Experience With an On-Site Coronary Computed Tomography-Derived Fractional Flow Reserve Algorithm for the Assessment of Intermediate Coronary Stenoses
Donnelly PM, Kolossváry M, Maurovich-Horvat P et al.

KEYWORDS
Coronary Computed Tomography-Derived Fractional Flow Reserve; Intermediate Coronary Stenoses

Fractional flow reserve (FFR) derived from coronary computed tomography angiography (CTA) is a new technique for the diagnosis of ischemic coronary artery stenoses. The aim of this prospective study was to evaluate the diagnostic performance of a novel on-site computed tomography-based fractional flow reserve algorithm (CT-FFR) compared with invasive FFR as the gold standard, and to determine whether its diagnostic performance is affected by interobserver variations in lumen segmentation. We enrolled 44 consecutive patients (64.6 ± 8.9 years, 34% female) with 60 coronary atherosclerotic lesions who underwent coronary CTA and invasive coronary angiography in 2 centers. An FFR value ≤0.8 was considered significant. Coronary CTA scans were evaluated by 2 expert readers, who manually adjusted the semiautomated coronary lumen segmentations for effective diameter stenosis (EDS) assessment and on-site CT-FFR simulation. The mean CT-FFR value was 0.77 ± 0.15, whereas the mean EDS was 43.6 ± 16.9%. The sensitivity, specificity, positive predictive value, and negative predictive value of CT-FFR versus EDS with a cutoff of 50% were the following: 91%, 72%, 63%, and 93% versus 52%, 87%, 69%, and 77%, respectively. The on-site CT-FFR demonstrated significantly better diagnostic performance compared with EDS (area under the curve 0.89 vs 0.74, respectively, p <0.001). The CT-FFR areas under the curve of the 2 readers did not show any significant difference (0.89 vs 0.88, p = 0.74). In conclusion, on-site CT-FFR simulation is feasible and has better diagnostic performance than anatomic stenosis assessment. Furthermore, the diagnostic performance of the on-site CT-FFR simulation algorithm does not depend on the readers' semiautomated lumen segmentation adjustments.