Yale University Open Data Access Research Articles

Placebo rates in Crohn’s disease randomized clinical trials: An individual patient data meta-analysis

Understanding placebo rates is critical for efficient clinical trial design. We assessed placebo rates and associated factors using individual patient data (IPD) from Crohn's disease (CD) trials. We conducted a meta-analysis of phase 2/3 placebo-controlled trials evaluating advanced therapies in moderate-to-severe CD (2010-2021). Deidentified IPD were obtained through Vivli Inc. and Yale University Open Data Access Project. Primary outcomes were clinical response and remission. Pooled placebo rates and 95% confidence intervals (CIs) were estimated using one- and two-stage meta-analytical approaches. Regression analyses identified patient-level factors associated with placebo rates. Using IPD from eight induction (n=1147) and four maintenance (n=524) trials, overall placebo clinical response and remission rates for induction were 27% (95%CI=23-32%) and 10% (95%CI=8-14%) respectively, and 32% (95%CI=23-42%) and 22% (95%CI=14-33%) for maintenance, respectively. Among bio-naïve patients, placebo response and remission rates during induction were 29% (95%CI=24-35%) and 11% (95%CI=8-15%) respectively, and 26% (95% CI=20-33%) and 10% (95% CI=8-14%) for bio-exposed, respectively. During maintenance, bio-naïve response and remission rates were 41% (95%CI=34-48%) and 32% (95%CI=24-40%), respectively, and 29% (95%CI=24-34%) and 16% (95%CI=13-21%) for bio-exposed, respectively. Higher baseline C-reactive protein concentration predicted lower placebo rates, while higher baseline albumin levels and body mass index increased the odds of placebo outcomes. Increased baseline Crohn's Disease Activity Index and 2-item patient-reported outcome scores predicted higher response rates in induction, lower response rates in maintenance, and lower remission rates in induction and maintenance. Patient- and trial-level characteristics influence placebo rates in CD trials. Careful implementation of eligibility criteria, outcome definitions, and patient stratification may reduce placebo rates.

P1013 Placebo rates in Crohn's disease: An individual patient data meta-analysis from multiple randomised controlled trials

Abstract Background Estimating placebo rates and their determinants is essential for designing efficient randomised clinical trials (RCTs) in inflammatory bowel disease. We conducted an individual patient data (IPD) meta-analysis of placebo data from RCTs in Crohn’s disease (CD). Methods MEDLINE, Embase, and CENTRAL were searched from 01/2010 to 01/2020 for contemporary phase 2 and 3 placebo-controlled RCTs of biologics in moderate-to-severe CD. De-identified IPD from eligible trials that were available through the Vivli and Yale University Open Data Access data-sharing platforms were obtained. Primary outcomes were placebo clinical response and remission. Pooled placebo rates and 95% CIs were estimated using a 2-stage meta-analytical approach. Significant patient-level factors (P<0.05) associated with placebo rates were identified using regression analyses. Results Placebo IPD were available from 8 induction (n=1147) and 4 maintenance trials (n=524). Pooled placebo clinical response and remission rates varied based on outcome definition and prior biologic exposure (Figure). In induction trials, overall placebo response and remission rates were 27% (95% CI 23-32%) and 10% (95% CI 8-14%), respectively. Among bio-exposed patients, placebo response and remission rates were 27% (95% CI 16-40%) and 9% (95% CI 6-15%), respectively, compared with 29% (95% CI 24-34%) and 11% (95% CI 8-15%) for bio-naïve patients. Overall placebo response and remission rates in maintenance trials were 32% (95% CI 23%-42%) and 22% (95% CI 14-33%), respectively. Corresponding placebo response and remission rates for bio-exposed patients were 21% (95% CI 8-46%) and 17% (95% CI 11-25%), and 36% (95% CI 28-44%) and 24% (95% CI 15-36%) for bio-naïve patients. Placebo rates were lowest when response was defined as a ≥100-point decrease from baseline in the Crohn's Disease Activity Index (CDAI) score, and when remission was defined using stool frequency (≤1.5) and abdominal pain (≤1) subscores. Higher baseline C-reactive protein concentrations were associated with lower odds of placebo response and remission, while higher baseline albumin levels increased the odds of these outcomes (Table). Increased baseline CDAI and 2-item patient-reported outcome (PRO2) scores predicted higher odds of placebo response in induction trials, yet this reduced the odds of remission in induction trials and of both outcomes in maintenance trials. Prior failure to tumour necrosis factor (TNF) antagonist therapy was more predictive than prior TNF antagonist/biologic exposure for reducing the odds of placebo response and remission. Conclusion Placebo rates and the patient-level factors influencing them vary according to study design, clinical outcome measure and outcome definitions.

