PURPOSE: Revascularization strategies in patients with ischemic heart failure (HF) should be based on evidence of reversible perfusion defects and myocardial viability. Myocardial viability assessment is preferably based on dual isotope PET using perfusion and metabolism tracers. However, in a nonnegligible subset of HF patients, reverse mismatch (RM) pattern (reduced glucose uptake relative to perfusion) of unknown origin is observed. We aimed to investigate determinants of RM and the impact of RM on the subsequent improvement in left ventricular function by revascularization. PATIENTS AND METHODS: Ninety-one patients (12 women, 25 with diabetes) with HF undergoing Rb perfusion PET and hyperinsulinemic-euglycemic clamp F-FDG viability PET were retrospectively reviewed. RESULTS: Follow-up time was 12 to 33 months. In 30 of 91 patients, hypometabolic myocardium exceeded the percentage of hypoperfused myocardium; however, only in 12 of 91 patients was the RM considered significant (percentage RM in the left ventricle, 42.5 ± 12.9 [reverse patients] vs 14.1 ± 8.6 [scar and hibernation patients]; P < 0.001). Diabetes status per se did not predict RM, but a significant inverse correlation between insulin sensitivity and RM was observed. The frequency of hospitalization, cardiac death, and myocardial infarctions were not significantly higher in RM patients. Reverse mismatch patients benefited from revascularization to the same extent as patient with normal metabolic patterns. CONCLUSIONS: Reverse mismatch is common among HF patients (~15%) and is inversely correlated to insulin sensitivity. It is not, however, associated with increased cardiac morbidity and mortality and does not predict a worse outcome after revascularization.
PURPOSE: Revascularization strategies in patients with ischemic heart failure (HF) should be based on evidence of reversible perfusion defects and myocardial viability. Myocardial viability assessment is preferably based on dual isotope PET using perfusion and metabolism tracers. However, in a nonnegligible subset of HF patients, reverse mismatch (RM) pattern (reduced glucose uptake relative to perfusion) of unknown origin is observed. We aimed to investigate determinants of RM and the impact of RM on the subsequent improvement in left ventricular function by revascularization. PATIENTS AND METHODS: Ninety-one patients (12 women, 25 with diabetes) with HF undergoing Rb perfusion PET and hyperinsulinemic-euglycemic clamp F-FDG viability PET were retrospectively reviewed. RESULTS: Follow-up time was 12 to 33 months. In 30 of 91 patients, hypometabolic myocardium exceeded the percentage of hypoperfused myocardium; however, only in 12 of 91 patients was the RM considered significant (percentage RM in the left ventricle, 42.5 ± 12.9 [reverse patients] vs 14.1 ± 8.6 [scar and hibernation patients]; P < 0.001). Diabetes status per se did not predict RM, but a significant inverse correlation between insulin sensitivity and RM was observed. The frequency of hospitalization, cardiac death, and myocardial infarctions were not significantly higher in RM patients. Reverse mismatch patients benefited from revascularization to the same extent as patient with normal metabolic patterns. CONCLUSIONS: Reverse mismatch is common among HF patients (~15%) and is inversely correlated to insulin sensitivity. It is not, however, associated with increased cardiac morbidity and mortality and does not predict a worse outcome after revascularization.