Virtual Resting Pd/Pa From Coronary Angiography and Blood Flow Modelling: Diagnostic Performance Against Fractional Flow Reserve

Abstract

Fractional flow reserve (FFR) has been established as a useful diagnostic tool. The distal coronary pressure to aortic pressure (Pd/Pa) ratio at rest is a simpler physiologic index but also requires the use of the pressure wire, whereas recently proposed virtual functional indices derived from coronary imaging require complex blood flow modelling and/or are time-consuming. Our aim was to test the diagnostic performance of virtual resting Pd/Pa using routine angiographic images and a simple flow model. Three-dimensional quantitative coronary angiography (3D-QCA) was performed in 139 vessels (120 patients) with intermediate lesions assessed by FFR. The resting Pd/Pa for each lesion was assessed by computational fluid dynamics. The discriminatory power of virtual resting Pd/Pa against FFR (reference: ≤0.80) was high (area under the receiver operator characteristic curve [AUC]: 90.5% [95% CI: 85.4-95.6%]). Diagnostic accuracy, sensitivity and specificity for the optimal virtual resting Pd/Pa cut-off (≤0.94) were 84.9%, 90.4% and 81.6%, respectively. Virtual resting Pd/Pa demonstrated superior performance (p<0.001) versus 3D-QCA %area stenosis (AUC: 77.5% [95% CI: 69.8-85.3%]). There was a good correlation between virtual resting Pd/Pa and FFR (r=0.69, p<0.001). Virtual resting Pd/Pa using routine angiographic data and a simple flow model provides fast functional assessment of coronary lesions without requiring the pressure-wire and hyperaemia induction. The high diagnostic performance of virtual resting Pd/Pa for predicting FFR shows promise for using this simple/fast virtual index in clinical practice.