DIA/FDA Statistics Forum 2014

This is an intertwined pair of sessions focused on statistical and modeling methods for dose finding studies. Dose selection remains one of the most challenging steps in clinical drug development, being directly associated with the success, or failure, of confirmatory studies and regulatory submissions. A common approach drug development has been to consider a dose finding study (also referred to as Phase 2B trial) as having the potential to be a pivotal study in the program. As a result, emphasis has been placed on multiple comparisons between active doses and control with strong control of Type I error, making studies that should, in principle, be exploratory in nature, look more like mini-Phase 3 studies. Poor dose selection has often been observed in such cases, eventually leading to costly failures in confirmatory programs or regulatory submissions. In the first session, three expert speakers (from regulatory agencies and industry) will review some of the critical issues related to dose finding studies and will present alternative, potentially more efficient dose selection approaches, including model-based methods.