
Funding | The authors received no funding for this work. | PMC10401797 | ||
Availability of data and materials | The key data are contained in the figures, tables, and additional files. The datasets used and/or analyzed during this study can be further obtained from the corresponding author, Jianmin Zhang, on reasonable request. | PMC10401797 | ||
Declarations | PMC10401797 | |||
Ethics approval and consent to participate | The study protocol was established, according to the ethical guidelines of the Helsinki Declaration and the Ethics Committee of Beijing Children's Hospital, Capital Medical University has approved the use of fully anonymized cohort data for research (ID: 2021-E-114-Y). All children's guardians provided written informed consent. | PMC10401797 | ||
Consent for publication | Not applicable, there are no details, images or videos on individuals within the manuscript. | PMC10401797 | ||
Competing interests | The authors declare that they have no competing interests. | PMC10401797 | ||
References | PMC10401797 | |||
Background | Janus tyrosine | POLYMYALGIA RHEUMATICA (PMR), DISEASE, PMR, PATHOGENESIS, INFLAMMATORY DISEASE | The authors have declared that no competing interests exist.Polymyalgia rheumatica (PMR) is a common inflammatory disease in elderly persons whose mechanism of pathogenesis has not been elucidated. Glucocorticoids are the main first-line treatments but result in numerous side effects. Therefore, there is a need to explore pathogenetic factors and identify possible glucocorticoid-sparing agents. We aimed to study the pathogenetic features of the disease and assess the efficacy and safety of Janus tyrosine kinase (JAK)-inhibitor tofacitinib in patients with PMR. | PMC10309604 |
Methods and findings | ± | DISEASE, PMR, ADVERSE EVENTS, SECONDARY | We recruited treatment-naïve PMR patients from the First Affiliated Hospital, Zhejiang University School of Medicine, between September 2020 and September 2022. In the first cohort, we found that the gene expression patterns of peripheral blood mononuclear cells (PBMCs) in 11 patients (10 female, 1 male, age 68.0 ± 8.3) with newly diagnosed PMR were significantly different from 20 healthy controls (17 female, 3 male, age 63.7 ± 9.8) by RNA sequencing. Inflammatory response and cytokine–cytokine receptor interaction were the most notable pathways affected. We observed marked increases in expression of IL6R, IL1B, IL1R1, JAK2, TLR2, TLR4, TLR8, CCR1, CR1, S100A8, S100A12, and IL17RA, which could trigger JAK signaling. Furthermore, tofacitinib suppressed the IL-6R and JAK2 expression of CD4In the second cohort, patients with PMR were randomized and treated with tofacitinib or glucocorticoids (1/1) for 24 weeks. All PMR patients underwent clinical and laboratory examinations at 0, 4, 8, 12, 16, 20, and 24 weeks, and PMR activity disease scores (PMR-AS) were calculated. The primary endpoint was the proportion of patients with PMR-AS ≤10 at weeks 12 and 24. Secondary endpoints: PMR-AS score, c-reactive protein (CRP), and erythrocyte sedimentation rate (ESR) at weeks 12 and 24. Thirty-nine patients with newly diagnosed PMR received tofacitinib, and 37 patients received glucocorticoid. Thirty-five patients (29 female, 6 male, age 64.4 ± 8.4) and 32 patients (23 female, 9 male, age 65.3 ± 8.7) patients completed the 24-week intervention, respectively. There were no statistically significant differences in primary or secondary outcomes. At weeks 12 and 24, all patients in both groups had PMR-AS <10. PMR-AS, CRP, and ESR were all significantly decreased in both groups. No severe adverse events were observed in either group. Study limitations included the single-center study design with a short observation period. | PMC10309604 |
Conclusions | PMR, PATHOGENESIS | We found that JAK signaling was involved in the pathogenesis of PMR. Tofacitinib effectively treated patients with PMR as glucocorticoid does in this randomized, monocenter, open-label, controlled trial (ChiCTR2000038253). | PMC10309604 | |
Trial registration | POLYMYALGIA RHEUMATICA | This investigator-initiated clinical trial (IIT) had been registered on the website (Xinlei Ma and co-workers investigate tofacitinib as a possible treatment for patients with polymyalgia rheumatica. | PMC10309604 | |
Author summary | PMC10309604 | |||
Why was this study done? | PATHOGENESIS, POLYMYALGIA RHEUMATICA (PMR), INFLAMMATORY DISEASE, PMR | Polymyalgia rheumatica (PMR) is a common inflammatory disease in elderly persons whose pathogenesis is not elucidated. Glucocorticoids are the first-line drugs for patients with PMR but result in numerous side effects. New therapeutic agents are urgently needed.We investigated whether Janus tyrosine kinase (JAK) signaling was involved in the pathogenesis of PMR and if JAK-inhibitor tofacitinib was effective in the treatment of patients with PMR. | PMC10309604 | |
What did the researchers do and find? | PMR | In this prospective study, we showed increased expression of many genes related to JAK signaling in PMR.We further found that patients with PMR receiving tofacitinib ( | PMC10309604 | |
