Validation Study of Image-Based Fractional Flow Reserve During Coronary Angiography

Abstract

Background— Fractional flow reserve (FFR), an index of the hemodynamic severity of coronary stenoses, is derived from invasive measurements and requires a pressure-monitoring guidewire and hyperemic stimulus. Angiography-derived FFR measurements (FFR angio ) may have several advantages. The aim of this study is to assess the diagnostic performance and interobserver reproducibility of FFR angio in patients with stable coronary artery disease. Methods and Results— FFR angio is a computational method based on rapid flow analysis for the assessment of FFR. FFR angio uses the patient’s hemodynamic data and routine angiograms to generate a complete 3-dimensional coronary tree with color-coded FFR values at any epicardial location. Hyperemic flow ratio is derived from an automatic resistance-based lumped model of the entire coronary tree. A total of 203 lesions were analyzed in 184 patients from 4 centers. Values derived using FFR angio ranged from 0.5 to 0.97 (median 0.85) and correlated closely (Spearman ρ=0.90; P <0.001) with the invasive FFR measurements, which ranged from 0.5 to 1 (median 0.84). In Bland–Altman analyses, the 95% limits of agreement between these methods ranged from −0.096 to 0.112. Using an FFR cutoff value of 0.80, the sensitivity, specificity, and diagnostic accuracy of FFR angio were 88%, 95%, and 93%, respectively. The intraclass coefficient between 2 blinded operators was 0.962 with a 95% confidence interval from 0.950 to 0.971, P <0.001. Conclusions— There is a high concordance between FFR angio and invasive FFR. The color-coded display of FFR values during coronary angiography facilitates the integration of physiology and anatomy for decision making on revascularization in patients with stable coronary artery disease. Clinical Trial Registration— URL: https://www.clinicaltrials.gov . Unique identifier: NCT03005028.