Optimal grading of stenosis severity on coronary computed tomographic angiography in patients with suspected coronary artery disease

Abstract

Background: Several multi-point stenosis grade reporting systems on coronary computed tomographic angiography (cCTA) have been suggested to provide information in therapeutic decision-making. The aim of this study is to evaluate current reporting systems of cCTA to find out the most optimal model in predicting revascularization therapy.Methods: Between January 2006 and March 2009, we retrospectively enrolled patients who underwent cCTA for the evaluation of clinically suspected coronary artery disease (CAD) and subsequent invasive coronary angiography (ICA) within 6 months. cCTA stenosis severity was measured semi-quantitatively on a 3-D workstation, and was reclassified according to four different reporting systems: (1) model A (<50; ≥50%), (2) B (<40; 40-69; ≥70%), (3) C (<30; 30-49; 50-69; ≥70%), and (4) D (<25; 25-49; 50-74; ≥75%). Endpoint was defined as significant CAD and revascularization at the discretion of physician based on ICA finding. The area under the curves (AUC) on receiver operating characteristic (ROC) curves was compared to each other.Results: Four hundred and eighty-five patients (male 66%, 63±11 years) were included for final analysis. Three hundred and ninety-eight (82%) patients had significant CAD and 312 (64%) patients underwent revascularization. Among four different multi-tiered stenosis grading models, model C had the largest AUC in the prediction of significant CAD (0.869) and revascularization (0.796 vs. A=0.726, p<0.001; vs. B=0.785, p=0.018; vs. D=0.783, p=0.048). When the pretest probability was combined with cCTA grading, its AUC demonstrated better predictive value (AUC 0.831, p < 0.001).Conclusion: Multi-grade reporting system of cCTA with cut-off of 30, 50, and 70% stenosis is the most optimal method to identify patients who need revascularization therapy. The pretest probability of CAD has incremental value to cCTA stenosis grading in predicting revascularization.