Abstract 12339: Erythropoietin Use in Pediatric Ventricular Assist Device Patients

Abstract

Background: Anemia is common in pediatric ventricular assist device (VAD) patients. Erythropoietin (EPO) has been used to reduce blood transfusions in children after VAD placement. A study in adults showed EPO use after left VAD implantation was associated with higher rates of suspected pump thrombosis. No studies have evaluated the use of EPO in pediatric patients receiving VAD support. Objectives: We aim to assess EPO use in all pediatric patients who underwent placement of a VAD and to assess outcomes for children who received EPO compared to those who did not. Methods: We identified patients from the Pediatric Health Information System who underwent VAD placement from 2010-2020. Use of EPO was described across all sites. Cohorts were compared to assess differences in blood transfusions received post-VAD, incidence of stroke and pump-exchange. Multivariable Cox proportional hazard and logistic regression models were constructed to adjust for confounding factors for outcomes of VAD exchange and stroke, respectively. Results: A total of 1260 patients were included, of which 336 (27%) received EPO. Patients who received EPO were younger (median 3.5y vs. 9y, p<0.001) and more likely to have congenital heart disease (55% vs. 45%, p=0.002). Patients on EPO received fewer blood transfusions compared to those not on EPO [median 0/month (IQR 0-0.9) vs. 0/month (IQR 0-2.3), p<0.001]. VAD exchange and stroke were more common in patients on EPO (10.1% vs. 5.5%, p=0.004 and 37.8% vs. 25%, p<0.001, respectively). After adjusting for confounding factors, the use of EPO was not independently associated with the need for VAD exchange (adjusted hazard ratio 1.2, 95% confidence 0.78 - 1.9, p=0.378). However, EPO remained independently associated with stroke during the hospitalization (adjusted hazard ratio 1.6, 95% confidence 1.2 - 2.1, p=0.002). Conclusion: EPO was used in approximately 1 out of 4 pediatric VAD patients. Those who received EPO were younger and more likely to have congenital heart disease. The use of EPO is associated with fewer post-VAD blood transfusions. EPO was not independently associated with VAD exchange but remained independently associated with stroke during the VAD hospitalization.