Principles of use of surrogate markers and endpoints

Abstract

Surrogate end-points are markers of biological mechanisms and require a mechanistic view on diseases. Thus, the validation of candidate surrogate end-points for the cardiovascular risk of sex steroids requires that a separate evaluation be made of the various clinical endpoints in which different biological mechanisms are likely to operate. The case of venous thromboembolic (VTE) disease and arterial diseases is complex since they are multifactorial and multiple surrogate end-points are likely to be relevant. In addition sex steroids have many effects on biological mechanisms and the selection and validation of surrogates from the available bulk of changes in multiple mechanisms is a large enterprise. In addition, the combination of candidates in algorithms is required and at very early development stages only. In practice, the validation of candidates requires the detection of changes or levels achieved in intervention trials with clinical end-points and such evaluations are scarce. The observations of increased risk of VTE for first-time users of sex steroids indicate the occurrence of a susceptible subgroup, suggesting that effects in this group may be dissimilar to those in the population as a whole or quantitatively strongly different.