Elastic meta-analytic-predictive prior for dynamically borrowing information from historical data with application to...

Area of Science:

• Biopharmaceutical research
• Statistical methodology in clinical trials
• Drug development and regulation

Background:

• Biosimilars are highly similar biological products to reference products, requiring rigorous trials to demonstrate equivalence.
• Establishing biosimilarity often relies on two-arm randomized clinical trials comparing the biosimilar to the reference product.
• Existing methods may not fully leverage historical data on the reference product, potentially limiting trial power.

Purpose of the Study:

• To introduce and evaluate the elastic meta-analytic-predictive (EMAP) prior method for biosimilar trials.
• To enhance the statistical power of biosimilar equivalence trials by utilizing historical data from the reference product.
• To ensure robust type I error rate control in biosimilar trials using the proposed EMAP prior.

Main Methods:

• The EMAP prior method extracts prior information from historical reference product studies via meta-analysis.
• An elastic function adaptively discounts the meta-analytic-predictive (MAP) prior based on data congruence between historical and trial data.
• The method allows for flexible information borrowing, ranging from full to no borrowing, ensuring consistency and controlled type I errors.

Main Results:

• Extensive simulations demonstrate that the EMAP prior method outperforms the robust MAP prior.
• The EMAP prior achieves comparable or higher statistical power in biosimilar trials.
• The proposed method provides better control over type I error rates, ensuring trial validity.

Conclusions:

• The EMAP prior method is an effective information-borrowing strategy for biosimilar equivalence trials.
• This novel approach enhances trial power and maintains statistical rigor, crucial for regulatory approval.
• The EMAP prior offers a statistically sound way to leverage historical data in biosimilar development.