Utility of 18 F-Flurpiridaz PET Relative Flow Reserve in Differentiating Obstructive From Nonobstructive Coronary Artery Disease

BACKGROUND: Absolute quantification of myocardial blood flow (MBF) on positron emission tomography perfusion imaging improves the identification of coronary artery disease (CAD). However, distinguishing MBF impairment due to obstructive CAD from nonobstructive CAD remains challenging. We aimed to evaluate the incremental diagnostic value of positron emission tomography derived relative flow reserve (RFR) in the diagnosis of obstructive CAD. METHODS: This is a post hoc analysis of the multicenter phase III trial of 18 F-flurpiridaz positron emission tomography. Patients with available MBF quantification were included. Reduced stress MBF (sMBF) was defined as sMBF below the median (2.2 mL/min per gram). Obstructive CAD on quantitative invasive coronary angiography was defined as ≥70% stenosis. RFR was calculated as a ratio of the minimal segment sMBF over the highest reference vascular territory sMBF. RFR performance for predicting obstructive CAD was evaluated through receiver operating characteristic analysis and the net reclassification index of multivariable regression models. RESULTS: The study included 231 patients (71% male; 56% with established CAD) drawn from the original cohort of 755 trial participants. No patients had 3-vessel CAD. In a per-vessel-based analysis, 82% of vessels with reduced sMBF had no obstructive CAD on invasive coronary angiography. RFR was significantly lower for vessels with obstructive CAD (0.55 versus 0.80; P <0.0001). In vessels with reduced sMBF, RFR was independently associated with obstructive CAD even after accounting for stress total perfusion deficit and myocardial flow reserve (odds ratio, 3.08 95% CI, 1.49–6.38; P =0.002). Although the addition of RFR did not significantly improve discrimination (area under the curve, 0.806 versus 0.822; P =0.11), it significantly improved reclassification of vessels with and without obstructive CAD (net reclassification index, 0.93; obstructive CAD net reclassification index, 0.44; nonobstructive CAD net reclassification index, 0.49; P <0.0001). CONCLUSIONS: RFR provides complementary diagnostic information beyond existing positron emission tomography parameters and may help refine the diagnosis of obstructive CAD in patients with reduced flows. REGISTRATION: URL: https://clinicaltrials.gov ; Unique identifier: NCT01347710.