Sensitivity Analysis for PFS

• Activity Number: 583
• Type: Contributed
• Date/Time: Wednesday, August 4, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Alicia Zhang*+ and Alan Rong and Ying Tian and Michael Wolf
• Companies: Amgen Inc. and Amgen Inc. and Amgen Inc. and Amgen Inc.
• Keywords: sensitivity analysis ; PFS ; oncology therapeutic trials

Abstract

Sensitivity analysis for PFS has become increasingly important in oncology therapeutic trials due to the fact that more trials use PFS (from central review) as the primary endpoint. Potential biases could mislead the primary analysis results and hence the impact should be assessed. Those biases come from many aspects including the tumor assessment frequency, tumor assessment interval, truncation of central review by local disease progression, open-label study and use of the subsequent anti-cancer therapies. In this presentation, we will present some key issues for PFS analyses raised from several P3 mCRC studies and propose some sensitivity analysis methods with different degrees of conservativeness to address these issues. Interpretation of those analyses including the impact to the estimation of treatment effect and the control of type 1 error rate will be discussed.