Impact of Hemolysis During Microaxial Flow Pump Support on Early Outcomes After Durable Left Ventricular Assist Device Implantation

Abstract

The impact of hemolysis during microaxial flow pump (mAFP; Impella, Danvers, Massachusetts, US) support on early outcomes after durable left ventricular assist device (d-LVAD) implantation is unknown. Three hundred and eleven consecutive patients undergoing d-LVAD implantation after mAFP support (Impella 5.0/5.5 72.3%) were retrospectively included. The incidence and predictors of hemolysis (plasma-free hemoglobin >20 mg/dl or lactic dehydrogenase (LDH) >2.5-fold the upper reference limit) before d-LVAD implantation were assessed, along with its impact on early post-d-LVAD outcomes. The primary outcome was a composite of hemocompatibility-related adverse events (HRAEs: stroke/gastrointestinal bleeding/pump thrombosis). Hemolysis occurred in 40.8%. Impella 2.5/CP versus 5.0/5.5 was the single independent predictor of hemolysis (adj-hazard ratio [HR] = 2.68, 95% confidence interval [CI] = 1.04-6.94, p = 0.031). Post-d-LVAD HRAEs occurred more frequently among patients with hemolysis (31.9% vs. 20.6%; p = 0.041), mainly driven by hemorrhagic stroke and gastrointestinal bleeding. At multivariate analysis, hemolysis remained independently associated with HRAEs (adj-HR = 1.62, 95% CI = 1.02-2.58; p = 0.041). Patients with hemolysis were more likely to need a temporary right ventricular assist device following d-LVAD implantation (28.3% vs. 16.8%; p = 0.012), with no difference in mortality (23.6% vs. 21.2%; p = 0.355). In conclusion, among patients undergoing d-LVAD implantation with mAFP bridge, hemolysis is common, occurs more frequently among patients supported with Impella 2.5/CP, and is an independent predictor of post-d-LVAD HRAEs.