Validation of Surrogate Endpoints

Surrogate Endpoint Validation

Our service for the validation of surrogate endpoints proposes to focus on other, more complete data, when there is a lack of high quality data for measuring efficacy of the most important clinical endpoint.

This service is especially useful when suspecting a strong relationship between the clinically important endpoint (e.g. overall survival) and another endpoint occurring earlier in the disease process (e.g. progression free survival), where we would then conduct an analysis  based on this surrogate endpoint.

In order to guarantee adequate credibility of such an approach, several aspects of the relationship between the surrogate and the clinically important endpoints need to be verified. Two requirements for this approach are:

  • Verification of the relationship between the endpoints (for a given population, indication and class of drugs with similar mechanism of treatment);
  • Prediction of the efficacy on the clinically important endpoint based on surrogate endpoint information.

The following steps are taken for surrogate endpoint validation:

  • Comprehensive literature search for biological plausibility of dependency between surrogate and clinically important endpoints for specified indication;
  • Systematic review of relevant clinical studies restricted to:
    • Indication/population,
    • Class of treatment (of similar mechanism of action);
  • Extraction of information of effects on surrogate and clinically important endpoints;
  • Regression-based model for prediction of the effect on the clinically important endpoint based on knowledge of the effect on a surrogate;
  • Model-based estimation of the effect on the particular treatment;
  • Wide sensitivity analysis, comprehensive discussion of the assumptions and limitations, final conclusions.