Estimating interactions and subgroup-specific treatment effects in meta-analysis without aggregation bias: A within-trial framework.

BackgroundChronic low back pain (LBP) is the leading cause of pain and disability, resulting in a major socioeconomic impact. The Cochrane Review which examined the effect of spinal manipulative therapy (SMT) for chronic LBP concluded that SMT is moderately effective, but was based on conventional meta-analysis of aggregate data. The use of individual participant data (IPD) from trials allows for a more precise estimate of the treatment effect and has the potential to identify moderators and/or mediators. The aim is (1) to assess the overall treatment effect of SMT for primary and secondary outcomes in adults with chronic LBP, (2) to determine possible moderation of baseline characteristics on treatment effect, (3) to identify characteristics of intervention (e.g., manipulation/mobilization) that influence the treatment effect, and (4) to identify mediators of treatment effects.MethodsAll trials included in the Cochrane Review on SMT for chronic LBP will be included which were published after the year 2000, and the search will be updated. No restrictions will be placed on the type of comparison or size of the study. Primary outcomes are pain intensity and physical functioning. A dataset will be compiled consisting of individual trials and variables included according to a predefined coding scheme. Variables to be included are descriptive of characteristics of the study, treatment, comparison, participant characteristics, and outcomes at all follow-up periods. A one-stage approach with a mixed model technique based on the intention-to-treat principle will be used for the analysis. Subsequent analyses will focus on treatment effect moderators and mediators.DiscussionWe will analyze IPD for LBP trials in which SMT is one of the interventions. IPD meta-analysis has been shown to be more reliable and valid than aggregate data meta-analysis, although this difference might also be attributed to the number of studies that can be used or the amount of data that can be utilized. Therefore, this project may identify important gaps in our knowledge with respect to prognostic factors of treatment effects.Systematic review registration:PROSPERO CRD42015025714

41 Background: Surrogate endpoints of overall survival (OS) are increasingly used as the basis of cancer drug approvals, but a strong association between their treatment effects is essential for validation as appropriate surrogates. We summarize the current evidence landscape for meta-analytic validation of surrogate endpoints in colorectal cancer (CRC) and assess surrogacy evaluation quality through a novel framework to inform the selection of primary endpoints in future trials. Methods: We developed a framework to grade the quality of surrogate endpoints in oncology. We then searched MEDLINE to identify meta-analytic validation studies of endpoints in CRC at the trial level. We consulted a research librarian to develop a search strategy in PubMed using keywords and MeSH terms. Articles published before 07/06/2024 were included and two reviewers independently evaluated each. Grades were assessed according to seven evidence domains: data source (A-B), disease population (0-3), surrogate and true endpoint definitions (0-3), number of trials evaluated (0-4), quality and strength of individual (0-2) and trial-level associations (0-8). Discordance in blinded assessments was resolved by consensus decision. Results: Eighteen articles were identified containing 39 evaluations of 12 unique endpoints. Results are presented for disease-free survival (DFS), Progression-free survival (PFS), overall response rate (ORR), time to progression (TTP) and disease control rate (DCR). PFS was most frequently assessed (n=11). Surrogates were primarily evaluated in stage IV disease (n=34) and analyses of biologic agents (n=29). Individual patient data (IPD) was used in 13 evaluations. A23 was the highest endpoint quality score assigned using IPD and B6 was the lowest using summary trial data (SD). Only DFS consistently demonstrated high-quality surrogacy for OS at the trial level using IPD (≥A18, R 2 ≥ 0.70). Conclusions: DFS consistently shows high quality of surrogacy, whereas other endpoints show inconsistent surrogacy evaluation ranging from low to moderate quality in contemporary CRC trials. Standardizing the methodology of endpoint validation using IPD may improve surrogacy estimates. Assessment of new, or revision of prior surrogate endpoints may be warranted, particularly in the context of metastatic disease. Endpoint Data source (frequency) Treatment setting (frequency) Treatment era (frequency) Trial-level association, R 2 (range) Surrogate quality score (range) DFS (n=5) IPD: 5 Adjuvant Pre-biologics: 4Post-biologics: 1 0.70-0.93 A18-23 PFS (n=11) IPD: 3SD: 8 Metastatic Pre-biologics: 2Post-biologics: 9 0.34-0.99 B6-A22 ORR (n=9) IPD: 1SD: 8 Metastatic Pre-biologics: 1Post-biologics: 8 0.03-0.74 B6-A13 TTP (n=5) SD: 5 Metastatic Post-biologics: 5 0.27-0.65 B7-11 DCR (n=2) SD: 2 Metastatic Post-biologics: 2 0.82-0.89 B7-8

Protocol for balanced versus saline trialists: living systematic review and individual patient data meta-analysis of randomised controlled trials (BEST-Living study)

In this study, we compared the efficacy of the overall meta-analysis estimates that used only the available aggregate data (AD) studies against those that combined the available AD and individual patient data (IPD) studies. We introduced some modifications to the existing two-stage method for combining the AD and IPD studies. We evaluated the effects of these modifications on the estimates of the overall treatment effect, and investigated the influence of the number of studies included in the meta-analysis, N, and the ratio of AD: IPD on these estimates. We used percentage relative bias (PRB), root mean-square-error (RMSE), and coverage probability to assess the overall efficiency of these estimates. The results revealed the superiority of estimates from the combined AD: IPD studies over those that utilized only the available AD in terms of both the accuracy and the RMSE. We found that the current method for combining the AD:IPD studies provided poor coverage probabilityand that the proposed methods generated improved coverage probability by more than 40% while maintaining the level of bias and RMSE at par to their existing counterparts. These findings validated the importance of utilizing both the AD and IPD studies whenever they are available, and demonstrated the significance of proper technique for combining these studies in order to obtain better overall estimates.

