OBJECTIVES: This study was designed to compare head-to-head fractional flow reserve (FFR) derived from coronary computed tomography angiography (CTA) (FFRCT) and cardiac magnetic resonance (CMR) stress perfusion imaging for prediction of standard-of-care-guided coronary revascularization in patients with stable chest pain and obstructive coronary artery disease by coronary CTA.
BACKGROUND: FFRCT is a novel modality for noninvasive functional testing. The clinical utility of FFRCT compared to CMR stress perfusion imaging in symptomatic patients with coronary artery disease is unknown.
METHODS: Prospective study of patients (n=110) with stable angina pectoris and 1 or more coronary stenosis ≥50% by coronary CTA. All patients underwent invasive coronary angiography. Revascularization was FFR-guided in stenoses ranging from 30% to 90%. FFRCT ≤0.80 in 1 or more coronary artery or a reversible perfusion defect (≥2 segments) by CMR categorized patients with ischemia. FFRCT and CMR were analyzed by core laboratories blinded for patient management.
RESULTS: A total of 38 patients (35%) underwent revascularization. Per-patient diagnostic performance for identifying standard-of-care-guided revascularization, (95% confidence interval) yielded a sensitivity of 97% (86 to 100) for FFRCT versus 47% (31 to 64) for CMR, p < 0.001; corresponding specificity was 42% (30 to 54) versus 88% (78 to 94), p < 0.001; negative predictive value of 97% (91 to 100) versus 76% (67 to 85), p < 0.05; positive predictive value of 47% (36 to 58) versus 67% (49 to 84), p < 0.05; and accuracy of 61% (51 to 70) versus 74% (64 to 82), p > 0.05, respectively.
CONCLUSIONS: In patients with stable chest pain referred to invasive coronary angiography based on coronary CTA, FFRCT and CMR yielded similar overall diagnostic accuracy. Sensitivity for prediction of revascularization was highest for FFRCT, whereas specificity was highest for CMR.