Bleeding Complications between Bivalirudin and Heparin for Extracorporeal Membrane Oxygenation in Neonates with Congenital Diaphragmatic Hernia.

Introduction: Neonates with congenital diaphragmatic hernia (CDH) who undergo repair while on extracorporeal membrane oxygenation (ECMO) are at risk of developing post-operative bleeding complications. Balanced anticoagulation is critical to maintain ECMO flow and avoid bleeding. Heparin has historically been our first-line anticoagulant; however, recently, we transitioned to bivalirudin, a direct thrombin inhibitor. The objective of this pilot study was to compare post-operative surgical bleeding complications between the two groups.

Methods: We performed a single center retrospective cohort study of patients who underwent CDH repair while on ECMO between 2008 and 2023. Neonates were stratified based on the type of anticoagulant initiated after CDH repair. Outcomes included bleeding requiring surgical re-operation, intracranial hemorrhage, volume of blood products transfused, number of circuit changes, days on ECMO, and overall survival.

Results: Among 62 neonates with CDH who underwent repair on ECMO, 44 (71%) were managed post-CDH repair with heparin and 18 (29%) with bivalirudin. One (5.6%) neonate managed with bivalirudin underwent re-operation following CDH repair for a bleeding complication compared to 17 (38.6%) managed with heparin (p = 0.022). In addition, the bivalirudin cohort utilized half of the total blood product volume compared to the heparin cohort (p = 0.020). Despite these benefits, there were no significant differences between groups for incidence of intracranial hemorrhage, number of circuit changes, days on ECMO, and overall survival.

Conclusion: Anticoagulation with bivalirudin in neonates who underwent CDH repair while on ECMO was associated with decreased surgical bleeding complications and less total blood product transfused. This pilot analysis is the first to compare heparin to bivalirudin and stresses the importance of a multicenter study.