What do these findings mean? | PMR, PATHOGENESIS | These results suggest that JAK signaling is involved in the pathogenesis of PMR.In this study, tofacitinib monotherapy had similar benefits to glucocorticoid treatment in patients with PMR. | PMC10309604 | |
Data Availability | The raw sequence data reported in this paper have been deposited in the Genome Sequence Archive (Genomics, Proteomics & Bioinformatics 2021) in National Genomics Data Center (Nucleic Acids Res 2022), China National Center for Bioinformation / Beijing Institute of Genomics, Chinese Academy of Sciences (GSA-Human: HRA003316) that are publicly accessible at | PMC10309604 | ||
Introduction | pelvic girdle, morning stiffness, pain | DISEASE, PATHOGENESIS, PMR, POLYMYALGIA RHEUMATICA (PMR), INFLAMMATORY DISEASE | Polymyalgia rheumatica (PMR) is an inflammatory disease in elderly patients (>50 years old) characterized by pain and morning stiffness in the shoulder, neck, and pelvic girdle [PMR is a multigene susceptibility disease, and many alleles may be involved in the pathogenesis of this disease [ | PMC10309604 |
Methods | PMC10309604 | |||
Ethics statement | This study was approved by the Medical Ethical Committee of the First Affiliated Hospital, Zhejiang University School of Medicine (the approval number: IIT20200070C-R1). The formal consent was obtained from participant in a written informed document before the study start. The study was performed according to the recommendations of the Declaration of Helsinki. | PMC10309604 | ||
Study design | This project had 2 phases, an observational study, and a Phase II, randomized, monocenter, open-label, controlled, noninferiority trial. Trial reporting was guided by the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline ( | PMC10309604 | ||
Study flowchart and treatment design. | HC | PMR, RECRUITMENT, POLYMYALGIA RHEUMATICA | Flowchart shows recruitment, randomization, and study design. HC, healthy control; IL-6R, interleukin 6R; PBMC, peripheral blood mononuclear cell; PMR, polymyalgia rheumatica; RT-PCR, real-time PCR. | PMC10309604 |
Participants | PMR | We included diagnosed PMR patients who fulfilled the 1982 Chuang criteria [ | PMC10309604 | |
RNA sequence analysis and semiquantitative real-time PCR | PMR | Peripheral blood mononuclear cells (PBMCs) were isolated from 11 patients with new diagnosed PMR and 20 healthy controls (HCs) according to the protocol [ | PMC10309604 | |
Isolation of CD4 | PMR | We collect the PBMCs from patients with PMR and then isolate the CD4 | PMC10309604 | |
Intervention, randomization, and outcomes | PMR | Patients with newly diagnosed PMR were randomized (1/1) to the tofacitinib (5 mg BID) group and the glucocorticoids (Pred 15 mg/day, gradually tapered) group using an envelope by a statistician who was not involved in the trial conduct. Rheumatologists enrolled participants, while statistician assigned participants to interventions. Statistician created 94 random numbers (from 0 to 100) using an online tool (The primary endpoint was the proportion of patients with PMR-AS < 10 at weeks 12 and 24. Secondary endpoints: PMR-AS score, CRP, and ESR at weeks 12 and 24. Patient information on demographic data, clinical features, serological profiles, and medications was obtained from medical records. | PMC10309604 | |
Statistical analysis | PMR, REMISSION, SECONDARY | Results are expressed as median and standard deviation (SD). In cohort 1, comparisons of the quantitative baseline characteristics data between 11 PMR patients and 20 HCs were performed using the nonparametric Mann–Whitney test, while qualitative data were analyzed using Fisher exact test.In cohort 2, the complete remission (CR) rate (PMR-AS < 10) at 12 and 24 weeks is the primary endpoint, and the CR rate of the patients in the experimental group is predicted as 95%, the CR rate of the control group was 85%, the margin of superiority/noninferiority was 0.1, the sample distribution ratio of the 2 groups was 1:1, calculated by PASS software, 38 cases are needed in each group. According to the 18% to 20% dropout, finally, a minimum of 47 cases per group is required. A sample size of 47 participants per group was estimated to provide at least 80% power to demonstrate the For the primary efficacy analysis, we included all patients who finished the follow-up. The proportion of patients with PMR-AS < 10 in 2 groups were analyzed by Fisher exact test. The secondary efficacy analysis was assessed using repeated ANOVA, including PMR-AS score, CRP, and ESR between 2 groups at weeks 0, 12, and 24. Safety was assessed for patients who were randomly assigned and received at least 1 dose of the study drug and analyzed by Fisher exact test. | PMC10309604 | |
Results | PMC10309604 | |||
The patients with PMR were in a state of high inflammation | PMR | The baseline characteristics of the 11 PMR patients and 20 HCs ( | PMC10309604 | |
Inflammatory gene expression profiles in patients with PMR | PMR | The principal component analysis (PCA) and heatmap showed that gene expressions of PBMCs in patients with PMR differed from those of HC ( | PMC10309604 | |
