Erythropoietin Concentrations and Neurodevelopmental Outcome in Preterm

OBJECTIVE.Erythropoietin therapy is effective in decreasing transfusions to varying degrees in preterm infants. Recent animal studies using erythropoietin doses to achieve serum concentrations 1000 mU/mL report neuroprotective effects. We evaluated the relationship between erythropoietin concentrations and neurodevelopmental outcome in extremely low birth weight infants. METHODS. Preterm infants who weighed 1000 g at birth were randomly assigned to erythropoietin (400 U/kg 3 times per week) or placebo/control. Therapy was initiated by 4 days after birth and continued through the 35th postmenstrual week. All infants received supplemental parenteral and enteral iron. Peak serum erythropoietin concentrations were obtained every 2 weeks. Follow-up evaluation included anthropometric measurements, Bayley scales of mental and psychomotor development, neurologic examination, and determination of overall neurodevelopmental impairment. Data were collected at 18 to 22 months’ corrected age by certified examiners who were masked to the treatment group. Analyses were performed to identify correlations between erythropoietin concentrations and outcomes. RESULTS. Sixteen extremely low birth weight infants were enrolled; 1 infant died at 2 weeks (placebo/control), and 15 had erythropoietin concentrations measured (7 erythropoietin, 8 placebo/control). Peak erythropoietin concentrations were significantly different between groups during the study (erythropoietin: 2027 1464 mU/mL; placebo/control: 26 11 mU/mL). Before follow-up, 3 infants died (1 erythropoietin, 2 placebo/control), and 12 were available for follow-up (6 erythropoietin, 6 placebo/control). At 18 to 22 months’ follow-up, none of the erythropoietin recipients and 2 of the placebo/control infants had Mental Development Index scores 70. Erythropoietin recipients had Mental Development Index scores of 96 11, and placebo/control infants had Mental Development Index scores of 78 7. Psychomotor Development Index scores were similar between groups (87 13 vs 80 7). There were no differences between groups with respect to anthropometric measurements. Two of 6 infants in the erythropoietin group and 4 of 6 infants in the placebo/control group had some form of neurodevelopmental www.pediatrics.org/cgi/doi/10.1542/ peds.2005-3186 doi:10.1542/peds.2005-3186