Using Historical Data for a Transitional Decision of Single-Arm Phase II Clinical Trials

• Activity Number: 427
• Type: Contributed
• Date/Time: Tuesday, August 3, 2010 : 2:00 PM to 3:50 PM
• Sponsor: IMS
• Author(s): Zunqiu Chen*+ and Yiyi Chen and Motomi Mori
• Companies: Oregon Health & Science University and Oregon Health & Science University and Oregon Health & Science University
• Keywords: Phase II oncology trials ; transitional decision

Abstract

Many phase II oncology trials are single-arm clinical trials that aim to decide whether the study treatment is promising enough for a phase III study. Such a decision is called a transitional ``Go / No Go'' decision. Traditionally, the transitional decision is based on hypothesis testing for one sample. The high failure rate in phase III oncology trials suggests that the ``Go / No Go'' decisions based on the traditional hypothesis testing method are often incorrect. This is partly due to ignoring the uncertainty associated with the specified null hypothesis. We propose a frequentist statistical decision rule that incorporates the uncertainty associated with the specified null. We compare the new decision rule with the traditional frequentist decision rule and the Thall and Simon's Bayesian decision rule, with an illustrative application in a real clinical trial.