Biomarker-based Bayesian randomized clinical trial design for identifying a target population

Area of Science:

• Biostatistics
• Clinical Trial Design
• Biomarker Discovery

Background:

• Incorporating biomarkers into clinical trials is crucial for developing targeted cancer therapies and immune-oncology agents.
• Identifying sensitive patient subpopulations often requires larger sample sizes, increasing costs and study duration.
• Existing methods may not efficiently identify or exclude specific patient groups based on biomarker data.

Purpose of the Study:

• To propose and evaluate a novel biomarker-based Bayesian (BM-Bay) randomized clinical trial design.
• To incorporate continuous or graded biomarkers for defining multiple patient subpopulations.
• To design efficient interim analyses with decision criteria for identifying sensitive patient populations for new treatments.

Main Methods:

• A biomarker-based Bayesian (BM-Bay) randomized clinical trial design is proposed.
• The design utilizes predictive biomarkers on a continuous or graded scale to define patient subpopulations.
• Interim analyses with specific decision criteria are incorporated for efficacy evaluation of time-to-event outcomes.

Main Results:

• Extensive simulation studies evaluated the operating characteristics of the BM-Bay design.
• The method demonstrated the probability of correctly identifying desired patient subpopulations.
• The expected number of patients required under various clinical scenarios was assessed.

Conclusions:

• The proposed BM-Bay design enables efficient and accurate identification of sensitive patient subpopulations.
• The decision criteria allow for both the inclusion of sensitive groups and exclusion of insensitive ones.
• The method is illustrated through the design of a randomized phase II immune-oncology clinical trial.