Futility Interim Analysis Based on Probability of Success Using a Surrogate Endpoint.

Area of Science:

• Clinical Trial Design
• Biostatistics
• Bayesian Methodology

Background:

• Evaluating long-term treatment efficacy in clinical trials is time-consuming and resource-intensive.
• Surrogate endpoints accelerate decision-making by correlating with primary endpoints.
• Leveraging historical data is crucial for optimizing clinical trial resources and minimizing sample sizes.

Purpose of the Study:

• To develop a Bayesian group sequential design methodology that utilizes historical surrogate endpoint data.
• To build an informative prior for the primary endpoint using early surrogate data from an interim analysis.
• To define a futility-stopping rule based on the predictive probability of trial success.

Main Methods:

• Employs the general theory of group sequential design within a Bayesian framework.
• Exploits documented historical relationships between final and surrogate endpoints.
• Integrates a robust approach combining surrogate prior with a vague component to mitigate prior-data conflicts.

Main Results:

• Demonstrates substantial enhancements in trial operating characteristics with good agreement between current and historical data.
• Maintains acceptable performance even with significant prior-data conflicts.
• Successfully applied to design a Phase III metastatic colorectal cancer trial (OS primary, PFS surrogate).

Conclusions:

• The proposed Bayesian methodology effectively leverages historical surrogate data for efficient clinical trial design.
• The approach enhances trial operating characteristics and provides robust futility-stopping rules.
• This strategy offers significant advantages for trials with long-term endpoints, particularly in oncology.