Objectives: The aim of this study was to determine whether the extent of viability or scar is important in the amount of recovery of left ventricular (LV) function, and to develop a model for predicting recovery after revascularization that could be tested in a randomized trial.
Background: F-18-fluorodeoxyglucose (FDG) positron emission tomography (PET) is used to define viable myocardium in patients with coronary artery disease (CAD) and severe LV dysfunction and to guide revascularization decisions. Whether this approach improves clinical outcomes has not been tested in a randomized trial. Before doing so, an objective model for prediction of recovery is required.
Methods: A total of 82 patients with CAD and an ejection fraction (EF) < or =35% had FDG PET perfusion imaging before revascularization. Complete follow-up was available on 70 patients (86%). Patients had radionuclide angiograms at baseline and three months post-revascularization.
Results: Diabetes (p = 0.029), time to operation (p = 0.008), and scar score (p = 0.001) were significant independent predictors of the change in EF. Previous coronary artery bypass graft confounded the effect of age. There was a significant interaction between the perfusion tracer used and mismatch score (p = 0.02). The multivariable prediction model incorporating PET and clinical variables had a goodness of fit with p = 0.001. Across tertiles of scar scores (I, small: 0% to 16%; II, moderate: 16% to 27.5%; III, large: 27.5% to 47%), the changes in EFs were 9.0 +/- 1.9%, 3.7 +/- 1.6%, and 1.3 +/- 1.5% (p = 0.003: I vs. III), respectively.
Conclusions: In patients with severe LV dysfunction, the amount of scar was a significant independent predictor of LV function recovery after revascularization. A combination of PET and clinical parameters predicts the degree of recovery. This model is being applied in a large randomized controlled trial to determine the effectiveness of therapy guided by FDG PET.