P280 Comparative responsiveness of disease activity indices in moderately-to-severely active Ulcerative Colitis: a post-hoc analysis of the UNIFI trial

Abstract Background Clinical, endoscopic, histological, and composite instruments are used to measure disease activity in ulcerative colitis (UC). We compared responsiveness of the Mayo Clinic Score (MCS), modified Mayo Clinic Score (mMS), partial Mayo Clinic Score (pMS), Robarts Histopathology Index (RHI), and UC-100 score after treatment with ustekinumab in patients with moderately-to-severely active UC. Methods Publicly available data from the phase 3 UNIFI induction trial (NCT02407236) comparing the efficacy and safety of ustekinumab versus placebo in patients with moderate-to-severe UC were obtained from Vivli (Vivli ID #7549) and Yale University Open Data Access Project (#2021-4848). Post-hoc analyses were performed on patient level data to ascertain outcomes of interest. Treatment assignment was the criterion for change in evaluation of responsiveness, which was quantified using the probability of a treated patient having a larger change than a placebo patient (win probability [WinP]) and estimated using nonparametric methods. Extent of responsiveness was interpreted according to Cohen’s benchmarks as very small (WinP<0.56); small (0.56≤ WinP<0.64); medium (0.64≤ WinP<0.71); and large (WinP≥0.71). Results Data from 961 patients with moderately-to-severely active UC randomised in UNIFI were included. At baseline, 319 patients received placebo and 624 patients received ustekinumab. A total of 912 (94.9%) patients completed the induction trial. Overall mean baseline index scores were: MCS 8.9 (±1.6), mMS 6.6 (±1.3), pMS 6.2 (±1.4), RHI 16.6 (±8.1), and UC-100 score 71.4 (±16.5) (Table). The UC-100 demonstrated a large degree of responsiveness (WinP 0.72 [95% CI: 0.66, 0.78]). The MCS (0.68 [0.62, 0.73]), mMS (0.69 [0.63, 0.75]), and pMS (0.65 [0.59, 0.71]) demonstrated similar, medium effect sizes (all comparisons p>0.05) (Figure). The RHI alone had a relatively small effect size (WinP 0.63 [0.56-0.70]). Among individual UC-100 components (Mayo Clinic stool frequency subscore, Mayo Clinic endoscopic subscore, and RHI), the endoscopic score was most responsive (WinP 0.59 [0.36-0.82]). The combination of endoscopy and RHI was associated with a medium degree of responsiveness (WinP 0.69 ([0.34, 1.00]). Conclusion UC disease activity indices are similarly responsive to change. Depending on trial design and measurement feasibility, different instruments can be considered optimal for evaluation of efficacy. The mMS is similarly responsive to the MCS without including the physician global assessment and may be more appropriate for determining trial eligibility and defining primary endpoint, whereas the UC-100 may be more appropriate for measuring treatment effect when an objective continuous measure is needed.

Relapse in clinically stable adult patients with schizophrenia or schizoaffective disorder: evidence-based criteria derived by equipercentile linking and diagnostic test accuracy meta-analysis

There is no consensus on defining relapse in schizophrenia, and scale-derived criteria with unclear clinical relevance are widely used. We aimed to develop an evidence-based scale-derived set of criteria to define relapse in patients with schizophrenia or schizoaffective disorder. We searched the Yale University Open Data Access (YODA) for randomised controlled trials (RCTs) in clinically stable adults with schizophrenia or schizoaffective disorder, and obtained individual participant data on Positive and Negative Syndrome Scale (PANSS), Clinical Global Impression Severity (CGI-S), Personal and Social Performance (PSP), and Social and Occupational Functioning Assessment Scale (SOFAS). Our main outcomes were PANSS-derived criteria based on worsening in PANSS total score. We examined their relevance using equipercentile linking with CGI-S and functioning scales, and their test-performance in defining relapse with diagnostic test accuracy meta-analysis against CGI-S worsening (≥1-point increase together with a score ≥4 points) and psychiatric hospitalisation. Based on data from seven RCTs (2354 participants; 1348 men [57·3%] and 1006 women [42·7%], mean age of 39·5 years [SD 12·0, range 17-89]; 303 Asian [12.9%], 255 Black [10.8%], 1665 White [70.7%], and other or unspecified 131 [5.6%]), an increase of 12 points or more in PANSS total (range 30-210 points) corresponded to clinically important deterioration in global severity of illness (≥1 point increase in CGI-S, range 1-7) and functioning (≥10 points decline in PSP or SOFAS, range 1-100). The interpretation of percentage changes varied importantly across different baseline scores. An increase of 12 points or more in PANSS total had good sensitivity and specificity using CGI-S as reference standard (sensitivity 82·1% [95% CI 77·1-86·4], specificity 86·9% [82·9-90·3]), as well as good sensitivity but lower specificity compared to hospitalisation (sensitivity 81·7% [74·1-87·7], specificity 69·2% [60·5-76·9]). Requiring either an increase in PANSS total or in specific items for positive and disorganization symptoms further improved test-performance. Cutoffs for situations where high sensitivity or specificity is needed are presented. An increase of either 12 points or more in the PANSS total score, or worsening of specific positive and disorganisation symptom items could be a reasonable evidence-based definition of relapse in schizophrenia, potentially linking symptoms used to define remission and relapse. Percentage changes should not be used to define relapse because their interpretation depends on baseline scores. German Research Foundation (grant number: 428509362).

