Drug Regimen Selection in Early-Stage Two-Arm Oncology Clinical Trials

• Activity Number: 583
• Type: Contributed
• Date/Time: Wednesday, August 4, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Guohui Liu and Xuedong Chi*+
• Companies: Millennium Pharmaceuticals, Inc. and Millennium Pharmaceuticals, Inc.
• Keywords: Regimen selection ; Oncology ; Clinical trials ; Bayes

Abstract

One of the main objectives in early stage clinical trial is the selection of drug regimen, such as dosing schedules or drug combinations, for further drug development. A small-scale two-arm clinical trial is usually conducted to pick the relatively superior arm based on efficacy or toxicity parameters. Steinberg and Venzon proposed a two stage approach based on statistical selection theorem which allows the trial to be stopped earlier if the difference in the number of responses between two randomized arms is larger enough. We implemented the first stage of that approach to have more confidence in picking the appropriate regimen. We also expanded this approach using Bayesian method, and applied to the scenario, where equal number of patients is not required for the two arms. The results showed that Bayesian approach is more conservative than the frequentist method.