Abundant inflammatory genes related to Janus tyrosine kinase (JAK) signaling in patients with PMR | Volcano results demonstrated that many genes expression were significantly increased; for example, TLR8 ( | PMC10309604 | ||
Tofacitinib suppressed the IL-6R and JAK2 expression of CD4 | Next, we isolated the CD4 | PMC10309604 | ||
Tofacitinib effectively treats patients with PMR as glucocorticoid does | PMR | In the second cohort, 35 patients with newly diagnosed PMR received tofacitinib and 32 patients received glucocorticoid who completed the 24-week follow-up ( | PMC10309604 | |
Tofacitinib effectively treats patients with PMR as glucocorticoid does. | DISEASE, PMR, POLYMYALGIA RHEUMATICA | Patients with newly diagnosed PMR were randomized to the tofacitinib (5 mg BID) group and the glucocorticoids (Pred 15 mg/day, gradually tapered) group for 24 weeks. All PMR patients underwent clinical and laboratory examinations at 0, 4, 8, 12, 16, 20, and 24 weeks, and PMR-AS were calculated. At weeks 12 and 24, all patients in both groups had PMR-AS <10. PMR-AS, CRP, and ESR were all significantly decreased at weeks 12, and 24 in both groups. CRP, c-reactive protein; ESR, erythrocyte sedimentation rate; PMR, polymyalgia rheumatica; PMR-AS, PMR activity disease score. | PMC10309604 | |
Baseline characteristics of PMR patients in tofacitinib and glucocorticoid group. | shoulder girdle, morning stiffness | DISEASE, PMR, REMISSION, POLYMYALGIA RHEUMATICA | CRP, c-reactive protein; ESR, erythrocyte sedimentation rate; EUL, ability to elevate the upper limbs (0 = lifted above the shoulder girdle, 1 = to the shoulder girdle, 2 = below the shoulder girdle, 3 = cannot lift); NSAID, nonsteroidal anti-inflammatory drug; PMR, polymyalgia rheumatica; PMR-AS, PMR activity disease score; SD, standard deviation; VS, visual analogue scale.PMR-AS: CRP (mg/dl) + patient self-evaluation (0–10 visual scale) + physician global assessment (0–10 visual scale) + [morning stiffness (min) × 0.1] + EUL (0–3).Furthermore, the blood samples were collected from 4 patients under treatment with tofacitinib at week 24, and it was isolated from 1 patient at week 12. The gene expression profiles of patients with PMR that were in remission after tofacitinib therapy (the PMR-R group) differed from those of patients in the newly diagnosed PMR with active disease activity (PMR) group (PMR-R versus PMR; 1,403 DEGs shown in In the tofacitinib group, 2 patients personally reduced the dose of tofacitinib from 5 mg BID to 5 mg QD at week 20 and then kept the dosage until week 24. Another patient reduced the dose of tofacitinib from 5 mg BID to 5 mg QD at week 8 and then kept the dosage until week 24. All 3 patients were always in remission even after reducing the dose of tofacitinib.One patient in the tofacitinib group had a rapid decrease in PMR-AS score (28.72 to 2.85) at the initial 4 weeks, but had an increase in PMR-AS score (7.77) at week 12, finally reduced to lower disease activity at week 24. One patient in the glucocorticoid group had a good response well (24.87 to 2.2) at the initial 16 weeks, but had disease exacerbation (2.2 to 7.6) at week 20, eventually a disease score of 4.97 at week 24. There were 4 patients lost to the follow-up due to traffic inconvenience but still had a good response to tofacitinib (PMR-AS score < 10) at weeks 8 to 12. In the control group, 5 cases had a good response to glucocorticoid (PMR-AS score < 10) at weeks 8 to 16 but were lost to the follow-up after week 16. | PMC10309604 |
Safety | death | There were no severe AEs or death in the 2 groups ( | PMC10309604 | |
AE of tofacitinib and glucocorticoid during the 24-week period. | malignancy, DVT, malignant tumors, cytopenia, Cancer | PULMONARY EMBOLISM, RECURRENCE, DEEP VEIN THROMBOSIS, DVT, MALIGNANT TUMORS, ADVERSE EVENT, CYTOPENIA, CANCER | AE, adverse event; ALT, alanine aminotransferase; AST, aspartate aminotransferase; DVT, deep vein thrombosis; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; PE, pulmonary embolism; SAE, severe adverse event; TC, total cholesterol; TG, triglyceride; ULN, upper limit of normal.Cancer: new malignancy or patients with malignant tumors who have been successfully treated for more than 5 years before screening, recent recurrence; cytopenia: Hb <90 g/L, or neutrophil <1 × 10 | PMC10309604 |
