A Signature Enrichment Design with Bayesian Adaptive Randomization.

Area of Science:

• Oncology
• Biostatistics
• Clinical Trial Design

Background:

• Precision cancer medicine relies on biomarker signatures for optimal treatment assignment.
• Current clinical trial designs may not fully leverage adaptive strategies for patient selection and treatment allocation.

Purpose of the Study:

• To propose a novel group sequential randomized phase II clinical trial design.
• To incorporate adaptive biomarker signature updates, patient enrichment, and Bayesian response-adaptive randomization.
• To introduce new performance criteria evaluating individual benefits and losses within clinical trials.

Main Methods:

• A group sequential randomized phase II design is proposed.
• The design features dynamic biomarker signature updates.
• Bayesian response-adaptive randomization is employed for treatment assignment.
• Enrichment strategies are utilized for patient selection.

Main Results:

• The proposed design enhances participants' chances of receiving optimal personalized treatments.
• It leads to higher response rates within the trial.
• Adaptive enrichment strategies increase the likelihood of discovering effective new drugs by identifying sensitive patient subsets.

Conclusions:

• The novel group sequential randomized phase II design offers significant advantages over traditional equal randomization methods.
• It improves individual patient outcomes and the efficiency of drug discovery in precision oncology.
• Simulation studies and a non-small-cell lung cancer trial example validate the design's benefits.