DIAGNOSIS OF NON-OBSTRUCTIVE CORONARY ATHEROSCLEROSIS IN MEN WITH SUSPECTED CORONARY ARTERY DISEASE

Abstract

Purpose: This study aimed to develop a method for detection of nonsignificant coronary lesions using clinical and functional parameters in suspected CAD men without obstructive coronary atherosclerosis by CAG. Methods. From 17.784 consecutive patients from the local CAG database we selected men with non-significant coronary lesions (Ј30 % of lumen) and with smooth coronary arteries by CAG with suspected CAD without acute coronary syndrome, unstable angina, myocardial infarction. Selected 1.957 patients were divided into two groups in a random way (1.389 - studied group and 568 - test group). We divided studied group in two subgroups: with nonsignificant coronary lesions (1.085 patients) and with smooth coronary arteries (304 patients). Results: Patients with nonsignificant coronary lesions compared to patient without were older, arterial hypertension, and II-III NYHA classes and angina pectoris were observed more often in these patients. Atherogenic index, echocardiographic indexes of aortic root diameter and left ventricular (LV) mass were higher in patients with nonsignificant coronary lesions, and echocardiographic signs of ascending aorta atherosclerosis were more frequent in these patients. For selection of scoring variables, we used the values of the Cramer coefficient and the information value index - IV. All quantitative variables were converted into categorical using the binning procedure. Based on the values of WoE, the model of logistic regression was constructed using stepwise variable selection. The model includes independent variables with an average and high prognostic ability: age, atherogenic index, arterial hypertension, echocardiographic signs of atherosclerosis of the ascending aorta. By converting the coefficients of the logistic regression equation obtained into scores, a scoring model was developed to assess the probability of non-obstructive coronary atherosclerosis in men. Based on ROC analysis, in studied group the specificity of the model for detection of nonsignificant coronary atherosclerosis was 67 %, the sensitivity was 65 %, the area under the ROC curve - 0.711. For patients of test group sensitivity was 68 % and specificity 63 %, the area under the ROC curve - 0.687. Conclusion. Nonsignificant coronary atherosclerosis can be predicted in men with suspected CAD before CAG by screening test based on age, atherogenic index, arterial hypertension and echocardiographic signs of aortic atherosclerosis.

Keywords

• men suspected coronary artery disease
• nonobstructive coronary atherosclerosis