Discussion | cancers, tumor necrosis, RA, pain, deaths, vascular syndrome | CANCERS, TUMOR NECROSIS, EVENT, DIABETES MELLITUS, GCA, GIANT CELL ARTERITIS, CHRONIC INFLAMMATION, PMR, HYPERTENSION, PATHOGENESIS, INFLAMMATORY DISEASE | PMR is a classic inflammatory disease with an acute onset but an unknown cause. In the first cohort, we found that the patients with PMR had a significant inflammatory gene expression profile related to JAK signaling via RNA sequence analysis. Furthermore, tofacitinib suppressed the IL-6R and JAK2 expression of CD4As we know, patients with active PMR or RA both commonly suffer from severe pain. During clinical practice, NSAIDs are used to relieve pain for patients with RA. Tofacitinib usually works at week 2 in the treatment of RA. The NSAIDs were allowed in the tofacitinib group in the first 2 weeks and then were stopped. We can still consider the therapeutic effect of tofacitinib for PMR during the 24-week period. Furthermore, no severe side effects were found in the 24-week period of treatment of tofacitinib. We also found that the response rate was higher as expected in patients with PMR. We think tofacitinib may have a high response rate in the new diagnosed PMR patients who were naïve to glucocorticoid or biological agents. If we treat PMR patients who were relapsing on prior glucocorticoid or MTX or other agents in the real world, the expected response rate of tofacitinib may be not high.IL-6 is one of the key pro-inflammatory cytokines involved in the pathogenesis of PMR [S100A8 is an inflammatory mediator that can induce the JAK1/2-dependent signaling [As we know, glucocorticoids are the first-line drugs in the treatment of PMR. However, they come with numerous side effects. JAK1 and JAK3 were involved in the chronic inflammation of media and large arteries in an animal model of giant cell arteritis (GCA) [Here, we did an IIT study to explore whether single tofacitinib effectively treats new patients with PMR as glucocorticoid does. Excitingly, patients with PMR had a good response to tofacitinib therapy or steroid therapy in our study. We also collect the safety data of the drug after 24 weeks. No severe side effects or deaths were found in the 24-week period of treatment of tofacitinib. Recent ORAL surveillance study demonstrated that RA patients aged ≥50 years with ≥1 additional cardiovascular risk factor (for example, current cigarette smoker, hypertension, diabetes mellitus) have higher risks of major adverse cardiovascular event (MACE) and cancers during a median follow-up of 4.0 years, comparing the combined tofacitinib doses (5 or 10 mg twice daily) with tumor necrosis factor inhibitors (TNFis) [Our study has some limitations. The number of patients was small recruited from our medical center. We do not finish the functional experiment of candidate genes involved in the pathogenesis of PMR. PMR is closely associated with GCA, which is mainly an inflammatory vascular syndrome involving medium—and large—arteries. In most cases, PMR occurs in isolation but may be found in 40% to 60% of patients with GCA [ | PMC10309604 |
Conclusions | PMR, REMISSION | Many inflammatory genes associated with JAK signaling are highly enriched in PBMCs from patients with PMR. Tofacitinib, a pan JAKi, effectively treated the PMR patients with clinical remission and a sharp decrease in CRP and ESR in this clinical trial study. Tofacitinib may effectively treat PMR patients. In the future, whether patients with PMR can be treated with tofacitinib alone should be carried out in the large phase randomized controlled trial. | PMC10309604 | |
Supporting information | PMC10309604 | |||
CONSORT 2010 checklist of information to include when reporting a randomized trial. | (DOC)Click here for additional data file.This supporting information contains the following items: (1) Original protocol in English. (2) Original protocol in Chinese. (3) Approved letter for this study in our hospital.(PDF)Click here for additional data file. | PMC10309604 | ||
KEGG and GO analysis of gene expression of PBMCs from PMR and HC. | (JPG)Click here for additional data file. | PMC10309604 | ||
Tofacitinib suppressed the IL-6R and JAK2 expression of CD4 | (TIF)Click here for additional data file. | PMC10309604 | ||
One representative case after tofacitinib therapy. | (TIF)Click here for additional data file. | PMC10309604 | ||
The gene expression profiles of patients with PMR that were in remission after tofacitinib therapy. | (TIF)Click here for additional data file. | PMC10309604 | ||
Primers for RT-PCR. | (DOCX)Click here for additional data file. | PMC10309604 | ||
Demographic and clinical parameters of 11 patients with PMR and 20 healthy controls in the first cohort. | (DOCX)Click here for additional data file. | PMC10309604 | ||
Differentially expressed genes (DEGs) between PMR and HC. | (XLSX)Click here for additional data file. | PMC10309604 | ||
Differentially expressed genes (DEGs) between PMR and PMR-R. | (XLSX)Click here for additional data file. | PMC10309604 | ||
Differentially expressed genes (DEGs) between PMR-R and HC. | (XLSX)Click here for additional data file. | PMC10309604 | ||
Abbreviations | necrosis | DISEASE, NECROSIS | anticitrullinated protein/peptide antibodyadverse eventantinuclear antibodycalcium pyrophosphate depositioncomplete remissionc-reactive proteindifferentially expressed genedeep vein thrombosiserythrocyte sedimentation ratefalse discovery ratefragments per kilobase of exon model per million mapped fragmentsgiant cell arteritisGene Ontologyhealthy controlhuman leukocyte antigen B27investigator-initiated clinical trialinterleukin 6Janus tyrosine kinaseJAK inhibitorKyoto Encyclopedia of Genes and Genomesmajor adverse cardiovascular eventmitogen-activated protein kinasenuclear factor kappa Bnonsteroidal anti-inflammatory drugperipheral blood mononuclear cellprincipal component analysispulmonary embolismpolymyalgia rheumaticaPMR activity disease scorerheumatoid arthritisrheumatoid factorstandard deviationsignal transducer and activator of transcriptionToll-like receptortumor necrosis factor inhibitorvisual analogue scale | PMC10309604 |
