Optimal Cutoff Value of Fractional Flow Reserve Derived From Coronary Computed Tomography Angiography for Predicting Hemodynamically Significant Coronary Artery Disease

Background: The optimal cutoff value of fractional flow reserve (FFR) derived from coronary computed tomography angiography (FFRCT) remains unclear. Methods: The current study population consisted of 93 patients with 139 vessels, who had suspected coronary artery disease by computed tomography angiography and underwent invasive FFR. We evaluated diagnostic performance of FFRCT according to different FFRCT cutoff values and FFRCT ranges with invasive FFR <= 0.80 as the reference standard. Results: In per-vessel analysis, median invasive FFR was 0.85 (interquartile range, 0.75-0.90), and 57 out of 139 vessels (41%) showed hemodynamically significant stenosis (<= 0.80). Median FFRCT was 0.77 (interquartile range, 0.66-0.84; mean difference [invasive FFR-FFRCT], 0.06 +/- 0.11). Per-vessel accuracy, sensitivity, specificity, positive predictive value, and negative predictive value were 73%, 95%, 59%, 61%, and 94% for the cutoff value of FFRCT <= 0.80, 81%, 86%, 78%, 73%, and 89% for FFRCT <= 0.75, and 83%, 74%, 89%, 82%, and 83% for FFRCT <= 0.70, respectively. Per-vessel accuracy across the different ranges of FFRCT <= 0.60, 0.61 to 0.70, 0.71 to 0.80, 0.81 to 0.90, and >0.90 with the cutoff value of FFRCT <= 0.80 were 95%, 74%, 32%, 93%, and 100%, respectively. Setting a gray zone of FFRCT 0.71 to 0.80 provided high positive predictive value (82%; n=42/51) in the range of FFRCT <= 0.70 and high negative predictive value (94%; n=48/51) in FFRCT >0.80. Conclusions: This study suggested that referral to invasive coronary angiography should be considered individually in the range of FFRCT 0.71 to 0.80, whereas dichotomous decision could be made in FFRCT <= 0.70 and >0.80. Future prospective studies evaluating clinical outcomes are needed to establish optimal FFRCT-based diagnostic algorithm.