High Rates of Thrombotic Events in Newly Diagnosed Multiple Myeloma Patients Enrolled on the Maia Trial

Background. Patients with newly diagnosed multiple myeloma (NDMM) have a nine-fold increased risk of developing venous thromboembolism (VTE). Despite the availability of the IMWG and NCCN VTE prophylaxis guidelines and updated risk stratification models such as SAVED and IMPEDE, the optimal thromboprophylaxis strategy is unknown, especially in the age of modern daratumumab-containing regimens. To continue highlighting the importance of appropriate thromboprophylaxis for patients with NDMM, we performed a post-hoc analysis of the randomized phase 3 MAIA trial to identify rates of venous and arterial thromboembolism. Methods. This retrospective study, carried out under YODA Project #2021-4663, used data from the randomized phase 3 MAIA trial that investigated lenalidomide and dexamethasone (Rd) with or without daratumumab (DRd) in 733 transplant ineligible (TIE) patients obtained from the Yale University Open Data Access (YODA) Project, which has an agreement with JANSSEN RESEARCH & DEVELOPMENT, L.L.C.. Thrombotic events (TE), including both venous and arterial events (V/ATE) were examined with VTE events including deep vein thrombosis (DVT), pulmonary embolism (PE), and superficial thrombophlebitis (SVT). Arterial events (ATE) included cerebrovascular accidents (CVA) defined as hemorrhagic, ischemic, or verebrobasilar strokes. The incidence and time to event of V/ATE were calculated and graded using the CTCAE v4, with Grade 1 representing nonactionable venous thrombosis, Grades 2-4 representing TEs requiring increasing higher levels of medical intervention, and Grade 5 indicating death was a result of the TE. Rates and time to events were compared between the DRd and Rd arms. Results. Of the included patients, median age was 72 years (range 64-72) and 48% were female. There was no signficant difference between baseline characteristics of sex, race, myeloma subtype, presence of high-risk cytogenetics, or ISS staging between patients that experienced a V/ATE event and those that did not. All grade V/ATE events were 18% (66/363) and 19% (70/365) in the DRd and Rd arms, respectively. Of these, most were VTE events in both arms (16%), with G2-4 actionable VTE event rate of 14.9% vs 14.8% for the DRd and Rd arms, respectively. G5 VTE events were 0.2% in both arms (no G5 ATEs). ATE were 2.2% and 2.7% in the DRd and Rd group, respectively, overall constituting a minority of events. The overall PE rate was higher in the DRd than Rd arm, 6.3% vs 4.7%, DVT was more frequent in the Rd arm, 9.6% vs 7.7%, and rates of SVT rates were similar between both arms at 1.9% and 2.2% for the DRd vs Rd arms, respectively (Table 1). Median time to onset of cumulative TEs was longer in DRd vs Rd arm, 7.15 (0.27 - 63.0) vs 4.65 (0.03 - 58.0) months, respectively. Among these, median time to onset of VTE was similar among the groups [6.57 (0.27 - 63.0) vs 4.48 (0.03 - 52.9) months, respectively] while the median time to onset of ATEs was longer in DRd vs Rd arm, 19.8 (1.70 - 26.6) vs 8.3 (0.67 - 58.5) months. The incidence of VTE events persisted past 24 months and incidence of ATE events increased over time (Figure 1). Analysis regarding prophylactic anticoagulation use and rates of TE based on prophylactic regimen are ongoing. Conclusion. This retrospective study reveals the highest reported rates of VTE in any modern therapy dataset. In patients with transplant ineligible NDMM, we show that the TE risk persists past 24 months of treatment in patients on DRd or Rd therapy. Similar to observations in the GRIFFIN trial, the addition of daratumumab did not increase the risk of TE. Our report highlights the critical need for further investigation of optimal prophylactic strategies and their duration for patients with NDMM.

Correlations between comorbidities in trials and the community: An individual-level participant data meta-analysis.

People with comorbidities are under-represented in randomised controlled trials, and it is unknown whether patterns of comorbidity are similar in trials and the community. Individual-level participant data were obtained for 83 clinical trials (54,688 participants) for 16 index conditions from two trial repositories: Yale University Open Data Access (YODA) and the Centre for Global Clinical Research Data (Vivli). Community data (860,177 individuals) were extracted from the Secure Anonymised Information Linkage (SAIL) databank for the same index conditions. Comorbidities were defined using concomitant medications. For each index condition, we estimated correlations between comorbidities separately in trials and community data. For the six commonest comorbidities we estimated all pairwise correlations using Bayesian multivariate probit models, conditioning on age and sex. Correlation estimates from trials with the same index condition were combined into a single estimate. We then compared the trial and community estimates for each index condition. Despite a higher prevalence of comorbidities in the community than in trials, the correlations between comorbidities were mostly similar in both settings. On comparing correlations between the community and trials, 21% of correlations were stronger in the community, 10% were stronger in the trials and 68% were similar in both. In the community, 5% of correlations were negative, 21% were null, 56% were weakly positive and 18% were strongly positive. Equivalent results for the trials were 11%, 33%, 45% and 10% respectively. Comorbidity correlations are generally similar in both the trials and community, providing some evidence for the reporting of comorbidity-specific findings from clinical trials.

The Yale University Open Data Access Project: A Resource for Research in Psychopharmacology.

The Yale University Open Data Access Project: A Resource for Research in Psychopharmacology.

