Association between coronary sinus flow estimated using dynamic coronary CT angiography and 13N-ammonia PET-derived myocardial flow reserve

Rationale and objectives: PET-derived myocardial flow reserve (MFR) is a powerful prognostic indicator in coronary artery disease. We devised a new method for CS flow quantification using 320-row dynamic coronary CT angiography (CCTA) and investigated the relationship between CS flow determined via CCTA and MFR obtained via 13N-ammonia PET (NH3-PET). Materials and methods: Forty patients with significant coronary stenosis on dynamic CCTA who subsequently underwent adenosine-stress NH3-PET were enrolled. Time-attenuation curves of the CS and the ascending aorta were extracted from dynamic CCTA data. CS start time (second) was defined as the timepoint between the peak aortic attenuation and the initial increase in CS attenuation. The association with patients with MFR < 2.0 and diagnostic performance were compared by CS start time and CAD-RADS (The coronary artery disease report and data system). Results: CS start time was significantly delayed in patients with MFR < 2.0 versus patients with MFR >= 2.0 (0.24 +/- 1.20 versus-1.47 +/- 1.29 sec, p = 0.0002). Multivariate logistic analysis showed that CS start time (odds ratio, 5.191; 95 % confidence interval, 2.038-22.194) and CAD-RADS (odds ratio, 4.207; 95 % confidence interval, 1.319-19.725) were independent associated factors with MFR < 2.0. The diagnostic performance of patients with MFR < 2.0 by receiver-operating-characteristic analysis was an area under the curve 0.841 for CS start time and 0.675 for CAD-RADS. Conclusion: CS flow measurement by dynamic CCTA is a new noninvasive method to estimate MFR. Delayed CS start time indicates a significant decrease in MFR, which is important information for therapeutic intervention in patients with moderate or severe coronary artery stenosis.