Abstract 2951: Estimation of Perfusion with 13 N-Ammonia Retention Underestimates the Frequency and Extent of Hibernating Myocardium when Compared to Quantification of Absolute Myocardial Blood Flow

Background: Viable dysfunctional myocardium can be classified as chronically stunned (normal resting perfusion) or hibernating (reduced resting flow). While 13 N-ammonia (NH 3 ) retention (late uptake) is typically used to estimate perfusion, the frequency and extent of hibernating myocardium (HM) may differ when quantification of dynamically-acquired absolute myocardial blood flow (MBF) is performed. Methods: Patients with stable ischemic cardiomyopathy (n = 25, EF 32 ± 2%, NYHA Class 2.1 ± 0.7) who were candidates for an ICD for the primary prevention of sudden death underwent imaging with NH 3 and insulin-clamp 18 F-2-deoxyglucose (FDG). Segmental perfusion (17 segment LV model, % peak segment) was assessed by both retention (20 minute summed image) and absolute MBF using a 3-compartment model (ml/min/g). Normal segments were defined as perfusion ≤ 80% peak segment. Segments with <50% peak segment FDG uptake were defined as scar. The remaining segments were considered HM if FDG/perfusion ratio was ≥ 1.2. Results: Of the 425 total segments, only 15 (3.5%) were considered HM when NH 3 retention was used to assess perfusion. In contrast, the number of HM segments increased markedly with quantification of absolute MBF (159 or 37%, p<0.001 vs. retention), with a commensurate reduction in the number of normally-perfused segments. Conclusions: The estimation of perfusion with NH 3 retention significantly overestimates MBF (Figure ) and hence underestimates the frequency and extent of HM. While the differentiation of chronically stunned from HM may not influence the decision for revascularization, the distinction may be important if HM is an independent risk factor for sudden death.