Predicting psychotic relapse following randomised discontinuation of paliperidone in individuals with schizophrenia or schizoaffective disorder: an individual participant data analysis

Predicting relapse for individuals with psychotic disorders is not well established, especially after discontinuation of antipsychotic treatment. We aimed to identify general prognostic factors of relapse for all participants (irrespective of treatment continuation or discontinuation) and specific predictors of relapse for treatment discontinuation, using machine learning. For this individual participant data analysis, we searched the Yale University Open Data Access Project's database for placebo-controlled, randomised antipsychotic discontinuation trials with participants with schizophrenia or schizoaffective disorder (aged ≥18 years). We included studies in which participants were treated with any antipsychotic study drug and randomly assigned to continue the same antipsychotic drug or to discontinue it and receive placebo. We assessed 36 prespecified baseline variables at randomisation to predict time to relapse, using univariate and multivariate proportional hazard regression models (including multivariate treatment group by variable interactions) with machine learning to categorise the variables as general prognostic factors of relapse, specific predictors of relapse, or both. We identified 414 trials, of which five trials with 700 participants (304 [43%] women and 396 [57%] men) were eligible for the continuation group and 692 participants (292 [42%] women and 400 [58%] men) were eligible for the discontinuation group (median age 37 [IQR 28-47] years for continuation group and 38 [28-47] years for discontinuation group). Out of the 36 baseline variables, general prognostic factors of increased risk of relapse for all participants were drug-positive urine; paranoid, disorganised, and undifferentiated types of schizophrenia (lower risk for schizoaffective disorder); psychiatric and neurological adverse events; higher severity of akathisia (ie, difficulty or inability to sit still); antipsychotic discontinuation; lower social performance; younger age; lower glomerular filtration rate; benzodiazepine comedication (lower risk for anti-epileptic comedication). Out of the 36 baseline variables, predictors of increased risk specifically after antipsychotic discontinuation were increased prolactin concentration, higher number of hospitalisations, and smoking. Both prognostic factors and predictors with increased risk after discontinuation were oral antipsychotic treatment (lower risk for long-acting injectables), higher last dosage of the antipsychotic study drug, shorter duration of antipsychotic treatment, and higher score on the Clinical Global Impression (CGI) severity scale The predictive performance (concordance index) for participants who were not used to train the model was 0·707 (chance level is 0·5). Routinely available general prognostic factors of psychotic relapse and predictors specific for treatment discontinuation could be used to support personalised treatment. Abrupt discontinuation of higher dosages of oral antipsychotics, especially for individuals with recurring hospitalisations, higher scores on the CGI severity scale, and increased prolactin concentrations, should be avoided to reduce the risk of relapse. German Research Foundation and Berlin Institute of Health.

External control arm analysis: an evaluation of propensity score approaches, G-computation, and doubly debiased machine learning

BackgroundAn external control arm is a cohort of control patients that are collected from data external to a single-arm trial. To provide an unbiased estimation of efficacy, the clinical profiles of patients from single and external arms should be aligned, typically using propensity score approaches. There are alternative approaches to infer efficacy based on comparisons between outcomes of single-arm patients and machine-learning predictions of control patient outcomes. These methods include G-computation and Doubly Debiased Machine Learning (DDML) and their evaluation for External Control Arms (ECA) analysis is insufficient.MethodsWe consider both numerical simulations and a trial replication procedure to evaluate the different statistical approaches: propensity score matching, Inverse Probability of Treatment Weighting (IPTW), G-computation, and DDML. The replication study relies on five type 2 diabetes randomized clinical trials granted by the Yale University Open Data Access (YODA) project. From the pool of five trials, observational experiments are artificially built by replacing a control arm from one trial by an arm originating from another trial and containing similarly-treated patients.ResultsAmong the different statistical approaches, numerical simulations show that DDML has the smallest bias followed by G-computation. In terms of mean squared error, G-computation usually minimizes mean squared error. Compared to other methods, DDML has varying Mean Squared Error performances that improves with increasing sample sizes. For hypothesis testing, all methods control type I error and DDML is the most conservative. G-computation is the best method in terms of statistical power, and DDML has comparable power at n=1000 but inferior ones for smaller sample sizes. The replication procedure also indicates that G-computation minimizes mean squared error whereas DDML has intermediate performances in between G-computation and propensity score approaches. The confidence intervals of G-computation are the narrowest whereas confidence intervals obtained with DDML are the widest for small sample sizes, which confirms its conservative nature.ConclusionsFor external control arm analyses, methods based on outcome prediction models can reduce estimation error and increase statistical power compared to propensity score approaches.

Depicting Risperidone Safety Profiles in Clinical Trials Across Different Diagnoses Using a Dopamine D2-Based Pharmacological Class Effect Query Defined by FAERS.

