Postoperative bridging anticoagulation and left ventricular assist system thrombosis

Abstract

Reduced emphasis on early postoperative bridging anticoagulation is one explanation for the increased incidence of HeartMate II (HM II) pump thrombosis. We conducted a single-center analysis of initial postoperative anticoagulation practices and their impact on the incidence of HM II pump thrombosis. Data was collected on 105 patients undergoing primary HM II implant from 2009 to 2014. A parenteral bridge was defined as use of parenteral anticoagulation prior to attainment of an international normalized ratio ≥ 2 on warfarin. A parenteral bridge was further characterized as early if initiated ≤ 3 days of implant and therapeutic if a manufacturer-specified goal partial thromboplastin time (PTT) was achieved during each of the first 3 days of administration. Pump thrombosis was “suspected” based upon suggestive clinical parameters leading to hospital admission with parenteral anticoagulant administration and “confirmed” by direct visualization of thrombus in the device. A majority of patients (70%) were treated with an initial parenteral bridge, which was started within 3 days of device implantation in 68% of cases. Therapeutic PTT levels were achieved in 52% of patients treated with a parenteral bridge. Patients who were bridged had lower Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) profiles (p = 0.039) and longer intensive care unit stay (p = 0.005). Pump thrombosis was suspected in 25% and confirmed in 13% of patients within 6 months of implant. Point estimates of pump thrombosis incidence demonstrated a lower event rate at 6 months in patients who received a therapeutic bridge (15.8%, 95% CI 6.3–29.1% for suspected; 7.9%, 95% CI 2.0–19.3% for confirmed) compared to those who did not receive a therapeutic bridge (29.9%, CI 19.3–41.1% for suspected; 16.4%, 95% CI 8.7–26.3% for confirmed). This trend was not sustained at 12 and 24 months. Cumulative incidence analyses showed no significant difference in the overall incidence of pump thrombosis between patients who did and did not receive a parenteral bridge. In patients undergoing HM II implantation, the use of initial postoperative parenteral bridging anticoagulation is common but frequently sub-therapeutic. Use of a parenteral bridge reaching therapeutic targets may decrease the 6-month but not the overall incidence of pump thrombosis.