To go or not to go in phase 2/3 oncology trials, a critical question with a unified answer.

The decision on how fast to go from single-arm Phase 1 after efficacy single-detection to randomized-controlled Phase 2/3 in oncology drug development is multifaceted and complex, if not often contentious. To facilitate the process, we have incorporated all viable options including the under-utilized adaptive Phase 2/3 designs into a comprehensive decision matrix for completeness and transparency. Illustrated with the 2-in-1 adaptive Phase 2/3 design, a formal decision analysis that explicitly optimizes the tradeoff between cost and benefit reveals a surprisingly robust efficacy bar for Go-No Go transition. It implies that similar robust bars exist for other adaptive Phase 2/3 designs, which remove a major hurdle for their wider adoption in practice.