Author + information
We read with interest the comments by Tousoulis and colleagues. They stressed the importance of coronary lesion morphology in predisposing to acute occlusion and myocardial infarction. Indeed, in our study (1), we also found a greater prevalence of border irregularity or ulceration in the 84 culprit as compared with the 291 nonculprit lesions (40% vs. 22%, p < 0.001). It was not mentioned in the published report, however, because this description is visual, subjective and highly dependent on the quality of the radiologic equipment and angiograms, and above all, the intraobserver and interobserver reproducibility in our experience is as low as 60%. Visual assessment of lesion eccentricity has the same limitations. We believe that such visually assessed variables should be analyzed cautiously.
We are also less optimistic than Tousoulis and colleagues in the ability of coronary angiography to identify potentially vulnerable lesions. Indeed, we found strong evidence that the symmetry index, the outflow angle and, to a lesser extent, the percent diameter severity separated future culprit stenoses from stable lesions within the following 36 months, using a univariate approach. However, use of various multivariate models to stratify the risk of individual stenoses and to predict the vulnerable lesion among stenoses of intermediate severity (40% to 70% diameter stenosis) yielded disappointing results, with positive predicting values (PPV) <50% (1). By contrast, stable lesions could be predicted with greater accuracy (PPV 87%). We infer that the occurrence of a future acute occlusion cannot be accurately predicted by angiography, even with the help of quantitative coronary analysis. We therefore certainly agree with Tousoulis and colleagues that use of other invasive or noninvasive techniques, such as those they mentioned, is mandatory to reach this goal and to help improve patient survival and care.
- American College of Cardiology