231 – Applications of Adaptive Designs in Clinical Trials
A Design Adapting Between Clinical Trial Phases
Keaven Anderson
Merck
Xiaoyun Li (Nicole)
Merck & Co.
At the time a study is designed there is imperfect information and it may not be completely clear that a drug is worth the investment required for a large, pivotal trial required for regulatory approval. An adaptive phase II/III study with an early futility analysis is sometimes chosen in order to shorten the development timeline and limit unneccessary patient exposure and investment. However, in a trial with a time-to-event endpoint, there may be many patients with relatively little follow-up if a trial is stopped early. Here we consider transformation of a trial from Phase III to a smaller Phase II at an interim analysis as an additional option. The intent is to allow collection of follow-up on patients already enrolled in order to get the best data possible for future decisions surrounding development of the drug. The objectives and required observed treatment e ffect and Type I error for a positive Phase II adaptation may be di fferent from the Phase III design. Compared to stopping a Phase III trial for futility, transforming to a Phase II trial with an intermediate treatment eff ect may be more cost-eff ective. The methods used here are related to group sequential design theory. The critical criteria for and timing of phase selection at interim analysis are discussed.
