Verification of Coronary Computed Tomography-Derived Fractional Flow Reserve Measurement Site for Detection of Significant Coronary Artery Disease

Background: The optimal site for measuring computed tomography (CT)-derived fractional flow reserve (FFR(CT)) to detect significant coronary artery disease (CAD) remains unknown. We investigated how diagnostic performance changes with FFR(CT) measurement site. Methods and Results: The diagnostic performance of FFR(CT), measured 1–2 cm distal to the stenosis vs. a far-distal site, in detecting significant CAD with invasive fractional flow reserve ≤0.8 was evaluated in 254 diseased vessels from 146 patients with stable or suspected CAD diagnosed by coronary CT angiography. Receiver operating characteristic curve analysis revealed a significantly larger area under the curve for FFR(CT) measured 1–2 cm distal to the stenosis than at a far-distal site (0.829 vs. 0.791, respectively; P=0.0305). The rate of reclassification of positive FFR(CT) was 19% for measurements made 1–2 cm distal to the stenosis, and diagnostic accuracy for FFR(CT) 0.71–0.80 improved from 36% to 58% (P=0.0052). Vessel-based diagnostic accuracy of FFR(CT) 1–2 cm distal to the stenosis and at a far-distal site was 75% and 65%, respectively (P<0.0001), with corresponding sensitivity of 87% and 94% (P=0.0039), specificity of 60% and 29% (P<0.0001), a positive predictive value of 73% and 62% (P=0.028), and a negative predictive value of 78% and 79% (P=0.958). Conclusions: Our data suggest measuring FFR(CT) 1–2 cm distal to the stenosis has better diagnostic performance for detecting physiologically significant CAD.