Abstract 19526: Fractional Flow Reserved Derived from Computed Tomographic Angiography (FFRCT) to Discriminate Individuals with versus without Ischemia: Results from the DeFACTO Trial (Determination of Fractional Flow Reserve by Anatomic Computed TOmographic Angiography)

Abstract

Background Non-invasive fractional flow reserve (FFR) derived from coronary computed tomographic (CT) angiography (FFRCT) is a novel method that employs computational fluid dynamics to diagnose coronary lesions that cause ischemia. To date, the per-patient diagnostic performance of FFRCT versus obstructive CAD by CT (>50% stenosis) to discriminate ischemia has been inadequately studied.

Methods A total of 407 vessels were evaluated in 252 patients from 17 centers in 4 countries who underwent CT, FFRCT, invasive coronary angiography and invasive FFR. The area under the receiver operator characteristics curve (AUC) for detecting abnormal FFR on a per-patient basis was examined considering: (1) all vessels interrogated by FFR (n=252 patients), (2) only vessels >2 mm interrogated by FFR (n=149 patients); 3) a decision rule which included all vessels interrogated by FFR, as well as vessels with maximal stenosis 90% which were considered negative and positive, respectively, for ischemia (n=252 patients); and 4) a decision rule for >2 mm vessels (n=229 patients). Per-patient AUCs were compared for FFRCT alone, CT stenosis >50% alone, and the combination of FFRCT and CT stenosis >50%. FFRCT was superior to CT stenosis for discrimination of individuals with ischemia (AUC 0.81 vs. 0.69, p2 mm vessels (AUC 0.82 vs. 0.74, p=0.0007), when employing the decision rule (0.83 vs. 0.67, p2 mm vessels (0.92 vs. 0.68, p<0.0001). The combination of FFRCT and CT stenosis was similarly superior to CT stenosis alone when considering all vessels interrogated by FFR (0.81 vs. 0.69, p2 mm vessels (0.82 vs. 0.74, p<0.001), when applying the decision rule (0.83 vs. 0.67, p2 mm vessels (AUC 0.91 vs. 0.68, p<0.0001). In contrast, no added discrimination was noted when CT stenosis was added to FFRCT, as compared to FFRCT alone (p=NS).

Conclusion FFRCT improves discrimination for identification of individuals who manifest ischemia. The addition of CT stenosis findings to FFRCT does not improve discrimination of individuals with ischemia.