Multivariate Network Meta-Analysis of Progression-Free Survival and Overall Survival

• Activity Number: 65
• Type: Topic Contributed
• Date/Time: Sunday, August 3, 2014 : 4:00 PM to 5:50 PM
• Sponsor: Health Policy Statistics Section
• Author(s): Jeroen Jansen*+ and Thomas Trikalinos
• Companies: Redwood Outcomes and Brown University
• Keywords: network meta-analysis ; time to event ; multivariate

Abstract

Cancer treatment efficacy is often quantified by time from treatment initiation to the occurrence of a particular event. Frequently studies report data on overall survival (OS), where the event is death from any cause, and on progression-free survival (PFS), where the event is death from any cause or disease progression, whichever occurred first. Separate meta-analyses of OS and of PFS data ignore their relationship. We present a method for the joint meta-analysis of OS and PFS that is based on a tri-state transition model with time-varying hazard rates modeled with fractional polynomials. We assume that, at any time, patients can be in one of 3 health states: "alive but not progressed", "alive and progressed", and "dead". PFS corresponds to time spent in the first state, and OS to time spent in the two alive states. The approach allows the joint network meta-analysis of OS and PFS, relaxes the proportional hazards assumption, extends to a network of more than two treatments, and simplifies the parameterization of decision and cost-effectiveness analyses. The data needed can be extracted directly from published survival curves. We demonstrate use with an example in melanoma.