References | PMC10309604 | |||
Purpose | diabetic retinopathy | DIABETIC RETINOPATHY | Real-world evaluation of a deep learning model that prioritizes patients based on risk of progression to moderate or worse (MOD+) diabetic retinopathy (DR). | PMC10715315 |
Methods | This nonrandomized, single-arm, prospective, interventional study included patients attending DR screening at four centers across Thailand from September 2019 to January 2020, with mild or no DR. Fundus photographs were input into the model, and patients were scheduled for their subsequent screening from September 2020 to January 2021 in order of predicted risk. Evaluation focused on model sensitivity, defined as correctly ranking patients that developed MOD+ within the first 50% of subsequent screens. | PMC10715315 | ||
Results | We analyzed 1,757 patients, of which 52 (3.0%) developed MOD+. Using the model-proposed order, the model's sensitivity was 90.4%. Both the model-proposed order and mild/no DR plus HbA1c had significantly higher sensitivity than the random order ( | PMC10715315 | ||
Conclusions | The model can help prioritize follow-up visits for the largest subgroups of DR patients (those with no or mild DR). Further research is needed to evaluate the impact on clinical management and outcomes. | PMC10715315 | ||
Translational Relevance | Deep learning demonstrated potential for risk stratification in DR screening. However, real-world practicalities must be resolved to fully realize the benefit. | PMC10715315 | ||
Introduction | visual loss, Diabetic retinopathy, diabetes | COMPLICATION, DIABETIC RETINOPATHY, DIABETES | Diabetic retinopathy (DR), a complication of diabetes, is a major cause of visual loss globally.However, DR screening faces a scaling challenge as the number of patients with diabetes rises.Of those with diabetes, 75% are in low-to-middle income countries, and the burden of diabetes and its related conditions disproportionately affects these populations.These challenges may be addressed by optimizing screening intervals personalized to a patient's risk of DR progression, potentially improving program efficiency and vision-related outcomes while also reducing cost. In recent times, deep learning (DL) systems have been applied to automated computational grading of fundus photos for DR assessment, and many have shown expert-level accuracy.In the present study, we conducted a prospective evaluation of the clinical usefulness of the aforementioned DR progression model to prioritize patients according to their risk of progression based on prior CFPs. This study was conducted in Thailand during the COVID-19 pandemic, which significantly impacted the country's national DR screening program—a program that serves an estimated 4.5 million patients with diabetes. | PMC10715315 |
Methods | PMC10715315 | |||
Inclusion and Exclusion Criteria | NPDR, diabetic macular edema, diabetes | EYE DISEASE, RETINA, PROLIFERATIVE, DIABETIC MACULAR EDEMA, DIABETES | The Thailand Prospective Study (TPS) (Thai Clinical Trials Registry #TCTR20190902002) explored the use of Google's automated DR screening software in a real-world clinical setting.Five hospitals across Thailand were involved in our study—namely, Phrao, Chomthong, Rajavithi, Khlong Luang, and San Patong Hospitals. Patients underwent a baseline visit as a part of the TPS. The inclusion criteria for TPS were all patients with diabetes in the national diabetes registry above age 18. Exclusion criteria for TPS were patients with a previous diagnosis of DME, severe NPDR, or proliferative DR; prior laser treatment of the retina or retinal surgery; other non-DR eye disease requiring referral to an ophthalmologist; or inability to have fundus photo taken of either eye for any reason. A subset of TPS patients who underwent DR screening between September 2019 and January 2020 and were graded as mild or no DR and not having diabetic macular edema in both eyes were analyzed for our study. A description of the grading approach is described in subsequent paragraphs. | PMC10715315 |
DL Progression Model | Details of the model have been published previously.The model produces a progression risk score per eye as a number between 0 and 1, representing the likelihood of an eye progressing in terms of DR severity within certain time windows. For this study, the likelihood of progression from no or mild DR to MOD+ within a 1-year window (i.e., at the subsequent annual screen) was used. The model evaluates each eye separately, and the maximum of the two per-eye scores was used for prioritization ranking of patients for this study. | PMC10715315 | ||
