Department of Biostatistics Seminar/Workshop Series

A Case Study of a Bayesian Adaptive Design in a Phase II Clinical Trial

Ben Saville, PhD

Assistant Professor, Department of Biostatistics, Vanderbilt University School of Medicine

Wednesday, November 9, 1:30-2:30pm, MRBIII Conference Room 1220

A Bayesian adaptive design is proposed for investigating the combination effect of venorelbine and metformin in patients with metastatic breast cancer in a Phase II clinical trial. The primary outcome is time to progression of disease, in which the hazard rate of the treatment group is compared to a historical control to determine if further investigation of the treatment regimen is warranted. Assuming an exponential distribution for time to disease progression, the posterior distribution of the hazard rate is shown to follow a gamma distribution. An interim analysis is conducted using Bayesian predicted probabilities. Frequentist operating characteristics of the Bayesian design are evaluated via simulation.