IBIS: identify biomarker-based subgroups with a Bayesian enrichment design for targeted combination therapy

Area of Science:

• Biostatistics
• Clinical Trial Design
• Biomarker Discovery

Background:

• Combination therapies offer improved cancer treatment effects but complicate clinical trial efficacy evaluation due to numerous subgroups and complex biomarker profiles.
• Targeted combination therapies necessitate innovative clinical trial designs to assess efficacy across diverse patient subgroups.

Purpose of the Study:

• To introduce IBIS (Identify BIomarker-based Subgroups), a statistical tool for biomarker-based subgroup identification and population enrichment in clinical trials.
• To facilitate the identification of patient subgroups most likely to benefit from investigational combination therapies.

Main Methods:

• IBIS employs subgroup division based on biomarker levels, Jensen-Shannon divergence for efficacy differentiation, and Bayesian hierarchical models (BHM) for robust efficacy evaluation.
• A hypothesis testing framework using Bayes factors is utilized for subgroup identification, supporting go/no-go decisions and population enrichment.

Main Results:

• Simulation studies demonstrate that IBIS achieves desired accuracy and precision in estimation.
• IBIS exhibits superior and robust performance in subgroup identification and population enrichment compared to traditional methods.

Conclusions:

• IBIS shows significant potential as a valuable tool for biomarker-based subgroup identification and population enrichment in clinical trials involving targeted combination therapies.
• The framework provides a method for obtaining design parameters for adaptive enrichment designs.