Adaptive group-sequential design with population enrichment in phase 3 randomized controlled trials with two binary...

Area of Science:

• Clinical Trial Design
• Biostatistics
• Pharmacometrics

Background:

• Co-primary endpoints offer a comprehensive assessment of experimental interventions compared to single endpoints.
• Traditional trial designs may dilute treatment effects in heterogeneous populations.

Purpose of the Study:

• To propose and develop an adaptive population enrichment design utilizing co-primary endpoints.
• To provide theoretical basis and stage-wise boundary values for futility and efficacy assessments.
• To evaluate the design's power under various efficacy and subgroup configurations.

Main Methods:

• A two-arm, two-stage parallel group design incorporating population enrichment at the interim analysis.
• Non-responsive subgroups are dropped, with efficacy testing conducted solely in the enriched population.
• Two binary co-primary endpoints are evaluated, requiring superiority in both for trial success.

Main Results:

• Stage I futility and efficacy boundary values are -0.1040 and 2.2761, respectively (alpha=0.025 per endpoint).
• Stage II futility and efficacy boundary value is 2.2419.
• The design increases power to detect effects in responsive subgroups with substantial treatment effect heterogeneity.

Conclusions:

• Adaptive population enrichment with co-primary endpoints enhances the ability to detect true treatment effects.
• Early dropping of non-responsive subgroups mitigates false-negative results caused by effect dilution.
• This design improves the efficiency and reliability of drug development trials in heterogeneous populations.