A Bayesian adaptive design for biosimilar trials with time-to-event endpoint.

Area of Science:

• Pharmacometrics and Biostatistics
• Clinical Trial Design
• Regulatory Science

Background:

• Biosimilar drugs require rigorous study designs to prove equivalence to reference products.
• Current biosimilar trial designs often rely on binary or continuous endpoints, limiting applicability.
• Time-to-event endpoints are crucial for many biological products but underutilized in biosimilar assessments.

Purpose of the Study:

• To propose a novel Bayesian adaptive design for biosimilarity trials with time-to-event endpoints.
• To enhance the efficiency and statistical power of biosimilar evaluations.
• To provide a robust framework for demonstrating biosimilarity using historical data.

Main Methods:

• Development of a Bayesian adaptive design incorporating a calibrated power prior for historical data utilization.
• Implementation of a two-stage procedure with a Bayesian biosimilarity index (BBI) for potential early stopping.
• Extensive simulations to evaluate the operating characteristics of the proposed design against naive methods.

Main Results:

• The proposed Bayesian adaptive design demonstrates favorable operating characteristics in simulations.
• The calibrated power prior effectively incorporates historical data for the reference drug.
• The two-stage procedure with BBI allows for efficient early stopping when biosimilarity is established.

Conclusions:

• The proposed Bayesian adaptive design offers an efficient and statistically sound approach for biosimilarity trials with time-to-event data.
• This novel design improves upon existing methods by leveraging historical data and enabling early trial termination.
• The framework provides a valuable tool for regulatory submissions and the development of biosimilar medicines.