Doppler Flow Velocity and Thermodilution to Assess Coronary Flow Reserve: A Head-to-Head Comparison With [15O]H2O PET

OBJECTIVES: This study sought to compare Doppler flow velocity reserve (CFRDoppl) and thermodilution-derived coronary flow reserve (CFRthermo) head-to-head with the gold standard for quantification of myocardial perfusion, [15O]H2O positron emission tomography (PET). BACKGROUND: Coronary flow reserve (CFR) is an important parameter for assessing coronary vascular function. To date, 2 techniques are available for invasive assessment of CFR: Doppler flow velocity and thermodilution. Although these techniques have been compared with each other, neither has been compared with [15O]H2O PET perfusion imaging. METHODS: CFR was assessed in 98 vessels of 40 consecutive stable patients with suspected coronary artery disease. Patients underwent [15O]H2O PET, followed by invasive angiography in conjunction with simultaneous measurements of fractional flow reserve, CFRDoppl, and CFRthermo. Both normal and obstructed arteries were included. RESULTS: The quality of Doppler flow velocity traces was significantly lower than that of thermodilution curves (p < 0.001). A moderate correlation was observed between CFRDoppl and CFRthermo (r = 0.59; p < 0.001). CFRDoppl correlated well with PET-derived CFR (CFRPET) (r = 0.82; p < 0.001). In contrast, the correlation between CFRthermo and CFRPET was only modest (r = 0.55; p < 0.001). This difference in correlation with CFRPET was significant (t = 4.9; df = 95; p < 0.001). Bland-Altman analysis revealed a tendency of CFRthermo to overestimate flow reserve at higher values. CONCLUSIONS: Coronary flow reserve, determined using Doppler flow velocity, has superior agreement with [15O]H2O PET in comparison with CFRthermo