The influence of angiographic endpoints on the outcome of lipid intervention studies. A proposal for standardization. REGRESS study group.

ABSTRACT

The aims of this study were to evaluate the influence of different coronary angiographic endpoints on the outcome of lipid intervention studies and to formulate a proposal for angiographic endpoint standardization. In recent angiographic intervention studies a confusing diversity in angiographic endpoints has been used to determine the outcome. In addition, differences in study populations (eg, bypass patients included or not) could influence results. This makes comparisons between studies cumbersome and raises the question to what extent the results of various studies may be subject to the selection of angiographic endpoints. The investigators compared three frequently used endpoints (mean segment diameter, minimum obstruction diameter, and % stenosis) in a group of 505 patients who had just finished a trial designed to assess the effect of cholesterol lowering by pravastatin. To exclude a potential bias this analysis was carried out at the time that the study was still blinded. They found poor intercorrelation coefficients for mean segment diameter calculated in different ways, ranging from 0.55 to 0.95, dependent on inclusion or exclusion of segments influenced by coronary angioplasty or bypass grafting and on whether or not a value of 0 was assumed for occluded segments and segments distal to occlusions. The correlation between mean segment diameter and minimum obstruction diameter was 0.79; between minimum obstruction diamete and % stenosis, 0.85; and between mean segment diameter and % stenosis, only 0.64. Different endpoints sometimes correlate poorly. This may lead to differences in results of angiographic intervention trials. The authors propose standardization by (1) using both mean sement diameter and minimum obstruction diameter as endpoints and (2) excluding from the primary analysis segments influenced by mechanical interventions.