|
Abstract:
|
A 2-in-1 adaptive Phase 2/3 design with well-controlled Type 1 error was proposed by Cong et al. The 2-in-1 design is beneficial to improve the overall development efficiency by: 1) building-in an early and informative Go/NoGo (to Phase 3) decision-making at Phase 2 with less cost (fewer participants); 2) potential registration path forward in either Phase 2 (if no Go) or 3 (if Go), respectively with overall Type 1 error well controlled for the whole Phase 2/3 study. The referred paper illustrates a general idea for the case when there is single primary endpoint/hypothesis with one analysis in Phase 3 portion of the adaptive design, while we extend to a widely-used oncology trial design with dual-primary endpoints/hypotheses and multiple interim analyses in a group sequential setting after Phase 3 expansion. Under the different assumptions of the correlation of the dual-primary endpoints, we showed that the overall Type 1 error is also well controlled for 2-in-1 adaptive Phase 2/3 design with dual-primary endpoints in the group sequential testing of the Phase 3 portion.
|
