Futility Rule and Unblind Sample Size Re-Estimation in Phase III Drug Development: Method and Practical Example

• Activity Number: 130
• Type: Contributed
• Date/Time: Monday, August 1, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Kyoungah See*+
• Companies: Eli Lilly and Company
• Keywords: Conditional Pronababilities ; Operating chracteristics ; Stopping Rules

Abstract

In this talk we shall focus our discussion on futility monitoring and sample size re-estimation in confirmatory phase 3 clinical trials based on unblinded interim results through practical case and simulation studies. These techniques are useful when a lack of historical information and uncertainty about variability in the data and primary effect size exist. In such settings, it is rather essential to unblind the interim data and an independent interim analysis committee is required. We shall discuss practical examples including some statistical methodology and some operational and regulatory issues. Sample size re-estimation allows an adjustment in sample size during the trial. Starting with a small but reasonable up-front sample size commitment, an increase is applied only if interim results meet the prespcified rule. The sample size re-estimation method follows Proschan et al.(2006), Gao et al. (2008), and Mehta and Pocock (2010). Analyses based on binary endpoints are explored.