An Integrated Bayesian Posterior Probability with Simon Two-Stage Design for a Randomized Phase II Clinical Trial

• Activity Number: 178
• Type: Contributed
• Date/Time: Monday, August 1, 2016 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Dung-Tsa Chen* and Po-Yu Huang and Hui-Yi Lin
• Companies: Moffitt Cancer Center and National Chung-Hsing University and Louisiana State University Health Sciences Center
• Keywords: Simon two-stage design ; Bayesian posterior probability

Abstract

Simon two-stage design has been widely used in phase II clinical trials to determine if a new drug agent has sufficient anti-tumor activity to warrant further Phase III studies. Because of its popularity, it has also used to evaluate two or more drug agents simultaneously in a randomized fashion to select a winner. However, this strategy has been criticized for lack of objective comparison among the agents. In this study, we integrate a Bayesian posterior probability into a two-arm randomized Simon two-stage design. The Bayesian posterior probability is defined as probability of the response rate in arm B higher than in arm A (assume two treatment arms A and B). The posterior probability is then applied to determine the winner when both arms pass the second stage. This trial design allows us to more objectively determine if it is worthwhile to conduct a large phase III trial.