Auxiliary variable-enriched biomarker-stratified design

Area of Science:

• Biostatistics
• Clinical Trial Design
• Precision Medicine

Background:

• Biomarker assessment is crucial for precision medicine clinical trials to identify patient subgroups for targeted therapies.
• Standard biomarker-stratified designs (BSD) can be inefficient and costly, particularly with low prevalence subgroups or high biomarker assessment costs.

Purpose of the Study:

• To introduce and evaluate an auxiliary variable-enriched biomarker-stratified design (AEBSD) for improving clinical trial efficiency.
• To reduce costs and patient waiting times in biomarker-driven clinical trials.

Main Methods:

• Proposed an auxiliary variable-enriched biomarker-stratified design (AEBSD) using an inexpensive, correlated auxiliary variable for enrichment.
• Developed an adaptive Bayesian method to dynamically adjust the positive predictive value (PPV) during the trial.
• Conducted numerical studies and presented an illustrative example.

Main Results:

• AEBSD reduces total trial costs compared to standard BSD when biomarker prevalence is low and the auxiliary variable's PPV exceeds prevalence.
• AEBSD allows immediate patient randomization post-screening, decreasing treatment waiting times.
• The adaptive Bayesian method effectively adjusts for uncertainty in the PPV.

Conclusions:

• AEBSD offers a more efficient and cost-effective approach for biomarker-stratified clinical trials, especially in precision medicine.
• This design facilitates faster patient enrollment and treatment initiation.
• An R package is available to implement the proposed methodology.