Long-term predictors of morbidity and mortality in patients following LVAD replacement

Background: As heart transplant guidelines evolve, the clinical indication for 73% of durable left ventricular assist device (LVAD) implants is now destination therapy. Although completely magnetically levitated LVAD devices have demonstrated improved durability relative to previous models, LVAD replacement procedures are still required for a variety of indications. Thus, the population of patients with a replaced LVAD is growing. There is a paucity of data regarding the outcomes and risk factors for those patients receiving first-time LVAD replacements.Methods: The study cohort consisted of all consecutive patients between 2006 and 2020 that received a first-time LVAD replacement at a single institution. Preoperative clinical and laboratory variables were collected retrospectively. The primary endpoint was death or need for an additional LVAD replacement. Data were subjected to Kaplan-Meier, univariate, and multivariate Cox hazard ratio analyses.Results: In total, 152 patients were included in the study, of which 101 experienced the primary endpoint. On multivariate analysis, patients receiving HeartMate 3 (HM3) LVADs as the replacement device showed superior outcomes (HR 0.15, 95% CI 0.065-0.35, p < 0.0001). Independent risk factors for death or need for additional replacement included preoperative extracorporeal membrane oxygenation (ECMO) (HR 4.44, 95% CI 1.87-14.45, and p = 0.00042), increased number of sternotomies (HR 5.20, 95% CI 1.87-14.45, and p = 0.0016), and preoperative mechanical ventilation (HR 1.98, 95% CI 1.01-3.86, and p = 0.045).Conclusions: Replacement with HM3 showed superior outcomes compared to all other pump types when controlling for both initial pump type and other independent predictors of death or LVAD replacement. Preoperative ECMO, mechanical ventilation, and multiple sternotomies also increased the odds for death or the need for subsequent replacement.