Background: Clinical outcome can be predicted by metabolism-perfusion positron emission tomography (PET) in patients with severe ischaemic left ventricular dysfunction. This study determined whether the amount of viable or non-viable myocardium detected with a PET scan or clinical-functional parameters might predict cardiovascular events.
Methods: All patients had previous myocardial infarction (>6 months previously) and left ventricular ejection fraction (LVEF) <40%. Metabolism-perfusion PET, echocardiogram and coronary angiography were provided. All subjects underwent short euglycaemic-hyperinsulinaemic clamp before the metabolism study. The dysfunctioning segment was defined as hibernating myocardium when metabolism was normal-moderately reduced with impaired perfusion (mismatch flow-metabolism). Cardiac death, hospital admission for myocardial infarction or heart failure were considered cardiovascular events.
Results: Ninety-three patients (71 males, aged 64.2 years) were studied. The LVEF was 30.2+/-7.7%; 48 (51.6%) suffered an anterior myocardial infarction. Fifty-three (54.1%) subjects were treated with coronary revascularization; all had optimal medical therapy. Cardiovascular events occurred in 20/93 patients at 1-year follow-up (event group). Age (P=0.7), diabetes mellitus (P=0.6) and rate of coronary revascularization (P=0.3) were not different in the two groups. Patients who experienced cardiovascular events had larger non-viable myocardium (5.8+/-2.7 vs. 4.1+/-2.6, P=0.01), lower metabolic rate glucose (1.3+/-0.6 vs. 1.7+/-0.7 ml . kg . min, P=0.04) but similar hibernating myocardium (1.6+/-1.6 vs. 1.7+/-2, P=0.8) and baseline LVEF (28.1+/-4.8 vs. 30.7+/-8.3%, P=0.08). Having more then five non-viable segments and a metabolic rate for glucose of <0.9 mg . kg . min predicted a worse prognosis (P=0.04, log rank, 3.89; and P=0.004, log rank, 8.1, respectively).
Conclusion: Non-viable myocardium revealed with PET predicts mid-term clinical prognosis. Insulin resistance seems to influence the outcome.