Joint Inference of Overall Survival and Progression-Free Survival

• Activity Number: 569
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Somesh Chattopadhyay*+ and Shenghui Tang and Rajeshwari Sridhara
• Companies: FDA and FDA and FDA
• Keywords: overall survival ; progression-free survival ; joint inference

Abstract

This presentation reflects the views of the author and should not be construed to represent FDA's views or policies.

Overall survival (OS), defined as the time from randomization to death due to any cause, and progression-free survival (PFS), defined as the time from randomization to either documented disease progression or death from any cause (whichever occurs earlier), are two important endpoints in oncology clinical trials. PFS is sometimes used as a surrogate reasonably likely to predict for OS. In some cases a drug or biologic product may improve PFS but does not improve OS. Frequently, where the PFS is the primary endpoint, mature PFS data are available at the time a regulatory decision is taken, but the OS data are not mature at that time. It is important to find out how these two endpoints are related and how we can make joint inference of PFS and OS. In this research we make joint inference of these two endpoints in both cases where mature PFS and OS data are available and where mature PFS data and partial OS data are available. We start with parametric models and explore the semiparametric and nonparametric cases.