Catheter Ablation of Ventricular Arrhythmia in Nonischemic Cardiomyopathy

BACKGROUND: The aim was to relate distinct scar distributions found in nonischemic cardiomyopathy with ventricular tachycardia (VT) morphology, late potential distribution, ablation strategy, and outcome. METHODS AND RESULTS: Eighty-seven patients underwent catheter ablation for drug-refractory VT. Based on endocardial unipolar voltage, 44 were classified as predominantly anteroseptal and 43 as inferolateral. Anteroseptal patients more frequently fulfilled diagnostic criteria for dilated cardiomyopathy (64% versus 36%), associated with more extensive endocardial unipolar scar (41 22-83 versus 9 1-29 cm(2); P<0.001). Left inferior VT axis was predictive of anteroseptal scar (positive predictive value, 100%) and right superior axis for inferolateral (positive predictive value, 89%). Late potentials were infrequent in the anteroseptal group (11% versus 74%; P<0.001). Epicardial late potentials were common in the inferolateral group (81% versus 4%; P<0.001) and correlated with VT termination sites (κ=0.667; P=0.014), whereas no anteroseptal patient had an epicardial VT termination (P<0.001). VT recurred in 44 patients (51%) during a median follow-up of 1.5 years. Anteroseptal scar was associated with higher VT recurrence (74% versus 25%; log-rank P<0.001) and redo procedure rates (59% versus 7%; log-rank P<0.001). After multivariable analysis, clinical predictors of VT recurrence were electrical storm (hazard ratio, 3.211; P=0.001) and New York Heart Association class (hazard ratio, 1.608; P=0.018); the only procedural predictor of VT recurrence was anteroseptal scar pattern (hazard ratio, 5.547; P<0.001). CONCLUSIONS: Unipolar low-voltage distribution in nonischemic cardiomyopathy allows categorization of scar pattern as inferolateral, often requiring epicardial ablation mainly based on late potentials, and anteroseptal, which frequently involves an intramural septal substrate, leading to a higher VT recurrence.