Statistical Issues in Adaptive Designs and Analysis of Clinical Trials with Predictive Biomarkers

• Activity Number: 198
• Type: Roundtables
• Date/Time: Monday, July 30, 2012 : 12:30 PM to 1:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Xiaoyun (Nicole) Li*+
• Companies: Merck & Co.
• Keywords: predictive biomarker ; adaptive design ; subgroup analysis ; personalized medicine ; clinical trial design ; optimization

Abstract

With the rapid development of molecular target therapies, the fact that many medicines benefit only a minority of the patients calls the urge of the development of personalized medicine, which uses one or a composite of predictive biomarkers as classifiers to select patients who are likely to benefit from the medicine. This roundtable discussion will focus on discussion of study designs in phase II/III adaptive design with potential predictive biomarker, as well as analysis challenges around them. Discussion topics include statistical issues of how to allow flexible enrichment of subgroup population during the trial and how to optimize the alpha spending for a predictive biomarker subgroup. It will also be really interesting to discuss whether it is more efficient to use Bayesian tool to help make informative interim decisions and whether we can use evidence-based design to help optimize the clinical trial design.