Generation of Ranked Lists and Clinical Workflow | CFPs from the baseline screening visit were input into the model. A list of deidentified participant IDs, ranked according to progression score output by the model (from highest to lowest) were provided to the Principal Investigator at each site.Patients were scheduled for their subsequent screening between September 2020 and January 2021 in the proposed order determined by the model from highest to lowest risk. Patients were provided with specific appointment dates by each participating hospital. However, in cases where a patient was unable to accept the given appointment, staff were instructed to find the nearest acceptable appointment day while ensuring that all participants will be screened within 12 to 14 months of their last DR screening, as per the standard of care. Because patients did not always show up on the appointed date, both the scheduled appointment and the actual screening dates were recorded to facilitate analysis. During the visit, clinical staff obtained CFPs and metadata according to routine screening protocol. Fundus images were obtained using Topcon Maestro 3D OCT-1, Topcon TRC NW300, Topcon TRC NW200 and Nidek AFC-210 cameras. | PMC10715315 | ||
Grading | diabetic macular edema, diabetes | RETINA, BLIND, DIABETIC MACULAR EDEMA, MACULAR EDEMA, DIABETES | Thai retina specialists did the grading of fundus images to obtain the baseline grades of patients with diabetes in the baseline visit (there were two retina specialists, each responsible for a separate set of hospitals assigned to them). After this, fundus images and metadata were transferred to coinvestigators at Google Health via a secure Cloud-based server. A single retina specialist from a pool of six U.S. board-certified retina specialists graded each fundus image for DR and diabetic macular edema to define the subsequent visit grade. This would serve as the progression ground truth for the DL model. Images were ordered randomly and assigned randomly to blind the retina specialist to the model scores. Analysis was performed to compare the DR/diabetic macular edema grades assigned by the retina specialist (progression ground truth) with the ranked list to assess the model's effectiveness in prospectively prioritizing cases. | PMC10715315 |
Statistical Analysis | diabetes | RETINA, DIABETES | Analysis was restricted to those patients with available hemoglobin A1c (HbA1c) results at baseline (because this was a required part of data collection), known follow-up dates, at least one image at the subsequent screen, and known MOD+ outcomes at the subsequent visit. Patients who did not attend within ±60 days of their appointment date, or attended significantly earlier than expected (<150 days since the baseline visit), were excluded from analysis. Images deemed poor quality by the retina specialist were also excluded from analysis. A CONSORT diagram illustrating participant inclusions and exclusions is presented in The analyses were performed by combining patients across sites. The goal of the analysis was to compare the ranking ability of different approaches to scheduling subsequent screenings:Approach A: Baseline - ordered randomlyApproach B: Baseline - all mild DR then all no DR, random order within each groupApproach C: All mild DR then all no DR, ranked by decreasing HbA1c measurement at baseline within each groupApproach D: Model proposed orderApproach E: Observed screening orderWhereas approach A is a simplified approximation to current practice of following a standardized screening interval, approaches B and C represent clinically informed ordering approaches based on available information (DR grade and HbA1c). Approach D represents our DL model, and approach E is the actual order observed in this study. Further detail on how rankings were combined across sites is provided in the The primary outcome measure was the sensitivity of the model, that is, its ability to rank patients by risk of MOD+ within the first 50% of subsequent screens. This was calculated by dividing the number of patients with MOD+ in the first 50% of rankings by the total number of patients with MOD+ at the subsequent screen.Approaches A and B were considered the baseline approaches for comparison with approaches C, D, and E. To test the superiority of the model's sensitivity, a permutation test (with 10,000 iterations) was performed under the null hypothesis that the ranking by the model is no different from random ranking (approach D vs A). A second permutation test was performed with the null hypothesis of ranking all mild DR then all no DR, random order within each group (approach D vs B). For each, alpha was set at 0.05. We also report statistical tests comparing approach C and approach E versus both stated null hypotheses in the In addition, we conducted a one-sided Mann–Whitney We devised additional ranking approaches using additional baseline variables, namely, age, duration of diabetes, and insulin use. We first restricted this analysis to a subset of patients where all relevant variables were present, yielding 1646 of 1757 patients. For each baseline variable, we devised a ranking scheme which used the baseline grades to rank the patients, and within those two groups (no DR and mild DR), we ranked the patients using the additional baseline variable. Using the progression ground truth, these rankings were then converted into a plot of fraction screened versus sensitivity plots. For a fair comparison, we also reran the previous approaches (A, B, C, D, and E) on this subset of patients. The results are presented in | PMC10715315 |
