PS31: Recent Statistical Advancements of Interim Analysis in Clinical Trials

In modern clinical trials, interim analyses are commonly pre-specified and conducted to inform decision-making regarding the overall trial. Regulatory agencies, such as the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA), have issued detailed guidance to provide recommendations and key considerations for implementing adaptive trial designs, with in-depth discussions on better practice to conduct interim analyses. This session brings together three speakers and a discussant from academia, industry, and FDA to highlight recent advances that address treatment-effect heterogeneity and improve interim monitoring. The talks cover Bayesian adaptive enrichment with hierarchical modeling and dynamic borrowing across prospectively defined subgroups; futility monitoring strategies paired with adaptive enrichment to avoid prematurely stopping trials when benefit is concentrated in subpopulations, with practical and regulatory considerations; and efficiency-enhanced interim analyses that integrate baseline covariates and auxiliary outcomes using empirical likelihood methods with machine learning and sample splitting within group sequential/alpha-spending frameworks. In summary, the presentations in this session will showcase innovative statistical methods and designs in clinical trials, offering insights from regulatory, industry and academia perspectives to improve decision-making through interim analyses.