Current methods are not designed to relate the incidence of individual adverse events reported in clinical trials to the plurality of adverse events accumulated in spontaneous reporting databases during real-world use. We have previously reported on a pharmacological class-effect query of clinical trial data defined by a disproportionality analysis of the US Food and Drug Administration Adverse Event Reporting System (FAERS) post-marketing data. The aim of the current analysis was to apply a dopamine D2-based pharmacological class-effect query to clinical trial safety data of an atypical antipsychotic tested across different patient populations. Patient-level adverse event data (n = 4400) from controlled clinical trials of the antipsychotic risperidone in schizophrenia, bipolar disorder, Alzheimer's disease psychosis, and autism were obtained through the Yale University Open Data Access (YODA) project. An Empirical Bayes Geometric Mean analysis was performed, and a three-fold threshold incidence level was applied to determine if a preferred term met criteria for being an antipsychotic class-related adverse event. In pooled data from seven trials of adult schizophrenia, class-specific adverse events were identified in 49% of patients treated with risperidone; in 49% of risperidone-treated patients in two trials in adolescent schizophrenia; in 65% of risperidone-treated patients in four trials in adult bipolar disorder; in 50% of risperidone-treated patients in two trials in adolescent schizophrenia; in 36% of risperidone-treated patients in one trial in Alzheimer's disease; and in 94% of risperidone-treated patients in one trial in autism. The cumulative curves of class-specific adverse events in risperidone clinical trials of schizophrenia were similar to those first reported for other atypical antipsychotic drugs. However, the class-specific adverse event curves were notably lower for Alzheimer's disease and higher for autism, suggesting that the diagnostic indication may have an important effect on the cumulative class-specific side-effect burden.

Vibration of effect in more than 16 000 pooled analyses of individual participant data from 12 randomised controlled trials comparing canagliflozin and placebo for type 2 diabetes mellitus: multiverse analysis

ObjectiveTo evaluate the impact of conducting all possible pooled analyses across different combinations of randomised controlled trials and endpoints.DesignMultiverse analysis, consisting of numerous pooled analyses of individual participant data.SettingIndividual patient...

Effect of age, sex, and morbidity count on trial attrition: meta-analysis of individual participant level data from phase 3/4 industry funded clinical trials

ObjectivesTo estimate the association between individual participant characteristics and attrition from randomised controlled trials.DesignMeta-analysis of individual participant level data (IPD).Data sourcesClinical trial repositories (Clinical Study Data Request and Yale University...

Reappraising the variability of effects of antipsychotic medication in schizophrenia: a meta-analysis.

It is common experience for practising psychiatrists that individuals with schizophrenia vary markedly in their symptomatic response to antipsychotic medication. What is not clear, however, is whether this variation reflects variability of medication-specific effects (also called "treatment effect heterogeneity"), as opposed to variability of non-specific effects such as natural symptom fluctuation or placebo response. Previous meta-analyses found no evidence of treatment effect heterogeneity, suggesting that a "one size fits all" approach may be appropriate and that efforts at developing personalized treatment strategies for schizophrenia are unlikely to succeed. Recent advances indicate, however, that earlier approaches may have been unable to accurately quantify treatment effect heterogeneity due to their neglect of a key parameter: the correlation between placebo response and medication-specific effects. In the present paper, we address this shortcoming by using individual patient data and study-level data to estimate that correlation and quantitatively characterize antipsychotic treatment effect heterogeneity in schizophrenia. Individual patient data (on 384 individuals who were administered antipsychotic treatment and 88 who received placebo) were obtained from the Yale University Open Data Access (YODA) database. Study-level data were obtained from a meta-analysis of 66 clinical trials including 17,202 patients. Both individual patient and study-level analyses yielded a negative correlation between placebo response and treatment effect for the total score on the Positive and Negative Syndrome Scale (PANSS) (ρ=-0.32, p=0.002 and ρ=-0.39, p<0.001, respectively). Using the most conservative of these estimates, a meta-analysis of treatment effect heterogeneity provided evidence of a marked variability in antipsychotic-specific effects between individuals with schizophrenia, with the top quartile of patients experiencing beneficial treatment effects of 17.7 points or more on the PANSS total score, while the bottom quartile presented a detrimental effect of treatment relative to placebo. This evidence of clinically meaningful treatment effect heterogeneity suggests that efforts to personalize antipsychotic treatment of schizophrenia have potential for success.

Gender-Based Differences in Response to Tumor Necrosis Factor Inhibitor Therapies for Ulcerative Colitis: Individual Participant Data Meta-Analyses of Clinical Trials.

Gender-based differences are reported in inflammatory bowel diseases (IBD) pathogenesis, but their impacts on IBD outcomes are not well known. We determined gender-based differences in response to treatment with tumor necrosis factor inhibitor (TNFi) therapies in individuals with ulcerative colitis (UC). We used the Yale University Open Data Access (YODA) platform to abstract individual participant data from randomized clinical trials to study infliximab and golimumab as induction and maintenance therapies in moderately to severely active UC. Using multivariable logistic regression, we examined associations between gender and the endpoints of clinical remission, mucosal healing, and clinical response for each study individually and in a meta-analysis. Of 1639 patients included in induction trials (Program of Ulcerative Colitis Research Studies Utilizing an Investigational Treatment-Subcutaneous [PURSUIT-SC], active ulcerative colitis trials [ACT] 1 and 2) and 1280 patients included in maintenance trials (Program of Ulcerative Colitis Research Studies Utilizing an Investigational Treatment-Maintenance [PURSUIT-IM], ACT 1 and 2), 696 (42.5%) and 534 (41.7%) were women, respectively. In a meta-analysis of induction trials, the adjusted odds ratios (aORs) of clinical remission (aOR, 0.55; 95% CI, 0.31-0.97), mucosal healing (aOR, 0.47; 95% CI, 0.27-0.83), and clinical response (aOR, 0.51; 95% CI, 0.29-0.90) in the treatment arm and of clinical remission in the placebo arm (aOR, 0.34; 95% CI, 0.15-0.82) were lower in men compared to women. There were no differences in outcomes by gender in the treatment and placebo arms in the meta-analysis of maintenance trials. Men are less likely to achieve clinical remission, mucosal healing, and clinical response compared to women during induction treatment with TNFi for UC, but not during the maintenance phase. Future studies delineating the mechanisms underlying these observations would be informative.

