Diagnostic Value of Relative Flow Reserve in Patients With Prior Coronary Artery Disease: A Post Hoc Analysis of the...

The relative flow reserve (RFR) derived from quantitative myocardial perfusion imaging is the ratio of absolute myocardial perfusion in a stenotic to normally perfused area and is considered the noninvasive equivalent of fractional flow reserve (FFR). In patients with prior coronary artery disease (CAD), detecting hemodynamically significant CAD using hyperemic myocardial blood flow (hMBF) is complicated by diffuse CAD and microvascular disease. In these patients, RFR may improve the diagnostic performance of myocardial perfusion imaging. Therefore, we studied the diagnostic value of RFR over hMBF in patients with prior CAD.

The PACIFIC-2 trial (functional stress imaging to predict abnormal coronary FFR) included symptomatic patients with prior myocardial infarction and/or percutaneous coronary intervention who prospectively underwent [¹⁵O]H₂O positron emission tomography perfusion imaging and invasive coronary angiography with 3-vessel FFR. RFR was assessed using positron emission tomography in an overall cohort incorporating all trial patients, and an optimal cohort of patients with angiographic 1- or 2-vessel disease (diameter stenosis ≥50%) and a nonstenotic reference vessel (diameter stenosis <30%). RFR was calculated as the ratio between the lowest to highest regional hMBF (overall cohort), or the lowest hMBF of a stenotic to the reference area (optimal cohort). Position emission tomography-derived flow indices were referenced by invasive FFR (≤0.80 deemed hemodynamically significant).

The overall cohort included 187 patients (63±9.3 years, 36 [19%] female), and the optimal cohort 80 patients (62±9.6 years, 19 [24%] female). Significant CAD was present in 87 (47%) and 43 (54%) patients, respectively. Correlations between RFR and FFR were 0.42 and 0.52 (P<0.001 for both). C statistics for hMBF and RFR were comparable in the overall (0.81 versus 0.78; P=0.288) and the optimal cohort (0.79 versus 0.82; P=0.512).

RFR proves clinically applicable, even without specific patient selection and knowledge of the coronary anatomy. However, RFR does not outperform absolute hyperemic myocardial perfusion for detecting FFR-defined significant CAD in patients with prior CAD and recurrence of symptoms.