LAUSR

BACKGROUND There are conflicting data on the impact of implantable cardioverter-defibrillator (ICD) shocks on subsequent mortality. OBJECTIVES The aim of this study was to determine whether the arrhythmic substrate leading to ICD therapy or the therapy itself increases mortality. METHODS The study cohort included 5,516 ICD recipients who were enrolled in 5 landmark ICD trials (MADIT-II, MADIT-RISK, MADIT-CRT, MADIT-RIT, RAID). The authors evaluated the association of device therapy with subsequent mortality in 4 separate time-dependent models: model I, type of ICD therapy; model II, type of arrhythmia for which ICD therapy was delivered; model III, combined assessment of all arrhythmia and therapy types during follow-up; and model IV, incremental risk associated with repeated ICD shocks. RESULTS When analyzed by the type of ICD therapy (model I), a first appropriate ICD shock was associated with increased risk of subsequent mortality with or without concomitant occurrence of inappropriate shock during follow-up (hazard ratio [HR]: 2.78 and 2.31; p < 0.001 and p = 0.12), whereas inappropriate shock alone was not associated with mortality risk (HR: 1.23; p = 0.42). Similarly, ICD therapy for ventricular tachycardia (VT) ≥200 beats/min or ventricular fibrillation (VF) (model II) was associated with increased risk of death with or without concomitant therapy for VT <200 beats/min (HRs: 2.25 and 2.62; both p < 0.001), whereas appropriate therapy for VT <200 beats/min or inappropriate therapy (regardless of etiology) did not reach statistical significance (all p > 0.10). Combined assessment of all therapy and arrhythmia types during follow-up (model III) showed that appropriate ICD shocks for VF, shocks for fast VT (≥200 beats/min) without prior antitachycardia pacing (ATP), as well as shocks for fast VT delivered after failed ATP, were associated with the highest risk of subsequent death (HR: all >2.8; p < 0.001). Finally, 2 or more ICD appropriate shocks were not associated with incremental risk to the first appropriate ICD shock (model IV). CONCLUSION The combined data from 5 landmark ICD trials suggest that the underlying arrhythmic substrate rather than the ICD therapy is the more important determinant of mortality in ICD recipients.