Background: Temporal patterns of ventricular tachyarrhythmia (VT/VF) have been studied only in patients who have received implantable cardioverter defibrillators (ICD) for secondary prevention of sudden death, and mainly in ischemic patients. The aim of this study was to evaluate VT/VF recurrence patterns in heart failure (HF) patients with biventricular ICD and to stratify results according to HF etiology and ICD indication.
Methods and results: We studied 421 patients (91% male, 66 +/- 9 years). HF etiology was ischemic in 292 patients and nonischemic in 129. ICD indication was for primary prevention in 227 patients and secondary prevention in 194. Baseline left ventricular ejection fraction (LVEF) was 26 +/- 7%, QRS duration 168 +/- 32 msec, and NYHA class 2.9 +/- 0.6. In a follow-up of 19 +/- 11 months, 1,838 VT/VF in 110 patients were appropriately detected. In 59 patients who had > or = 4 episodes, we tried to determine whether VT/VF occurred randomly or rather tended to cluster by fitting the frequency distribution of tachycardia interdetection intervals with exponential functions: VT/VF clusters were observed in 46 patients (78% of the subgroup of patients with > or = 4 episodes and 11% of the overall population). On multivariate logistic analysis, VT/VF clusters were significantly (P < 0.01) associated with ICD indication for secondary prevention (odds ratio [OR] = 3.12; confidence interval [CI] = 1.56-6.92), nonischemic HF etiology (OR = 4.34; CI = 2.02-9.32), monomorphic VT (OR = 4.96; CI = 2.28-10.8), and LVEF < 25% (OR = 3.34; CI = 1.54-7.23). Cardiovascular hospitalizations and deaths occurred more frequently in cluster (21/46 [46%]) than in noncluster patients (63/375 (17%), P < 0.0001).
Conclusions: In HF patients with biventricular ICDs, VT/VF clusters may be regarded as the epiphenomenon of HF deterioration or as a marker of suboptimal response to cardiac resynchronization therapy.