Original Article
Accuracy of 3-dimensional and 2-dimensional quantitative coronary angiography for predicting physiological significance of coronary stenosis: a FAVOR II substudy
Abstract
Background: Three-dimensional quantitative coronary angiography (3D-QCA) enables reconstruction of a coronary artery in 3D from two angiographic image projections. This study compared the diagnostic accuracy of 3D-QCA vs. 2-dimensional (2D) QCA in predicting physiologically significant coronary stenosis, using fractional flow reserve (FFR) as the reference standard.
Methods: All interrogated vessels in the FAVOR II China study and the FAVOR II Europe-Japan study were assessed by 2D-QCA and 3D-QCA according to standard operating procedures in core laboratories. QCA analysts were blinded to the corresponding FFR values.
Results: A total of 645 vessels from 576 patients with 3D-QCA, 2D-QCA, and FFR were analyzed. Using the conventional cut-off value of 50% for percent diameter stenosis (DS%), 3D-QCA was more accurate in predicting FFR ≤0.80 than 2D-QCA [accuracy 74.0% (95% CI: 69.9–77.7%) vs. 64.9% (95% CI: 61.3–68.7%), difference: 9.1%, P<0.001]. Sensitivity was higher by 3D-QCA compared with 2D-QCA [69.1% (95% CI: 63.0–75.1%) vs. 47.1% (95% CI: 40.5–53.6%), difference: 22.0%, P<0.001] and specificity was similar [76.5% (95% CI: 72.5–80.6%) vs. 74.4% (95% CI: 70.2–78.6%), difference: 2.1%, P=0.40]. Area under the receiver operating characteristic curve was significantly higher for 3D-QCA than for 2D-QCA [0.81 (95% CI: 0.77–0.84) vs. 0.66 (95% CI: 0.62–0.71), P<0.001].
Conclusions: 3D-QCA demonstrated better diagnostic performance in predicting physiologically significant coronary stenosis compared with 2D-QCA, when FFR was used as the reference standard.
Methods: All interrogated vessels in the FAVOR II China study and the FAVOR II Europe-Japan study were assessed by 2D-QCA and 3D-QCA according to standard operating procedures in core laboratories. QCA analysts were blinded to the corresponding FFR values.
Results: A total of 645 vessels from 576 patients with 3D-QCA, 2D-QCA, and FFR were analyzed. Using the conventional cut-off value of 50% for percent diameter stenosis (DS%), 3D-QCA was more accurate in predicting FFR ≤0.80 than 2D-QCA [accuracy 74.0% (95% CI: 69.9–77.7%) vs. 64.9% (95% CI: 61.3–68.7%), difference: 9.1%, P<0.001]. Sensitivity was higher by 3D-QCA compared with 2D-QCA [69.1% (95% CI: 63.0–75.1%) vs. 47.1% (95% CI: 40.5–53.6%), difference: 22.0%, P<0.001] and specificity was similar [76.5% (95% CI: 72.5–80.6%) vs. 74.4% (95% CI: 70.2–78.6%), difference: 2.1%, P=0.40]. Area under the receiver operating characteristic curve was significantly higher for 3D-QCA than for 2D-QCA [0.81 (95% CI: 0.77–0.84) vs. 0.66 (95% CI: 0.62–0.71), P<0.001].
Conclusions: 3D-QCA demonstrated better diagnostic performance in predicting physiologically significant coronary stenosis compared with 2D-QCA, when FFR was used as the reference standard.