Adverse events after antipsychotic discontinuation: an individual participant data meta-analysis

Adverse events can occur after antipsychotic discontinuation but evidence from antipsychotic drug trials is scarce. We aimed to estimate the occurrence of adverse events after discontinuing antipsychotics. For this two-stage individual participant data meta-analysis, we searched the Yale University Open Data Access Project's database for randomised controlled trials of antipsychotics from database inception until May 6, 2021. We included placebo-controlled antipsychotic randomised controlled trials with individual participant data of participants (aged ≥ 18 years, of any sex and ethnicity) with schizophrenia, schizoaffective disorder, or bipolar disorder. Studies were excluded if treatment with antidepressants, lithium, or antiepileptic drugs was initiated as additive therapy at the start of the placebo phase. Starting from the screening or washout phase, we divided participants who were randomised to placebo into two groups: the discontinuation group (participants who discontinued prestudy antipsychotics at the start of the screening or washout phase) and control group (participants who did not take prestudy antipsychotics for at least 4 weeks before the start of the screening or washout phase). Participants were excluded from the discontinuation and control groups if they discontinued prestudy treatment with antidepressants, lithium, or antiepileptic drugs up to 4 weeks before baseline, received an antipsychotic as a tolerability test, or received a long-acting injection of an antipsychotic within 12 weeks before baseline. In the discontinuation group, individuals were excluded if they discontinued prestudy antipsychotic treatment more than 3 days before, or any day after, the start of screening or washout phase. The prespecified primary outcome was occurrence of at least one new somatic adverse event with an onset within 4 weeks after the start of the screening or washout phase. We implemented a generalised linear model that accounted for potential confounders, to estimate the effect of antipsychotic discontinuation. This study is registered with PROSPERO (CRD42021224350). We identified 409 records of which 18 were eligible and included in the analysis. From these 18 studies, 692 individuals (242 [35·0%] women and 450 [65·0%] men) were eligible for the discontinuation group and 935 individuals (339 [36·3%] women and 596 [63·7%] men) were eligible for the control group (median age in both groups: 39 years [IQR 30-47]). New somatic adverse events occurred in 295 (43%) individuals in the discontinuation group and 293 (31%) individuals in the control group (OR 1·74; 95% CI 1·27-2·39; τ2=0·15; moderate strength of evidence). New psychiatric adverse events were also more frequent in the discontinuation group than the control group (OR 2·01; 95% CI 1·38-2·94). Longer duration of treatment before discontinuation (OR for doubling the duration of treatment: 1·08; 95% CI 1·01-1·14) was associated with a higher probability of new somatic adverse events after antipsychotic discontinuation, and tapered discontinuation (compared with abrupt discontinuation: 0·54; 0·32-0·91) and no history of somatic illness (compared with history of somatic illness: 0·63; 0·43-0·91) were associated with lower probabilities of new somatic adverse events after antipsychotic discontinuation. The risk of bias was moderate in 13 (72·2%) studies and serious in five (27·8%) studies. We detected moderate evidence of emerging somatic adverse events after discontinuation of first-generation and second-generation antipsychotics, particularly after discontinuation of longer durations of treatment. Tapered discontinuation can mitigate the risk of emerging somatic adverse events after antipsychotic discontinuation. These findings have implications for the safety of treatment discontinuation and could be used for tailored treatment planning. German Research Foundation.

Benefits and harms of Risperidone and Paliperidone for treatment of patients with schizophrenia or bipolar disorder: a meta-analysis involving individual participant data and clinical study reports

