Cardiac glycosides are not associated with increased mortality or hospitalization rates in ICD and CRT-ICD patients after adjustment for baseline-characteristics at one-year follow-up: Results from the German DEVICE registry

Aims: Despite lacking supporting randomized trials, cardiac glycosides (CGs) are widely used in heart failure and/or atrial fibrillation. Moreover, several pro- and retrospective studies and registry-data have recently raised serious concerns in terms of efficacy and safety of CGs in this field. We have therefore examined the association between CGs and clinical outcome of implantable cardioverter-defibrillator (ICD) and cardiac resynchronization (CRT-ICD) patients of the large German DEVICE registry.

Methods and results: Between 2007 and 2014, 3782 ICD and 1529 CRT-ICD patients were enrolled in the German DEVICE registry. Those two groups were analyzed independently according to medication with or without CGs. After adjustment for patient characteristics, CGs were not significantly associated with increased one-year mortality (HR 1.27, 95%-CI 0.91-1.76, p = 0.162), major adverse cardiac and cerebrovascular events (OR 1.36, 95%-CI 0.98-1.89, p = 0.063), ICD-shocks (OR 1.29, 95%-CI 0.95-1.74, p = 0.104) or the need for rehospitalization in ICD patients at one-year-follow-up. Similar findings were obtained in CRT-ICD patients. Regarding possible determinants for glycoside treatment, atrial fibrillation at enrollment was found to be most strongly associated with the prescription of glycosides in ICD (adjusted OR 3.25, 95%-CI 2.63-4.02) and CRT-ICD patients (adjusted OR 3.17, 95%-CI 2.39-4.19).

Conclusion: Overall harmful effects of CGs in ICD- and CRT-ICD patients could not be confirmed in DEVICE. Further large and randomized-controlled trials that investigate dose-dependent effects of CGs in addition to contemporary therapy of heart failure and atrial fibrillation are needed.