CTA Protocol and Analysis

CT scans were performed using a 64-row detector CT scanner (LightSpeed VCT; GE Healthcare) with retrospective gating until the 31 of December 2013 and using a 256-row system (Revolution CT; GE Healthcare) with prospective gating thereafter. Image acquisition was also performed for patients in atrial fibrillation and with a heart rate of up to 128 bpm. Intravenous injection of 80–100 mL iohexol (350 mg I/mL, Accupaque® 350, GE Healthcare) into an antecubital vein was performed using a power injector at a flow rate of 5 mL/s, followed by a saline chaser. The bolus tracking technique with manual triggering was applied to start the acquisition. The acquisition parameters were as follows: rotation speed, 0.35 s (64-row system) and 0.28 s (256-row system); tube potential, 100 or 120 kVp depending on body mass index; and tube current, 400–600 mA. Neither heart rate control nor vasodilatation (nitroglycerin) was used prior to the procedure as CTA was not coronary dedicated. The image reconstruction parameters were as follows: section thickness, 0.625 mm; kernel, soft tissue; algorithm, adaptive statistical image reconstruction 50% (64-row system) and adaptive statistical image reconstruction-V 50% (256-row system); display field-of-view, 28 cm; and cardiac phase, end-systolic. In addition to the 28 cm field-of-view, a dedicated cardiac reconstruction with a 20 cm field-of-view was provided for coronary artery analysis. CTA images were retrospectively read by two experienced radiologists, who were blinded from ICA images and reports. To determine the inter-observer variability using the κ of Cohen test, 15% of images were randomly selected and read by both radiologists separately. The four main coronary arteries were evaluated (right coronary artery, left main artery, left circumflex artery, and left anterior descending artery). The quality of each vessel was rated using a 3-point Likert scale in terms of delineation between lumen and wall of the artery, and filling with contrast material: optimal (no artifacts, strong attenuation of vessel lumen), suboptimal (moderate artifacts, low contrast, but acceptable for routine clinical diagnosis), or unanalyzable (severe artifacts or insufficient contrast impairing accurate evaluation). CAD analysis was undertaken on all vessels ≥ 2.5 mm whose quality was rated as optimal or suboptimal. The following classification for CAD was applied: no significant CAD (0–49% lumen diameter reduction), moderate CAD (50–69% lumen diameter reduction), and severe CAD (70–99% lumen diameter reduction or occlusion). If there was more than one stenosis on a vessel, only the most severe lesion was taken into account. The Agatston calcium score was also calculated in the coronary arteries.