Does the Type of Ventricular Assisted Device Influence Survival, Infection, and Rejection Rates Following Heart Transplantation?

Objective: To determine the influence of the type of left ventricular assisted device (LVAD) as predictor of survival, hospitalizations due to infection, and rejection following heart transplantation and to delineate any further predictors of such outcomes. Methods: Patients who received a left ventricular assist device (HeartMate [Thoratec Corp., Pleasanton, CA, USA] or Novacor [World Heart Corporation, Ottawa, Canada]) as a bridge to heart transplantation between October 1991 and June 2004 using the United Network for Organ Sharing (UNOS) Thoracic Registry database were evaluated. Comparison of survival curves between HeartMate and Novacor was performed using Kaplan-Meier survival method and Cox proportional hazard model adjusting for patient demographics and co-morbidities. Infection and rejection rates between the two devices were analyzed using multivariable logistic regression model. Results: The UNOS database provided 1255 patients with HeartMate and 154 patients with Novacor between 1991 and 2004. Unadjusted one- and five-year survivals for HeartMate and Novacor were 0.84 and 0.80 and 0.72 and 0.56, respectively, following heart transplantation. Adjusting for patient demographics and co-morbidities, no statistical significant difference in one-year survival was observed between those who received HeartMate and Novacor (HR = 1.49, p = 0.127). At five years, however, the HeartMate group had higher survival than Novacor (HR = 1.53, p = 0.043). All-time posttransplant hospitalizations due to infection and rejection were similar between HeartMate and Novacor recipients after adjusting for patient case mix. Conclusion: Controlling for patient case mix, there was no survival difference at one year between those who received Novacor versus HeartMate LVAD. At five years, a Novacor bridge to transplant patient on average had statistically significantly lower survival than HeartMate recipients. No difference in infection and rejection were observed. (J Card Surg 2009;24:250-255).