Results | PDR, nonproliferative, NPDR | PROLIFERATIVE | A total of 3507 patients who attended their baseline screening with no or mild DR were ranked by the model at five screening sites (Demographics and characteristics of analysis-eligible patients are shown in Demographics and Characteristics of Analyzed DatasetHbA1c, hemoglobin A1c.; MOD+, moderate or worse DR; NPDR, nonproliferative DR; PDR, proliferative DR.Values are mean ± standard deviation, number, or number (%). | PMC10715315 |
Proposed Versus Observed Order | The sites attempted to schedule patients in the order proposed by the model (approach D as described in Methods). Three sites (Rajavithi, San Patong, and Khlong Luang) implemented the model proposed order for the majority of patients (The actual screened order versus the model proposed order (according to the DL progression model), with each patient represented by a dot. Rajavithi, San Patong, and Khlong Luang closely follow the model proposed order, illustrated by a monotonic increasing relationship between the two axes. Phrao followed the order in numerous discrete batches owing to scheduling issues. | PMC10715315 | ||
Rate of Progression | The annual transition probabilities between various stages of DR has been studied as part of the assessment of cost effectiveness of screening. | PMC10715315 | ||
MOD+ Sensitivity Versus Fraction Screened | PMC10715315 | |||
Excluding Phrao | SENSITIVITY | Analysis of the three sites, after excluding Phrao, demonstrates that the sensitivity of the observed screened order (approach E) and model proposed order (approach D) was 100.0% and 86.7%, respectively (Sensitivity versus fraction screened for Rajavithi, San Patong, and Khlong Luang in aggregate, after excluding Phrao (see Results). The green line represents the model proposed order (approach D), the red line represents the actual observed order (approach E), the blue line represents in order of mild DR and then no DR (approach B), the orange line represents in order of mild/no DR and decreasing HbA1c within each group (approach C), and the black line represents a random order (approach A). | PMC10715315 | |
Including Phrao | SENSITIVITY | Using the model proposed order, 90.4% of individuals that developed MOD+ were successfully ranked within the first 50% of subsequent screens (Sensitivity versus fraction screened for Rajavithi, San Patong, Khlong Luang, and Phrao in aggregate (see Results). The green line represents the model proposed order (approach D), the red line represents the actual observed order (approach E), the blue line represents in order of mild DR and then no DR (approach B), the orange line represents in order of mild/no DR and decreasing HbA1c within each group (approach C), and the black line represents a random order (approach A). Each accompanying line for black and blue represents one run of the randomization (which is applicable when there are ties, such as between mild DR patients or between no DR patients). To show the variability that might occur, we plot 100 accompanying lines with a faint color around the main solid line representing the average value. | PMC10715315 | |
Comparison With Alternative Ranking Approaches | We compared three alternative grading approaches on a subset of patients where data was available for relevant baseline variables. Plots can be seen in | PMC10715315 | ||
Discussion | diabetic, vision loss | MICROANEURYSMS | In this prospective study, we demonstrate the capability of our DL model to prioritize diabetic patients for a subsequent screening visit, based on their personalized risk of DR progression. Our evaluation assessed the model's clinical usefulness during COVID-19 recovery efforts in Thailand, with the objective of mitigating delays in scheduling follow-up screening appointments for those at higher risk. The DL model was implemented at sites with high patient volumes, where prioritization of patients based on DR risk could potentially lead to a more efficient allocation of resources and ensure that patients with a higher risk of DR progression receive timely eye clinic visits for appropriate management. Furthermore, our approach directly tackles the problem of optimizing screening intervals by risk stratifying the two largest groups of patients: those without any DR and those with mild DR.The proposed screening order determined by the DL model (approach D) showed a significantly higher sensitivity (defined as fraction of patients who were MOD+ at second screen who were screened in the first 50% of subsequent screens) compared with baseline methods (approach A and B). The observed screening order (approach E) was highly consistent with the model proposed order at three of the four screening sites, with the exception of the rural Phrao site. At the three consistent-ordering sites (i.e., excluding Phrao), our model successfully provided a more accurate personalized risk assessment to optimize screening intervals. As a result, a greater proportion of patients at higher risk were scheduled for their subsequent annual visit prior to those with lower risk, compared with the baseline methods, ensuring overall earlier detection of progression to MOD+. However, owing to few MOD+ cases at these three sites, there was a lack of power to measure a significant improvement in sensitivity compared with the baseline approaches. Including Phrao in the aggregate analysis, our results showed that simply ordering patients by their baseline grade (mild DR followed by no DR) for the subsequent visit identified 44.3% of those that progressed to MOD+ at 10.2% screened. Although the model proposed order and rankings by grade and HbA1c also demonstrated the same sensitivity at 10.2% screened, both quickly outperformed the baseline approaches as the fraction screened increased, suggesting that both the DL model and HbA1c were good signals for prioritizing patients for subsequent screenings. A notable advantage of the DL model compared with rankings by baseline grade and HbA1c is that it solely requires CFPs (which are already taken for DR screening purposes) as an input, making it a more practical solution for implementation if HbA1c values are not already available. However, in settings where the DL system cannot be deployed, or where HbA1c monitoring is routine, it may be practical and feasible to prioritize subsequent screening visits based on baseline grade and decreasing HbA1c levels.This study provides data not just on the DL model's performance, but also real-world insights on implementation of the solution. Three of the four evaluated sites successfully implemented the DL model into real-world practice by following the model proposed order for the majority of patients. However, several real-world deployment challenges were encountered during the study. For instance, the attendance order was disrupted by patients who missed their appointment and required rescheduling during the follow-up period. Additionally, patients in certain rural areas traveled to the screening site in groups using prearranged transport, making it challenging to adhere to the model proposed order. This was particularly the case at the Phrao site, which had the greatest number of cases overall and also most of the cases that progressed to MOD+ at the subsequent screen. Because Phrao comprised the majority of patients and MOD+, this factor substantially impacted the sensitivity results using the observed order. Nevertheless, our experience demonstrates the complexities that impact the real-world use of artificial intelligence in health care and highlights the importance of careful evaluation, design, and implementation of systems for improved artificial intelligence delivery.This work aims to personalize follow-ups according to the patient's risk of progression, which may result in some patients being seen sooner or later than currently recommended guidelines. International guidelines recommend annual DR screening for those without DR or with mild DRLimitations of this study include DR grading variability, particularly for subtle retinal changes such as microaneurysms.In conclusion, our study demonstrates the ability of a DL model to help schedule follow-up visits for the two largest subgroups of DR patients based on their risk of progression. To our knowledge, this is the first real-world deployment of a DR progression model in a health care setting. Our findings suggest that such a model could enable earlier identification and treatment of patients at greatest risk of progression before irreversible vision loss occurs. At the same time, it could decrease the burden of screening attendance for lower risk patients by extending their follow-up intervals. However, additional studies are needed to further validate the DL model with a robust dataset for analysis and assess its scalability in screening programs. Further research is needed to evaluate the long-term impact of adapting screening intervals to reflect personalized risk profiles on cost-effectiveness, clinical management, and patient outcomes. | PMC10715315 |
Supplementary Material | PMC10715315 | |||
Supplement 1 | PMC10715315 | |||
Supplement 2 | PMC10715315 | |||
Supplement 3 | PMC10715315 | |||
Supplement 1 | PMC10715315 | |||
Acknowledgments | The authors would like to thank the following people for their advice and assistance with the study, or manuscript feedback: Roy Lee, Yun Liu, Naama Hammel, Michael Howell, Rajroshan Sawhney, Tayyeba Ali, and graders Eric Feinstein, Nicolas Yannuzzi, Ehsan Rahimy, Nathan Haines, Nathan Cutler, and Malav Joshi.Funded by Google and Rajavithi Hospital.Disclosure: | PMC10715315 | ||
References | PMC10715315 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.