More Disease, Fewer Arrhythmic Events: A Competing‐Risk Signal for SGLT2 Inhibitors in ICD Recipients

INTRODUCTION: SGLT2 inhibitors (SGLT2i) reduce heart failure events across the ejection fraction spectrum, but their association with device-adjudicated ventricular arrhythmias in real-world primary-prevention ICD recipients remains poorly characterized. METHODS AND RESULTS: In a prospectively maintained primary-prevention ICD registry (n = 62; January 2021 onward), 36 patients (58%) were receiving SGLT2i at implantation. The primary endpoint was first appropriate ICD shock for sustained ventricular tachyarrhythmia; death without prior shock was the pre-specified competing event. Cumulative incidence functions were estimated by the Aalen-Johansen method. Over a median follow-up of 752 days, 22 first appropriate shocks occurred. SGLT2i-treated patients had more advanced structural disease (median LVEF 31% vs. 51%) and lower baseline arrhythmic burden (NSVT 36% vs. 69%) than untreated patients. Cumulative incidence of appropriate shock differed significantly between groups (Gray test p = 0.008). In a cause-specific Cox model adjusted for LVEF and amiodarone use, SGLT2i was associated with a lower hazard of appropriate shock (HR 0.34; 95% CI 0.13-0.91; p = 0.031), confirmed by Fine-Gray subdistribution analysis (sHR 0.36; 95% CI 0.13-0.99; p = 0.047). Results were consistent in an HFrEF sub-cohort (LVEF 35% or less; HR 0.24; 95% CI 0.06-0.92; p = 0.037). Adjustment for baseline NSVT attenuated the association to non-significance (HR 0.43; p = 0.12), reflecting confounding by arrhythmic substrate. CONCLUSION: In this single-center competing-risk analysis, SGLT2i use at ICD implantation was associated with fewer device-adjudicated arrhythmic events. Whether this reflects direct electrophysiological modification or favorable patient selection requires confirmation in larger cohorts.