BackgroundMatching Adjusted Indirect Comparison (MAIC) is a statistical method used to adjust for potential biases when comparing treatment effects between separate data sources, with aggregate data in one arm, and individual patients data in the other. However, acceptance of MAIC in health technology assessment (HTA) is challenging because of the numerous biases that can affect the estimates of treatment effects – especially with small sample sizes, increasing the risk of convergence issues. We suggest statistical approaches to address some of the challenges in supporting evidence from MAICs, applied to a case study.MethodsThe proposed approaches were illustrated with a case study comparing an integrated analysis of three single-arm trials of entrectinib with the French standard of care using the Epidemio-Strategy and Medical Economics (ESME) Lung Cancer Data Platform, in metastatic ROS1-positive Non-Small Cell Lung Cancer (NSCLC) patients. To obtain convergent models with balanced treatment arms, a transparent predefined workflow for variable selection in the propensity score model, with multiple imputation of missing data, was used. To assess robustness, multiple sensitivity analyses were conducted, including Quantitative Bias Analyses (QBA) for unmeasured confounders (E-value, bias plot), and for missing at random assumption (tipping-point analysis).ResultsThe proposed workflow was successful in generating satisfactory models for all sub-populations, that is, without convergence problems and with effectively balanced key covariates between treatment arms. It also gave an indication of the number of models tested. Sensitivity analyses confirmed the robustness of the results, including to unmeasured confounders. The QBA performed on the missing data allowed to exclude the potential impact of the missing data on the estimate of comparative effectiveness, even though approximately half of the ECOG Performance Status data were missing.ConclusionsTo the best of our knowledge, we present the first in-depth application of QBA in the context of MAIC. Despite the real-world data limitations, with this MAIC, we show that it is possible to confirm the robustness of the results by using appropriate statistical methods.Trial registrationNA.

69 Inclusion of both patient level and study-level covariates in a meta-analysis

The traditional endpoint for assessing efficacy of chemotherapies for advanced/recurrent gastric cancer is overall survival (OS), but OS requires prolonged follow-up. We investigated whether progression-free survival (PFS) is a valid surrogate for OS. Using individual patient data from the GASTRIC meta-analysis, surrogacy of PFS was assessed through the correlation between the endpoints and through the correlation between the treatment effects on the endpoints. External validation of the prediction based on PFS was also evaluated. Individual data from 4069 patients in 20 randomized trials were analyzed. The rank correlation coefficient between PFS and OS was 0.853 (95% confidence interval [CI] = 0.852 to 0.854). The R (2) between treatment effects on PFS and on OS was 0.61 (95% CI = 0.04 to 1.00). Treatment effects on PFS and on OS were only moderately correlated, and we could not confirm the validity of PFS as a surrogate endpoint for OS in advanced/recurrent gastric cancer.

NM2 Using Randomized and Observational DATA to Predict Heterogeneous Treatment Effects

Endothelin Receptor Antagonists for Kidney Protection: Lessons from the SONAR Trial.

Spontaneous preterm birth is an important cause of neonatal mortality and morbidity. An increasing body of evidence suggests that uteroplacental ischemia plays an important role in the etiology of spontaneous preterm birth. We aimed to study whether antiplatelet agents reduce the risk of spontaneous preterm birth. We included data from an individual participant data meta-analysis of studies that had evaluated the effect of antiplatelet agents to reduce preeclampsia (Perinatal Antiplatelet Review of International Studies Individual Participant Data). The meta-analysis included 31 studies that randomized women to low-dose aspirin-dipyridamole or placebo-no treatment as a primary preventive strategy for preeclampsia. For the current study we analyzed data from 17 trials (28,797 women) that supplied data on type of delivery (spontaneous compared with indicated birth). Primary endpoints were spontaneous preterm birth at less than 37 weeks, less than 34 weeks, and less than 28 weeks of gestation. We analyzed outcomes for each trial separately using χ statistics and combined in an individual participant data meta-analysis using a binary logistic regression model. Women assigned to antiplatelet treatment compared with placebo or no treatment had a lower risk of spontaneous preterm birth at less than 37 weeks (relative risk [RR] 0.93, 95% confidence interval [CI] 0.86-0.996) and less than 34 weeks of gestation (RR 0.86, 95% CI 0.76-0.99). The RR of having a spontaneous preterm birth at less than 37 weeks of gestation was 0.83 (95% CI 0.73-0.95) for women who have had a previous pregnancy and 0.98 (95% CI 0.89-1.09) for women in their first pregnancy. The treatment effect was stable in all other prespecified subgroups. Antiplatelet agents reduce spontaneous preterm birth in pregnant women at risk for preeclampsia.

