Incorporating Biomarkers and Auxiliary Data in the Design of Adaptive Clinical Trials

• Activity Number: 9
• Type: Invited
• Date/Time: Sunday, July 31, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Ashish Sanil*+ and Scott Berry
• Companies: Berry Consultants and Berry Consultants
• Keywords: adaptive design ; bayesian adaptive design

Abstract

Clinical trial designs are typically constructed for comparing the primary endpoint across treatment arms. But some primary endpoints may be observed only after a significant amount of time has elapsed, and for some (e.g., survival) not all patients might have an event during the trial period. Often in such cases, early in the observation period, data are available on biomarkers and other auxiliary endpoints (tumor response, disease progression, etc.) that contain some indirect information about the primary endpoint. Adaptive trial designs based on Bayesian modeling have the natural ability to formally incorporate information from these early indicators for better interim assessments of the primary endpoints. Such leveraging of the early data lead to more efficient adaptive trial designs. We will discuss various aspects of such adaptive Bayesian designs and present examples.