The value of combined analysis of plaque characteristics and stenosis based on coronary CT angiography in improving CT diagnostic performance for lesion-specific myocardial ischemia / 中华放射学杂志

ABSTRACT

Objective:To probe the diagnostic performance of the combined evaluation of stenosis and plaque characteristics based on coronary computed tomography angiography (CCTA) in identification of myocardial ischemic lesions, using the invasive coronary angiography (ICA)-based fractional flow reserve (FFR) as the gold standard.Methods:From November 2018 to March 2020, the patients with suspected or known coronary artery disease and scheduled for ICA at 5 clinical trials centers were enrolled in this study. All the patients underwent CCTA, ICA and FFR in turn in one week. The luminal stenosis and plaque characteristics were measured and assessed including plaque burden, volume ratios of calcification and non-calcification, lesion length and CT vulnerable features. All culprit vessels were divided into FFR≤0.8 and FFR>0.8 groups, and the parameters of plaque characteristics were compared. The correlation of ischemic lesions with CCTA stenosis and plaque characteristics was analyzed by the logistic regression analysis. The ROC curve was used to evaluate the sensitivity and specificity of CCTA stenosis rate and plaque characteristics, meanwhile the area under curve (AUC) of each parameter was compared by Delong test.Results:Three hundred and sixty-six culprit vessels in 317 patients were analyzed in this study (169 vessels in ischemia group and 197 in nonischemia group). The plaque burden [34.3% (30.3%, 38.8%) vs. 32.4% (28.5%, 37.9%); Z=-2.622, P=0.009], proportion of CT vulnerable features [26.9% (45/169) vs.11.7% (23/197); χ 2=15.311, P<0.001] and lesion length [22.1 (14.4, 35.0) mm vs. 17.6 (11.0, 26.0) mm; Z=-4.388, P<0.001] in FFR≤0.8 group were higher than those in FFR>0.8 group. The results of logistic regression analysis revealed that CCTA stenosis, lesion length, and CT vulnerable features were significant predictors for myocardial ischemia (OR values 3.794, 2.461, 1.027; P<0.001, P=0.002, P=0.002). The diagnostic performance of CCTA ≥50% stenosis alone in identification of ischemic lesions was low (AUC=0.625). When it combined high-risk plaque characteristics and lesion length, the AUC was improved to 0.714 with a statistical significance. Conclusions:CCTA stenosis, lesion length, and CT vulnerable features are major predictors in identification of myocardial ischemic lesions, and the combination will significantly improve the diagnostic performance of CCTA ≥50% stenosis.