Video-on-Demand Library

This content is restricted to members of PSI. If you are already a member please login. To join PSI or to see all the benefits of membership click here.

04 June 2019

Learning from the rapidly emerging data in drug development of cancer immunotherapies is essential to identify those patients benefiting most. However, sometimes data is only available from a different indication and there remains residual uncertainty whether the observed pattern applies to the new indication under investigation or not. Adaptive designs provide a valuable tool to address this residual uncertainty. A case study in breast cancer will be discussed where an ongoing trial in an all-comer population was amended to account for emerging external data from a later line of therapy suggesting a much larger treatment effect size in a biomarker defined subgroup. Statistical, clinical and operational challenges in embedding the ongoing all-comer trial into the adaptive framework while maintaining its statistical integrity will be discussed. The amended adaptive design is a two-stage enrichment design which allows for potential stopping at stage 1 for overwhelming efficacy or futility in the subgroup or all-comers population, respectively, as well as for population selection for stage 2. Target population selection after stage 1 is based on the observed treatment effect. Type I error protection in the strong sense is achieved by the joint use of p-value combination tests and a closed testing procedure. Operational characteristics of the design will be discussed, including consideration of alternative criteria for subgroup selection such as Bayesian conditional power. Finally, considerations regarding the minimal detectable difference in this adaptive design will be provided.

Related Documents

Part of Collection(s)

Topic(s)

Upcoming Events

Latest Jobs