Abstract:
|
Multi-stage phase II/III seamless trial designs have become increasingly popular in clinical research and development. In practice, it is typical to consider a surrogate endpoint at stage one (phase II study) and a clinical endpoint at stage two (phase III study), especially for early decision-making, when the clinical outcome takes longer to observe, or when the study objectives at each stage are similar but vary slightly (e.g., dose selection vs. efficacy confirmation). A common approach is to assume a relationship between the surrogate and clinical endpoints to bridge the two phases. Our proposed transitional seamless design combines a phase II and phase III study with different objectives and endpoints, with and without adaptations. This design allows for the estimation of nuisance parameters (such as the correlation between the surrogate and clinical endpoints) and enables investigators to update both the treatment effect size of the clinical outcome and the sample size re-estimation based on the observed treatment effect in the surrogate endpoint. The impact of the adaptive methods on controlling of the FWER under the proposed design was examined using simulations.
|