PS18: Enrichment Design: Challenges Arising in Statistical Approaches and Clinical Interpretations

Enrichment design has become increasingly popular in clinical trials for both therapeutic treatments and diagnostic purposes. It provides several advantages, including increased efficiency, higher statistical power, reduced sample size requirements, etc. While "all-comers" trial designs often struggle to demonstrate meaningful treatment effects, enrichment designs aim to address this challenge by intentional selecting or prioritizing patient subgroups who are more likely to benefit from an investigational therapy or more likely to exhibit measurable outcomes, based on patients' clinical characteristics, biomarkers, or early response indicators. The rise of molecular diagnostics and biomarker-driven therapeutics has accelerated the development of enrichment designs. Now enrichment designs have been widely adopted in clinical trials where • Event rates are low, and traditional recruitment would require large sample size • Biomarker-defined subgroups are central to the treatment procedure and outcomes • Disease heterogeneity complicates signal detection • Precision medicine approaches demand targeted evaluation Despite the advantages, enrichment designs may raise important statistical, operational, and regulatory considerations. Challenges include determining how results from an enriched population with specific patient characteristics can be extrapolated to the general patient population; optimizing the process for identifying potential participants who meet the enrichment criteria while avoiding bias and preventing type I error inflation in subsequent study phases, etc. These considerations require careful planning to minimize bias, ensure generalizability, and preserving validity and interpretability of study results. Successfully addressing these challenges requires close collaboration among multidisciplinary teams, including clinicians, biostatisticians, pharmaceutical/medical device developers, and regulatory scientists, to develop robust designs that meet both scientific and regulatory standards. This session will invite clinicians, biostatisticians, regulatory scientists to discuss the issues and questions that emerge in enrichment design studies. Speakers will present examples from real-world practices and discuss regulatory considerations, statistical approaches, and clinical perspectives on the challenges encountered during the planning, implementation, and regulatory evaluation of clinical trials with enrichment designs. Topics and questions include, but are not limited to: For Regulatory Scientists: • What are the regulatory considerations for treatment effects/device performances derived from enriched study populations that may differ in composition from the intended patient population? What is the impact of enrichment designs on product labeling and marketing claims? For Statisticians: • Statistical methodologies developed to address challenges in enrichment design-based clinical trials, such as Type I error rate control (for example, in adaptive enrichment designs), bias mitigation, etc. For Clinicians and Industry Developers: • What are the clinical considerations before, during, and after implementing clinical trials with enrichment designs? How do enrichment designs impact the clinical interpretation of trial outcomes?