Audit Strategies for Progressive-Free Survival

• Activity Number: 473
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Meihua Wu*+ and Kalyanee Kalyanee Appann and William Mietlowski and Yong Zhang and Can Cai and Paramita Saha
• Companies: University of Michigan and Novartis and Novartis and Novartis and Novartis and Novartis
• Keywords: progressive-free survival ; reader bias ; blinded independent central review ; auxiliary variable estimator ; inverse probability censoring weight

Abstract

Progression-free survival (PFS), defined as time to documented disease progression or death without documented progression is often used as the primary endpoint in randomized oncology trials. The local radiologist assessment of PFS is used for patient management but might have reader bias leading to a biased estimate of hazard ratios (HR). Blinded independent central review (BICR) has been proposed to reduce potential reader bias in PFS. However, a 100% BICR could increase the cost, prolong trial duration, and introduce informative censoring.

We first consider the auxiliary variable estimator (Dodd 2010) that reduces cost/time by combining local assessment data and sub-sample BICR audit data. We then investigate the degree of discordance between the local assessment and BICR. Finally, we address potential informative censoring by: 1) continuing follow up until the PFS event is confirmed by BICR in combination with auxiliary variable estimator; 2) using the Inverse probability of censoring weight (IPCW) approach (Robins and Finkelstein, 2000) in combination with auxiliary variable estimator. We evaluate our proposed strategies with extensive simulations and clinical trial data.