BACKGROUND

Impaired renal function is associated with increased mortality among patients with implantable cardioverter-defibrillators (ICDs). The relationship between renal function at time of ICD generator replacement and subsequent appropriate ICD therapies is not known.

METHODS AND RESULTS

We identified 441 patients who underwent first ICD generator replacement between 2000 and 2011 and had serum creatinine measured within 30 days of their procedure. Patients were divided into tertiles based on estimated glomerular filtration rate (eGFR). Adjusted Cox proportional hazard and competing risk models were used to assess relationships between eGFR and subsequent mortality and appropriate ICD therapy. Median eGFR was 37.6, 59.3, and 84.8 mL/min/1.73 m(2) for tertiles 1-3, respectively. Five-year Kaplan-Meier survival probability was 34.8%, 61.4%, and 84.5% for tertiles 1-3, respectively (P < 0.001). After multivariable adjustment, compared to tertile 3, worse eGFR tertile was associated with increased mortality (HR 2.84, 95% CI [1.36-5.94] for tertile 2; HR 3.84, 95% CI [1.81-8.12] for tertile 1). At 5 years, 57.0%, 58.1%, and 60.2% of patients remained free of appropriate ICD therapy in tertiles 1-3, respectively (P = 0.82). After adjustment, eGFR tertile was not associated with future appropriate ICD therapy. Results were unchanged in an adjusted competing risk model accounting for death.

CONCLUSIONS

At time of first ICD generator replacement, lower eGFR is associated with higher mortality, but not with appropriate ICD therapies. The poorer survival of ICD patients with reduced eGFR does not appear to be influenced by arrhythmia status, and there is no clear proarrhythmic effect of renal dysfunction, even after accounting for the competing risk of death.