Use of bivalirudin for anticoagulation in pediatric extracorporeal membrane oxygenation (ECMO)

This study describes the use of bivalirudin in children on extracorporeal membrane oxygenation (ECMO). Pediatric patients receiving bivalirudin were compared to patients receiving heparin as the anticoagulant on ECMO. Data was collected for children under 18 years of age supported by ECMO from January 2016 to December 2019. Data collected included demographics, diagnosis, ECMO indication, type, and duration, indication for bivalirudin use, dose range, activated partial thromboplastin time (aPTT) levels, minor and major bleeding, hemolysis, and mortality. Forty pediatric patients received ECMO; eight received bivalirudin primarily for anticoagulation. The median age was 4 months (IQR 0.5, 92) in the heparin cohort, 0.6 months (IQR 0.0, 80.0) in the primary bivalirudin cohort. The indication for ECMO was respiratory in 5 patients (18%) in the heparin group versus 6 (75%) in the primary bivalirudin group, cardiac in 18 (67%) in heparin versus 1 (12.5%) in primary bivalirudin, and extracorporeal-cardiopulmonary resuscitation (E-CPR) in 4 (15%) in heparin versus 1 (12.5%) in primary bivalirudin. Bivalirudin was the initial anticoagulant for eight patients (66.6%) while three (25%) were switched due to concern for heparin-induced thrombocytopenia (HIT) and one (8%) for heparin resistance. The median time to achieve therapeutic aPTT was 14.5 hours compared to 12 hours in the heparin group. Sixty-five percent of aPTT values in the bivalirudin and 44% of values in the heparin group were in the therapeutic range in the first 7 days. Patients with primary bivalirudin use had significantly lower dose requirement at 12 (p = 0.003), 36 (p = 0.007), and 48 (p = 0.0002) hours compared to patients with secondary use of bivalirudin. One patient (12.5%) had major bleeding, and two patients (25%) required circuit change in the primary bivalirudin cohort. Bivalirudin may provide stable and successful anticoagulation in children. Further large, multicenter studies are needed to confirm these findings.