BackgroundSchizophrenia and bipolar disorder are severe mental illnesses which are highly prevalent worldwide. Risperidone and Paliperidone are treatments for either illnesses, but their efficacy compared to other antipsychotics and growing reports of hormonal imbalances continue to raise concerns. As existing evidence on both antipsychotics are solely based on aggregate data, we aimed to assess the benefits and harms of Risperidone and Paliperidone in the treatment of patients with schizophrenia or bipolar disorder, using individual participant data (IPD), clinical study reports (CSRs) and publicly available sources (journal publications and trial registries).MethodsWe searched MEDLINE, Central, EMBASE and PsycINFO until December 2020 for randomised placebo-controlled trials of Risperidone, Paliperidone or Paliperidone palmitate in patients with schizophrenia or bipolar disorder. We obtained IPD and CSRs from the Yale University Open Data Access project. The primary outcome Positive and Negative Syndrome Scale (PANSS) score was analysed using one-stage IPD meta-analysis. Random-effect meta-analysis of harm outcomes involved methods for coping with rare events. Effect-sizes were compared across all available data sources using the ratio of means or relative risk. We registered our review on PROSPERO, CRD42019140556.ResultsOf the 35 studies, IPD meta-analysis involving 22 (63%) studies showed a significant clinical reduction in the PANSS in patients receiving Risperidone (mean difference − 5.83, 95% CI − 10.79 to − 0.87, I2 = 8.5%, n = 4 studies, 1131 participants), Paliperidone (− 6.01, 95% CI − 8.7 to − 3.32, I2 = 4.3%, n = 13, 3821) and Paliperidone palmitate (− 7.89, 95% CI − 12.1 to − 3.69, I2 = 2.9%, n = 5, 2209). CSRs reported nearly two times more adverse events (4434 vs. 2296 publication, relative difference (RD) = 1.93, 95% CI 1.86 to 2.00) and almost 8 times more serious adverse events (650 vs. 82; RD = 7.93, 95% CI 6.32 to 9.95) than the journal publications. Meta-analyses of individual harms from CSRs revealed a significant increased risk among several outcomes including extrapyramidal disorder, tardive dyskinesia and increased weight. But the ratio of relative risk between the different data sources was not significant. Three treatment-related gynecomastia events occurred, and these were considered mild to moderate in severity.ConclusionIPD meta-analysis conclude that Risperidone and Paliperidone antipsychotics had a small beneficial effect on reducing PANSS score over 9 weeks, which is more conservative than estimates from reviews based on journal publications. CSRs also contained significantly more data on harms that were unavailable in journal publications or trial registries. Sharing of IPD and CSRs are necessary when performing meta-analysis on the efficacy and safety of antipsychotics.

P433 Sex-based differences in response to tumor necrosis factor inhibitor induction therapy for ulcerative colitis: a pooled analysis of individual patient-level clinical trials data

Abstract Background There are increasing data on sex-based differences in IBD epidemiology, phenotype and outcomes. The purpose of this study was to characterize sex-based differences in response to tumor necrosis factor inhibitor (TNFi) therapies in patients with ulcerative colitis (UC). Methods We conducted a pooled analysis of individual-level data from randomized clinical trials (RCTs) on induction therapy with infliximab and golimumab, accessed through the Yale University Open Data Access platform. We analyzed patients in the treatment and placebo arms separately. Using multivariable logistic regression modeling, we compared the primary outcome of clinical remission, and secondary outcomes of clinical response and mucosal healing, according to clinical trial definitions, between male and female patients with UC at week 6 (golimumab) or week 8 (infliximab) of induction therapy. Effect estimates were expressed as adjusted odds ratios (aOR) and 95% confidence intervals (CI). Results We included 1639 patients (696, 42.5% were women, Table) from two trials using infliximab (ACT-1 and 2) and one trial using golimumab (PURSUIT). Male patients were older than female patients in the treatment arm of the overall cohort [mean age (standard deviation) 31.31 (16.53) years and 29.34 (15.19) years, respectively). All other baseline characteristics, such as sex, race, Mayo score at screening, and concomitant systemic corticosteroid and immunomodulator use were comparable. Compared to female patients, male patients were significantly less likely to achieve clinical remission in both treatment and placebo arms (aOR 0.55, 95% CI 0.31, 0.97 and 0.34, 95% CI 0.15, 0.82, respectively, Figure). Compared to female patients, male patients were significantly less likely to achieve mucosal healing and clinical response in the treatment arm (aOR 0.47, 95% CI 0.27, 0.83 and 0.51, 95% CI 0.29, 0.90, respectively), but there was no significant difference in these secondary outcomes by sex in the placebo arm. Conclusion Based on this analysis of pooled data from RCTs, we demonstrate that clinical remission, response and mucosal healing with TNFi induction therapies are less likely in male patients compared to female patients with UC. Clinical remission is also less likely in male patients on placebo. These finding may lead to further research on sex-based differences in treatment outcomes, including mechanistic studies.

DOP58 Non-white race is associated with decreased efficacy of tumor necrosis factor antagonist therapy in Ulcerative Colitis: A post-hoc analysis of individual level data from golimumab clinical trials

Abstract Background Observational studies suggest non-white patients with inflammatory bowel disease (IBD) have worse clinical outcomes.1 There are limited data on whether race impacts response to biologic therapy. We therefore aimed to evaluate the efficacy of the tumor necrosis factor (TNF) antagonist golimumab comparing white to non-white participants, using individual participant level data from phase 2/3 randomized clinical trials of TNF antagonist therapy in ulcerative colitis (UC). Methods We conducted a pooled analysis of individual-level data from the induction and maintenance trials of golimumab in UC accessible through Yale University Open Data Access Project (YODA). There were insufficient non-white participants in infliximab studies accessible through YODA, precluding meaningful analysis. We analyzed patients in the placebo and treatment arms separately. Our primary outcome was clinical response and secondary outcomes were clinical remission and endoscopic healing according to clinical trial definitions. We compared white and non-white (defined as Black, Asian, or Other race) participants using multivariable logistic regression a priori adjusting for age, sex, treatment group, baseline Mayo score, immunomodulator and corticosteroid use. Effect estimates were expressed as adjusted odds ratios (aOR) and 95% confidence intervals (95% CI). Results A total of 1,006 participants were included in the induction trial (PURSUIT-SC; 18% non-white) and 783 participants in the maintenance trial (PURSUIT-M; 17% non-white). Non-white participants had significantly lower odds of week 6 clinical response (aOR 0.43, 95%CI 0.28–0.66), clinical remission (aOR 0.41, 95%CI 0.22–0.77) and endoscopic remission (aOR 0.48, 95%CI 0.30–0.74) compared to white participants (Figure). Non-white participants also had a lower adjusted odds of week 30 clinical response (aOR 0.64, 95%CI 0.40–1.01), clinical remission (aOR 0.45, 95%CI 0.28–0.74), and endoscopic remission (aOR 0.62, 95%CI 0.41–0.96). By week 54, the lower odds of these outcomes among non-whites were no longer statistically significant (Figure). Conclusion Non-white UC patients were less likely to achieve clinical response, clinical remission, and endoscopic healing with golimumab compared to white patients in these clinical trials. Further studies are needed to understand these differences and whether they are observed with other drugs or outside the context of clinical trials. Reference

