How Accurate is the Fractional Flow Reserve Derived From Routine Coronary Angiography? Results from TCT 2018, FAST-FFR study: Accuracy of Fractional Flow Reserve Derived From Coronary Angiography

The FAST-FFR study conducted by Fearon and his colleagues, and presented at the annual TCT conference 2018, has shown that Fractional Flow Reserve (FFR_angio) measured from the coronary angiogram alone has a high sensitivity, specificity and accuracy compared with pressure-wire derived Fractional Flow Reserve (FFR). The results of this study were published ahead of print in Circulation.

Although FFR with a pressure wire is very useful, is in the guidelines and improves patient outcomes, the uptake has not been seen to be as robust as expected due to limitations of the technique such as wire handling, characteristics, drift, the need for hyperemia etc. Therefore, there was always an impending need to derive FFR from angiography without the insertion of a needle or hyperemia. By eliminating these two requirements, and displaying FFR values of the entire coronary tree, the FFR_angio Accuracy versus Standard FFR (FAST-FFR) study was a prospective, multicenter, international trial which aimed to determine the accuracy of FFR_angio.

In patients with suspected coronary artery disease, routine coronary angiography was performed. Then, a coronary pressure wire and hyperemic stimulus were used to measure FFR in vessels having coronary lesions of varying severity. With the use of proprietary software, operators blinded to FFR then calculated FFR_angio based on angiograms of the respective arteries acquired in at least two different projections. Co-primary endpoints were the sensitivity and specificity of the dichotomously scored FFR_angio for predicting pressure wire-derived FFR using a cutoff value of 0.80. The study was powered to meet prespecified performance goals for sensitivity and specificity. A logistic mixed effects model accounted for multiple lesions per patient in the primary analysis. Pearson’s correlation coefficient between FFR and FFR_angio was reported among vessels with FFR and FFR_angio values above 0.50.

A total of ten centers in the United States, Europe, and Israel enrolled over 300 subjects and vessels. The mean FFR was 0.81 and 43% of vessels had an FFR≤0.80. This study reported that FFR derived from routine coronary angiography (FFR_angio) had a very high sensitivity, specificity, and diagnostic accuracy, all of which are >90% for predicting the reference standard, coronary pressure wire-derived FFR. The diagnostic accuracy of FFR_angio was 92% overall and remained high when only considering FFR values between 0.75-0.85 (87%). Additionally, FFR_angiovalues correlated well with FFR measurements (r=0.80, p<0.001) and the Bland Altman 95% confidence limits were between -0.14 and 0.12. The device success rate for FFR_angio was 99%.

When asked about the future of FFR_angio, Dr. William Fearon commented, “I think FFR_angio will replace a pressure wire, but we obviously need more data. The next step will be to do a clinical outcomes study comparing FFR_angio guided strategy randomized to patients in an invasive pressure wire derived strategy and demonstrate the outcomes.” Indeed, if confirmed in future studies, FFR_angio may eventually replace wire-based FFR measurement and substantially increase physiologic coronary lesion assessment in the catheterization laboratory, potentially contributing to improved patient outcomes.

To view the interview with Dr. C. Michael Gibson, click here.