Erythropoietin, transfusions, and outcomes of retinopathy of prematurity and brain injury in extremely preterm infants: A post hoc analysis of the Preterm Erythropoietin Neuroprotection Trial (PENUT)

Background Erythropoietin is perceived as both a neuroprotectant and a biomarker for hypoxic stress. Objective To explore correlations between serum erythropoietin (Epo) concentrations, perinatal risk factors, red blood cell transfusions and recombinant human erythropoietin (rHuEpo) with outcomes including retinopathy of prematurity (ROP) and brain injury on magnetic resonance imaging (MRI) in extremely preterm infants. Methods This is a post hoc analysis of data from the Preterm Erythropoietin Neuroprotection Trial of preterm infants born between 24 0/7 and 27 6/7 weeks gestation and randomized to placebo or rHuEpo treatment (N = 941). Serum Epo concentrations were collected within 24 hours (baseline) and at 7, 9, and 14 days. MRI was obtained at 36 weeks postmenstrual age (N = 220). Results Baseline Epo concentrations negatively correlated with gestational age, delayed cord clamping, and Apgar scores, and positively correlated with intraventricular hemorrhage and risk of death. Neither endogenous Epo at baseline nor trajectories from birth to 14 days were associated with ROP. In the placebo group, Epo at 1 week of life (r = 0.26, p = 0.033) and 2-week area under the curve (r = 0.28, p = 0.019) positively correlated with white matter injury. In the treatment group, Epo at 14 days negatively correlated with white matter injury (r = -0.35, p = 0.004). Grey matter injury negatively correlated with baseline Epo in the placebo group (r = -0.27, p = 0.01) but positively correlated in the treatment group (r = 0.23, p = 0.047). Transfusions were associated with severe ROP (p < 0.0001) and total brain injury on MRI (p = 0.007). Transfusion volumes in the first week of life were associated with a greater risk of severe ROP in males (p = 0.0006). Conclusions Endogenous Epo concentrations in preterm infants are influenced by perinatal variables and correlate with poor outcomes. The association of Epo with MRI results differed between placebo and rHuEpo treatment groups. Transfusions were associated with increased ROP and brain injury on MRI.