Optimizing subgroup selection in two-stage adaptive enrichment and umbrella designs

Area of Science:

• Clinical Trial Design
• Biostatistics
• Medical Research Methodology

Background:

• Identifying patient subgroups who benefit from novel therapeutics is crucial for personalized medicine.
• Traditional clinical trials may not efficiently allocate resources to the most responsive patient populations.
• Adaptive trial designs offer flexibility to optimize treatment allocation and subgroup identification.

Purpose of the Study:

• To develop and evaluate adaptive two-stage confirmatory clinical trial designs.
• To identify patient subgroups likely to benefit from a new treatment through subgroup-specific efficacy testing.
• To optimize trial parameters, such as recruitment probabilities, using a Bayesian decision-theoretic framework.

Main Methods:

• Design of two-stage confirmatory trials with interim analysis for adaptation.
• Implementation of adaptive elements using the conditional error rate approach to maintain overall error rates.
• Optimization of design parameters via a Bayesian decision-theoretic framework, maximizing a utility function incorporating subgroup prevalence.
• Comparison of designs with familywise error rate control (closed testing) and per-comparison error rate control (umbrella trials).

Main Results:

• Demonstration of adaptive trial designs that can efficiently identify treatment-benefiting subgroups.
• Numerical examples illustrating the optimization process for design parameters.
• Validation of the effectiveness of proposed adaptive designs in various error-controlling scenarios.
• The conditional error rate approach effectively protects overall statistical error rates during adaptation.

Conclusions:

• Adaptive two-stage trials provide a robust framework for identifying patient subgroups that benefit from new treatments.
• Bayesian optimization enhances trial efficiency by considering subgroup prevalence.
• The proposed methods offer flexibility and statistical rigor for modern clinical trial design, applicable to both traditional and umbrella trial structures.