Evaluating the Proportion of Treatment Effect Explained by a Continuous Surrogate Marker in Logistic or Probit...

Area of Science:

• Biostatistics
• Clinical Trial Design
• Pharmacometrics

Background:

• Surrogate endpoints accelerate drug development by shortening clinical trials and reducing costs.
• Validating surrogate endpoints is crucial for reliable drug development.
• Current methods for estimating the proportion of treatment effect explained (PTE) by surrogates have limitations, including model validity issues and estimates outside the [0, 1] range.

Purpose of the Study:

• To develop and validate alternative statistical measures for evaluating the proportion of treatment effect explained by surrogate endpoints.
• To provide easily estimable measures within standard statistical software.
• To enhance the interpretability and accuracy of surrogate endpoint validation in clinical trials.

Main Methods:

• Proposed novel measures for estimating the proportion of treatment effect explained (PTE) in logistic or probit regression models.
• Utilized Ordinal Dominance (OD) curves for visual interpretation of the proposed measures.
• Conducted simulations to compare the performance of the new measures against existing methods.

Main Results:

• The alternative measures are easily estimable using standard binary linear regression modeling capabilities.
• Ordinal Dominance (OD) curves provide a visually intuitive method for understanding the measures.
• Simulations demonstrated that the proposed measures offer more accurate estimates with reduced bias, lower variability, and narrower confidence intervals compared to traditional methods.

Conclusions:

• The developed alternative measures provide a more robust and accurate approach to validating surrogate endpoints in clinical trials.
• These methods facilitate more reliable and cost-effective drug development by improving the assessment of treatment effects.
• The proposed measures and visualization tools enhance practical usability for researchers and clinicians involved in clinical trial design and analysis.