The reverse flow-metabolism mismatch pattern seen in myocardial viability PET scans reflects altered substrate utilization

Abstract

Background: Cardiac PET utilizing N-13 ammonia (NH 3) to quantify flow and F-18 fluorodeoxyglucose (FDG) to assess metabolism is often used to assess for viable myocardium in patients with ischemic cardiomyopathy. The flow-metabolism patterns characteristic of normal (NL) [normal NH 3/ normal FDG] and hibernating (HIB) [↓ NH 3/ normal or ↑ FDG], and infarcted (INF) [↓ NH 3/ ↓ FDG] myocardium are well established. However, a fourth “reverse mismatch” pattern (RMM) [normal NH3 / ↑ FDG] is often seen, but its etiology is not well understood. We postulated that RMM does not result from improper scaling of PET images, but reflects a true defect in FDG uptake in normally perfused, dysfunctional regions. Methods: The PET scans from 18 patients showing both NL and RMM regions were quantitatively analyzed. Tracer activity (counts/pixel) and MBF (ml/g/min) were measured in identical 1-cm regions of interest (ROI's), selected in myocardial segments identified by experienced readers as NL, RMM, HIB or INF. Values were normalized to the most normal ROI for that patient. Results: A total of 81 ROI's were analyzed. Compared to NL, flow in RMM was somewhat decreased (p = 0.03), but was significantly greater than in HIB and INF (p <0.005). Uptake of FDG in NL and HIB was equal, but was decreased in RMM and INF. Whereas the normalized FDG:NH 3ratio was near unity in NL and INF, it was decreased in RMM and increased in HIB. * NL (n = 23) RMM (n = 27) HIB (n = 10) INF (n = 21) Normalized MBF 0.95 ± 0.12 0.84 ± 0.22 * 0.55 ± 0.20 * # 0.40 ± 0.18 * # Normalized FDG Activity 0.97 ± 0.08 0.66 ± 0.12 * 1.07 ± 0.35 * 0.47 ± 0.16 * # Normalized NH 3 Activity 0.97 ± 0.07 0.94 ± 0.08 0.63 ± 0.18 * # 0.47 ± 0.15 * # Normalized FDG:NH 3 Ratio 0.99 ± 0.08 0.71 ± 0.13 * 1.77 ± 0.53 * # 1.06 ± 0.40 # Data are mean ± SD of ROI values as % of corresponding value measured in the most normal ROI for the same patient: * p < 0.05 vs. NL; p < 0.05 vs. HIB # p < 0.005 vs. RMM Conclusions: The RMM pattern often seen in PET viability studies can be mathematically described by an FDG:NH 3ratio <1.0. It reflects a deficit in FDG uptake that is out of proportion to the slight decrease in perfusion to affected regions, not from improper image scaling. This pattern stands in marked contrast to the HIB pattern and likely reflects diminished glucose uptake in regions with viable, but jeopardized, myocardium.