|
Abstract:
|
An effective and valid predictive enrichment strategy is of critical importance in oncology drug development when the new treatment is likely to benefit only a subset of patient population. Jiang (2007) proposed a biomarker adaptive threshold design in phase III setting, which combines a test for treatment in the overall population and establishment of a biomarker cut point for identifying a sensitive population. In reality, based on mechanism of action and clinical evidence from early phase of drug development, it can be beneficial to start considering the patient enrichment as early as Phase II study. However, a reliable biomarker might not exist until later in the drug development since biomarker is often concurrently being developed with clinical studies. The impact of accuracy of the measurements used to identify enriched population can be profound. In this presentation, we modify Jian (2007) into phase 2 single arm design setting and demonstrate the impact of biomarker measurement error on the operation characteristics of adaptive threshold design under various assumptions.
|
ASA Meetings Department
732 North Washington Street, Alexandria, VA 22314
(703) 684-1221 • meetings@amstat.org
Copyright © American Statistical Association.