Adaptive Multi-Armed Survival Designs Using a Short-Term Endpoint for Treatment Selection

• Activity Number: 99
• Type: Invited
• Date/Time: Monday, July 30, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Pharmaceutical Research and Manufacturers of America
• Author(s): Gernot Wassmer*+
• Companies: ADDPLAN Inc.

Abstract

Adaptive trial methodology for multi-armed trials have been extensively discussed in the past. A general principle to construct test procedures which control the Familywise Type I error rate in the strong sense is available (see, e.g., Bretz et al., Statistics in Medicine, 2009, 28, 1181 - 1217). Using survival data, a problem arises when using information of patients which are under risk at interim. A similar problem was recently considered for designs with subpopulation selection (Jenkins et al., Pharmaceut. Statistics 2011, 10, 347-356; see also Friede et al., Statistics in Medicine 2011, 30, 1528-1540). Here, only the results for the survival endpoint may be used for monitoring the study and potentially adapting the design. We propose a solution for multi-armed designs that can be used within a multi-stage group sequential design. We illustrate through the use of case studies how these trials can be designed in practice. This includes the assessment of effect sizes and the correlation structure between short-term and the primary endpoint, the assessment of sample size, timing of interim analysis, and selection rule on the performance of the