Surrogate end-points: the case of trials on coronary atherosclerotic plaque regression.

The experimental progression-regression human model offers an excellent opportunity for testing the pathophysiological hypotheses that have arisen in the study of atherosclerosis. Nevertheless, it is still necessary to demonstrate that the modifications of angiographic patterns are strictly related to changes in coronary events, and this evidence is not yet supported to date by performed trials. Ten randomized controlled trials designed with the aim of angiographically evaluating the evolution of coronary lesions in patients with coronary heart disease made over the past 13 years were reviewed to perform a meta-analysis. Quality of design and execution of the trials, and the analysis of the primary and secondary end-points chosen for each trial, as well as the methods used to measure the changes in atherosclerotic plaques were also assessed. The phenomenon of plaque regression in human beings seems to occur, but clinical, methodological and physiopathologic heterogeneity between examined trials made it unfeasible to perform a meta-analysis. Coronary events are the consequence of the interaction of known and unknown factors, and coronary arterial narrowing from a mild to moderate degree is only one of these factors. The choice of the regression of atherosclerotic plaques as the primary end-point of a trial should probably be confined to phase 2 studies, while proof of clinical efficacy should be based on harder end-points that are representative of the very objective of medical interventions: amelioration and/or prolongation of a patient's life.