Hemolysis, pump thrombus, and neurologic events in continuous-flow left ventricular assist device recipients

An ongoing challenge in the management of patients with heart failure who receive left ventricular assist devices (LVADs) is achieving optimal anticoagulation. Adverse prothrombotic events include hemolysis or pump thrombus (H/T) and neurologic events (NEs), and all limit the success of LVAD therapy. Our aim was to study the incidence and clinical outcomes associated with these events in a large single-center cohort.

We retrospectively reviewed our prospectively collected database of all patients receiving a HeartMate II (Thoratec Corp, Pleasanton, CA) LVAD from 2005 to 2012. Demographic, clinical, and outcome data were analyzed using standard statistical methods. All adverse events were recorded.

Of 193 patients receiving LVADs, we identified 39 H/T events in 26 (13.4%) patients and 22 NEs in 19 (9.8%) patients. Seventy-four percent of events occurred in the last 3 years of the series, during which time 63% of implants were placed. Of patients with H/T, 8 (31% of those having H/T, 4.1% of total) had more than 1 event and 4 (15.4% of those having H/T, 2.1% of total) underwent pump exchanges. Five (23%) patients had NEs after H/T, and 6 (32%) died as a result of the NE. Of patients with H/T, 27% had preceding episodes of infection, 31% had an international normalized ratio (INR) of less than 1.5, 31% had an INR of 1.5 to 2, 15% had a history of clotting or were hypercoagulable, and 4% had anticoagulation intentionally withheld. Lactate dehydrogenase (LDH), plasma hemoglobin, INR, and platelet determinations were significantly different at the time of H/T compared with baseline values. The survival at 6 months (alive or having undergone transplantation) for those with a prothrombotic event compared with those without was 70% versus 75.2% (p = 0.5).

The incidence of H/T or NEs is significant and results in major morbidity after LVAD placement. Infection and suboptimal anticoagulation are associated with the majority of these events. Identification of patients at higher risk for hemolysis (ie, infection) may allow for modification of anticoagulation regimens to reduce these risks and improve clinical outcomes.