Risk of Mortality for Ventricular Arrhythmia in Ambulatory LVAD Patients

BACKGROUND: There are limited data regarding the incidence and prognostic significance of ventricular arrhythmias (VA) in ambulatory continuous flow left ventricular assist device (LVAD) patients. METHODS: Sixty-one consecutive patients from November 1, 2006 through December 31, 2010 with an LVAD and implantable cardioverter defibrillator that survived to discharge from the LVAD implantation admission were studied. Follow-up began from date of discharge with both devices in situ and ended with death, transplant, on June 1, 2011. Pre-LVAD VA history was related to the primary endpoints of post-LVAD VA, mortality, and the combined endpoint of post-LVAD VA/mortality. RESULTS: During a mean follow-up of 622 days 19 patients (31%) experienced VA (14 episodes of VT, 5 episodes of VF). Pre-LVAD VA was predictive of post-LVAD VA (hazard ratio HR 2.91, P = 0.026) and the combined post-LVAD VA/mortality endpoint (HR 2.70, P = 0.021) but only displayed a nonsignificant association with mortality (HR 2.30, P = 0.11). In multivariate analysis, pre-LVAD VA remained a significant predictor of post-LVAD VA (HR 2.84, P = 0.03) and the combined post-LVAD VA/mortality endpoint (HR 2.65, P = 0.025). Post-LVAD VA was the strongest univariate predictor of mortality (HR 13.92, P < 0.001) and remained so after multivariate adjustment (HR 9.69, P = 0.001). Post-LVAD VA occurred at a mean of 1 year from mortality events with 45% within 1 month. CONCLUSIONS: Pre-LVAD VA is a significant predictor of post-LVAD VA but not of mortality. VA in the continuous flow LVAD population carries a significant risk of mortality often within the first month.