An Audit Strategy for Blinded Independent Central Review of Progression in Cancer Clinical Trials Using Bayesian Predictive Model

• Activity Number: 664
• Type: Invited
• Date/Time: Thursday, August 13, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Soumi Lahiri* and Satrajit Roychoudhury
• Companies: Novartis Pharmaceuticals and Novartis Pharmaceuticals
• Keywords: Progression-free survival ; Blinded independent central review ; Hazard ratio ; Bayesian predictive model ; Posterior predictive probability

Abstract

As progression-free survival (PFS) is commonly used as the primary endpoint in oncology phase 3 trials, FDA has generally required a complete-case blinded independent central review (BICR) of PFS to assess and reduce potential bias in the local site evaluation (LE). However, recent publications and FDA analyses showed a high correlation between LE and BICR assessments of the PFS treatment effect, which questions whether complete-case BICR is necessary. One potential alternative is to use BICR as an audit tool to detect evaluation bias in the LE. We propose a BICR audit strategy as an alternative to a full BICR to provide assurance of the presence of a treatment effect using Bayesian predictive model. We develop a model to calculate the posterior predictive probability of observing a clinically meaningful hazard ratio (HR) from the complete sample by BICR, given the BICR HR from audited sample and the LE HR from the complete sample. If the probability crosses the pre-specified threshold value, then no complete BICR is needed. Otherwise, go for complete BICR. Performance and implementation of this method are evaluated through simulations as well as using real data retrospectively.