Statistical Consideration Framework of Enrichment Designs: Leveraging Early Data for Late-Phase Success

Music City Center 

In this presentation, we introduce a statistical consideration framework for how to effectively leverage early-phase data to select appropriate enrichment designs for confirmatory trials in the context of oncology drug development. Our approach focuses on two key metrics for evaluating different design options: overall power and overall value, of which the latter considers population size and treatment effect size. The framework quantitatively integrates available information and a variety of considerations, such as extrapolation of enrichment observed in early endpoints to clinical endpoints, strategies to enhance power by accounting for the correlation between the enriched population and the all-comer population and assessing multiplicity strategies. We will provide general guidance supported by illustrative numeric examples

Enrichment design

Early endpoints

Multiplicity