Access to data from clinical trials in the COVID-19 crisis: open, flexible, and time-sensitive

- There is a great need to swiftly develop vaccines and treatments against COVID-19 in the current pandemic. Yet, the time pressure and difficult circumstances to conduct clinical trials increases the risk of incomplete data sets, suboptimal study designs, and thus spurious results and reversals of decisions. - Open access to clinical trial and other patient-relevant data is necessary in order to provide transparency, and the possibility for validating clinical-trial analyses, and fully exploiting the information emerging from clinical trials. Thus, open access is critical in order to enable the successful and swift development of an effective intervention.

Neutropenia in Myelosuppressed Cancer Patients Treated with Recombinant Human Erythropoietin (EPO)

Recombinant human erythropoietin (EPO) has been used extensively for 3 decades to treat various forms of anemia, including anemia associated with cancer treatment. Multiple randomized, placebo controlled clinical trials conducted since the late 1980's found that EPO treatment ameliorated anemia, reduced transfusion requirements, and improved quality of life measures in cancer patients receiving myelosuppressive chemotherapy. However, results of several of these trials suggested that EPO treatment might also exacerbate neutropenia in these patients. This possible side effect had been anticipated by a pre-clinical observation that neutrophil production in long-term marrow cultures was suppressed by EPO. However, chemotherapy-induced neutropenia was expected in these clinical trials and not studied, nor reported, in detail. Recently, primary data from 5 prospective randomized, placebo controlled clinical trials of EPO treatment in cancer patients undergoing myelosuppressive chemotherapy, sponsored by Johnson &amp; Johnson, became available through the Yale Open Data Access (YODA) project*. These data allowed us to compare serial neutrophil counts (ANCs) from 1650 study subjects with various forms of cancer who participated in these 5 clinical trials and were randomly assigned to either EPO or placebo treatment. Individual clinical trial sample sizes varied from 71 to 922 (Table below). ANC values were recorded weekly in 2 trials but less frequently in the other 3. In our analysis we determined the percent of study subjects with recorded ANC values &lt;1500/mm3, &lt;500/mm3, and &lt;200/mm3 during the initial 12 weeks of each trial for individuals assigned either to placebo or to EPO treatment. Mantel-Haenszel odds ratios were used as measures of differences between treatment groups. In addition, Poisson regression was used to assess differences in the total numbers of ANC measurements &lt;500/mm3 per study subject per 12 weeks on study that had been recorded in each treatment group. In each trial more study subjects assigned to EPO than to placebo treatment were found to have had at least one ANC &lt;500/mm3 reflecting severe neutropenia (Table). Study-specific odds ratios for this difference ranged from 1.24 to 2.08 and the Mantel-Haenszel odds ratio for the combined studies was 1.48 with a 95% Confidence Interval (CI) of 1.08 - 2.03, p =0.014. A statistically significant odds ratio was also found for ANC values of &lt;200/mm3 (1.73, 1.10-2.70, p=0.016), but not for neutropenic ANC values overall, i.e. &lt;1500/mm3 (1.18, 0.93-1.50, p=0.17). Total numbers of recorded ANC values &lt;500/mm3 per study subject per 12 weeks on study were also significantly different - 0.66 (95% CI, 0.57-0.76) in the EPO treatment group vs. 0.49 (0.42-0.60) in the placebo group, p=0.0075. This analysis of 5 randomized placebo controlled clinical trials carried out from 1988 to 2002 finds that EPO treatment of individuals undergoing myelosuppressive chemotherapy was associated with statistically significant increases in the incidence of severe neutropenia and suggests that competition between the erythroid and myeloid lineages may occur in the setting of suppressed hematopoietic reserves when the erythroid lineage is pharmacologically amplified. *This study, carried out under YODA Project 2017-2486, used data obtained from the Yale University Open Data Access Project, which has an agreement with JANSSEN RESEARCH &amp; DEVELOPMENT, L.L.C.. The interpretation and reporting of research using this data are solely the responsibility of the authors and does not necessarily represent the official views of the Yale University Open Data Access Project or JANSSEN RESEARCH &amp; DEVELOPMENT, L.L.C.. Table Disclosures No relevant conflicts of interest to declare.