Abstract
The ability of recombinant human erythropoietin (rhEPO) to protect preterm infants against perinatal intrauterine herpes virus infection-induced brain injury was studied. In total, 120 women infected with perinatal intrauterine herpes virus were randomized into four groups: A, B, C and D, and were given 1,500 IU (mother, pre-partum), 3,000 IU (mother, pre-partum), 250 IU/kg (infant, post-natal), and no rhEPO, respectively. Hemoglobin (Hb), reticulocyte (Ret), hematocrit (Hct), neuron specific enolase (NSE), myelin basic protein (MBP), and S100 protein B (S100B) levels were measured immediately (T(0)) and at 1 week (T(1)), 2 weeks (T(2)), and 4 weeks (T(3)) post-delivery. Linear regression analysis was performed to analyze inter-indicator correlation, and ROC risk models were established to determine the predictive value of Hb, Ret and Hct for brain injury immediately after delivery. The brain injury incidence rate of group A (10%) was significantly lower than group D (33.3%) and group B (6.7%) significantly lower than groups C (26.7%) and D. At T(0), Hb, Ret and Hct in groups A and B were significantly higher than in group C and D, while from T(1) to T(3), groups A, B and C showed significantly higher values than group D. NSE, MBP and S100B showed an inverse trend, with groups A and B lower at T(0) and groups A, B and C lower from T(1)-T(3). Hb and NSE, MBP and S100B were negatively correlated, while no correlation was found between Ret and NSE, MBP and S100B. Finally, Hct and NSE, MBP and S100B were negatively correlated. The optimal cut-off values for Hb and Hct for brain injury diagnosis immediately post-partum were 170 g/l (sensitivity 99%, specificity 95.7%) and 28.5% (sensitivity 79.4%, specificity 100%), respectively. Ret did not show predictive value. In conclusion, pre-partum rhEPO treatment showed greater protective effects than post-natal administration, and this may be the regulation of Hb and Hct levels in post-natal preterm infants. In addition, a dose-dependent effect was displayed.