A Bayesian Adaptive Design for Concurrent Trials Involving Biologically-Related Diseases

• Activity Number: 575 - Trial Design with Subpopulation Heterogeneity: What Is the Role of Real-World Evidence?
• Type: Topic Contributed
• Date/Time: Thursday, August 6, 2020 : 3:00 PM to 4:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Xun Jiang*
• Companies: Amgen
• Keywords: Bayesian; adaptive design; clinical trial

Abstract

In this paper, we develop a Bayesian adaptive design framework for a clinical program where an investigational product is to be studied concurrently in a set of phase II trials for a set of biologically related diseases with the goal of demonstrating superiority to a control in each disease. The proposed approach borrows information on treatment effectiveness using Bayesian model averaging (BMA) to combine inference from analyses based on combinations of informative conjugate power priors that either strongly favor the null hypothesis or the hypothesized alternative in each disease indication. The approach allows for information borrowing without making untenable assumptions regarding how treatment effect parameters relate across the different diseases being studied and provides a powerful framework for information borrowing in a general setting where each disease has a potentially different endpoint (e.g., some binary and some continuous). Information borrowing is induced through elicited prior model probabilities that satisfy a dependence criterion which we show to be sufficient and necessary to result in information borrowing in the BMA context.