Control of Overall Type I Error in Clinical Trials with Both Surrogate and Final Endpoints

• Activity Number: 25
• Type: Contributed
• Date/Time: Sunday, August 6, 2006 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Chung-Kuei Chang*+
• Companies: Cephalon, Inc.
• Keywords: surrogate endpoint ; accelerate approval ; overall type I error ; interim analysis ; modified Simes procedure ; adjusted p-value

Abstract

For serious diseases with an unmet medical need, FDA may grant an accelerated approval based on a surrogate endpoint and stipulate post-marketing studies using the endpoint of primary interest (the final endpoint). FDA may also request the sponsor to design phase 3 studies with enough power for both surrogate and final endpoints and adjust for multiplicity. Assuming a non-negative but unknown correlation of the two endpoints, we show how to control the overall type I error for testing both endpoints with multiple interim analyses. An adjusted p-value for the interim and final analyses is first defined for each endpoint. After showing that these adjusted p-values follow the uniform distribution on (0, 1) under the null hypothesis, we demonstrate that a modified Simes procedure applied to the adjusted p-values controls the overall type I error through a phi-inverse transformation.