Incorporating Historical Data in Bayesian Phase I Trial Design: Evaluating the Similarity in Dose-Toxicity Relationship Between Subgroups

• Activity Number: 505
• Type: Contributed
• Date/Time: Wednesday, August 12, 2015 : 8:30 AM to 10:20 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Kentaro Takeda* and Satoshi Morita
• Companies: Astellas and Kyoto University Graduate School of Medicine
• Keywords: dose finding ; historical data ; prior effective sample size ; continual reassessment method ; phase I study design ; Bayesian approach

Abstract

In oncology area, following a phase I dose-finding trial completed in a certain population of patients, further phase I trials are often conducted to determine the maximum tolerated dose (MTD) for different patient subpopulations. This may be due to concerns about possible differences in treatment tolerability between subgroups. We propose a Bayesian approach to incorporating historical data to establish prior distributions for a dose-finding trial to develop an anti-cancer agent. Our proposed approach aims to appropriately borrow strength from a previous trial to improve the MTD determination in another population of patients. We design a further phase I trial using the Bayesian continual reassessment method (CRM) to utilize all available prior information through prior distributions of the model parameters. We propose a historical-to-current parameter to evaluate the similarity in dose-toxicity relationship between subgroups. We present a simulation study of this proposed method to explore its properties. The simulation results compared with a standard approach that ignores historical information are very encouraging.