Bayesian Adaptive Phase II/III Clinical Trial Design Using the Relationship Between Tumor Response and Survival

• Activity Number: 470
• Type: Contributed
• Date/Time: Wednesday, August 4, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Donald Arthur Berry and Haiying Pang*+
• Companies: MD Anderson Cancer Center and MD Anderson Cancer Center
• Keywords: Bayesian ; adaptive ; tumor response ; survival ; seamless phase II/III

Abstract

Phase III cancer clinical trials use overall survival or progression-free survival as the primary endpoint. Conventional methods are limited by event rate, even in cancers with poor prognoses. Tumor response is a secondary endpoint but is ignored in drawing conclusions about progression-free or overall survival. We model the relationship between tumor response and survival. We use the information about a treatment's effect on the tumor to understand its consequent effect on survival. If the relationship happens to be weak, we learn that. But when tumor response is informative about survival, the trial is more efficient in the sense of statistical power, sample size and trial duration. We use simulation to find the operating characteristics of the proposed method and compare these with operating characteristics of the conventional approach.