Combining individual patient data from multiple trials require specific modelling techniques in order to provide accurate predictions

One‐stage individual participant data meta‐analysis models should account for within‐trial clustering, but it is currently debated how to do this. For continuous outcomes modeled using a linear regression framework, two competing approaches are a stratified intercept or a random intercept. The stratified approach involves estimating a separate intercept term for each trial, whereas the random intercept approach assumes that trial intercepts are drawn from a normal distribution. Here, through an extensive simulation study for continuous outcomes, we evaluate the impact of using the stratified and random intercept approaches on statistical properties of the summary treatment effect estimate. Further aims are to compare (i) competing estimation options for the one‐stage models, including maximum likelihood and restricted maximum likelihood, and (ii) competing options for deriving confidence intervals (CI) for the summary treatment effect, including the standard normal‐based 95% CI, and more conservative approaches of Kenward‐Roger and Satterthwaite, which inflate CIs to account for uncertainty in variance estimates. The findings reveal that, for an individual participant data meta‐analysis of randomized trials with a 1:1 treatment:control allocation ratio and heterogeneity in the treatment effect, (i) bias and coverage of the summary treatment effect estimate are very similar when using stratified or random intercept models with restricted maximum likelihood, and thus either approach could be taken in practice, (ii) CIs are generally best derived using either a Kenward‐Roger or Satterthwaite correction, although occasionally overly conservative, and (iii) if maximum likelihood is required, a random intercept performs better than a stratified intercept model. An illustrative example is provided.

IntroductionAlzheimer's dementia (AD) is the most common cause of dementia, and several organisations, such as the National Institute for Health and Care Excellence, suggest that management of patients with AD...

Surrogate endpoints for overall survival in locally advanced head and neck cancer: meta-analyses of individual patient data

Individualized or personalized medicine has become a buzzword in the academic as well as public debate surrounding health care. The word personalized is appealing and transports the message of a new medicine evoking hopes for patients and physicians. However, personalized medicine is not necessarily about persons - it’s about subgroups and the more refined nosology of modern medicine which is based on much more profound knowledge on the pathological processes. To identify benefits and harms for these subgroups implicates several methodological issues, which I investigated in my PhD thesis. Subgroup analyses in randomized clinical trials (RCT) can have important impact on patient care if their results are true. However, most subgroup analyses have been shown to be false with detrimental effects on patients’ health. To investigate the planning of subgroup analyses in protocols of RCTs and the agreement with corresponding full journal publications, we established a cohort of RCT protocols and subsequent full journal publications. Protocols were approved between 2000 and 2003 by six research ethics committees in Switzerland, Germany, and Canada. We included 894 protocols of RCTs involving patients; 515 subsequent full journal publications were identified. About a third of protocols planned one or more subgroup analyses, but of those, only a small fraction (< 10%) provided a clear hypothesis for at least one subgroup analysis, and only a third planned an appropriate statistical test for interaction. 515 of 894 (58%) studies were published as journal article; of those, almost 50% reported at least one subgroup analysis. In 33% of all publications reporting subgroup analyses, authors stated that subgroup analyses were pre-specified but this was not supported by a third of the corresponding protocols. Furthermore, in those 86 publications in which authors claimed a subgroup effect, only 42% corresponding protocols reported a planned subgroup analysis. More than one third of statements in RCT publications about subgroup pre-specification had no documentation in the corresponding protocols. We conclude that subgroup analyses are insufficiently described in RCT protocols and investigators rarely specify the anticipated direction of subgroup effects. Credibility of claimed subgroup effects cannot be judged without access to RCT protocols. In statistical analysis, categorizing an inherently continuous predictor (e.g. age) raises several critical methodological issues. This problem also applies to investigation of interaction between e.g. treatment assignment and a continuous predictor in RCTs – e.g. do older patients benefit from a certain therapy compared to younger patients? We applied the new multivariable fractional polynomial interaction (MFPI) approach to investigate interaction between continuous patient baseline characteristics and the allocated treatment in an individual patient data meta-analysis of 3 RCTs (N=2299) from the intensive care field. In all included RCTs, patients requiring mechanical ventilation were randomized into two treatment groups: higher versus lower positive end expiratory pressure (PEEP) ventilation strategy. For each study, we used MFPI to calculate a continuous treatment effect function for four baseline characteristics and 3 outcomes. These functions were plotted with a 95% point wise confidence interval: 1. For each study separately, 2. For all studies combined (averaged function using a fixed effect model). This novel approach allows assessing whether treatment effects interact with continuous baseline patient characteristics and avoids categorisation-based subgroup analyses. These interaction analyses are exploratory in nature. However, they may help to foster future research using the MFPI approach to improve interaction analyses of continuous predictors in randomized trials and individual patient data meta-analyses.