Using a surrogate marker for early testing of a treatment effect

Area of Science:

• Biostatistics
• Clinical Trial Design
• Epidemiology

Background:

• Surrogate markers can potentially reduce costs and follow-up times in clinical studies.
• Assessing treatment effects using surrogate markers, especially with censored data, remains challenging.
• Existing methods lack robust approaches for early treatment effect testing with censored time-to-event data.

Purpose of the Study:

• To propose a novel test statistic for evaluating treatment effects using surrogate marker information in time-to-event outcomes.
• To develop robust nonparametric estimation and inference procedures for this novel test.
• To assess the power of study designs utilizing surrogate marker information for early treatment effect evaluation.

Main Methods:

• Development of a new test statistic for treatment effect assessment using early surrogate marker data.
• Application of robust nonparametric estimation and inference procedures.
• Evaluation of statistical power for future study designs based on surrogate marker data.

Main Results:

• The proposed test statistic effectively utilizes surrogate marker information for early treatment effect testing.
• Simulation studies demonstrate the procedure's validity and power compared to end-of-study tests.
• The method was successfully applied to data from the Diabetes Prevention Program.

Conclusions:

• The novel test statistic offers a valuable tool for earlier assessment of treatment effects in time-to-event studies.
• This approach has the potential to significantly impact clinical trial design by reducing duration and cost.
• The proposed methods provide a robust framework for utilizing surrogate markers in censored time-to-event data analysis.