LAUSR

Objective Positron emission tomography (PET) integrating assessment of perfusion with 13 N-ammonia (NH3) and viability with 18 F-fluorodeoxyglucose (FDG) has high accuracy to identify viable, hibernating myocardium. We tested whether quantification of myocardial blood flow (MBF) and washout (k2) can predict myocardial viability using FDG as standard of reference. Methods In 180 consecutive patients with ischemic cardiomyopathy, myocardium was categorized on a segment-level into normal, ischemic, hibernating, and scar. From dynamic images, stress MBF, rest MBF, and k2 were derived and myocardial flow reserve (MFR) and volume of distribution (VD) were calculated. Results Across myocardial tissues, all parameters differed significantly. The area under the curve (AUC) was 0.564 (95% CI 0.527-0.601), 0.635 (0.599-0.671), 0.553 (0.516-0.591), 0.520 (0.482-0.559), and 0.560 (0.522-0.597) for stress MBF, rest MBF, MFR, k2, and VD. The generalized linear mixed model correctly classified 81% of scar as viable, hibernating myocardium. If the threshold of rest MBF to predict viability was set to 0.45 mL·min −1 ·g −1 , sensitivity and specificity were 96% and 12%, respectively. Conclusion Quantitative NH3 PET parameters have low to moderate diagnostic performance to predict viability in ischemic cardiomyopathy. However, if rest MBF falls below 0.45 mL·min −1 ·g −1 , viability testing by FDG-PET may be safely deferred.