Thrombopoietin in preterm infants: gestational age-dependent response

PURPOSE: To investigate the factors affecting thrombopoietin (TPO) levels in preterm infants and to determine if TPO levels differ in infants born to mothers with preeclampsia and those infants with culture-proven sepsis. METHODS: Serial serum samples (N = 95) were obtained from 27 infants less than 33 weeks' gestation. Samples were analyzed for TPO using enzyme-linked immunosorbent assay. All samples had an accompanying complete blood count. Analysis of variance with post hoc analysis by least significant difference test, Mann-Whitney test, or chi2 was used to compare groups, as appropriate. Forward, stepwise linear regression was used to account for potential confounding variables. Data are expressed as mean +/- SD. RESULTS: TPO levels were not significantly correlated with the absolute platelet counts (R = -0.04, P = 0.69). TPO levels were significantly correlated with gestational age (R= 0.50, P < 0.001) when the platelet count was less than 150,000/mm3. TPO levels were significantly elevated in infants with platelets less than 150,000/mm3 born to mothers with preeclampsia compared with infants with sepsis (1184 +/- 98 vs. 579 +/- 363 pg/mL, P < 0.01). After adjusting for confounding variables using multivariate analysis (model: r2 = 0.43, P < 0.01), gestational age (r2 = 0.26) and preeclampsia (r2 = 0.03) remained significantly associated with TPO levels, whereas sepsis did not contribute to the variability of TPO. CONCLUSIONS: TPO response of infants with platelets less than 150,000/mm3 is dependent on gestational age. Infants with thrombocytopenia associated with preeclampsia have increased circulating levels of TPO. Infants with thrombocytopenia secondary to sepsis do not show an increase in TPO, but this appears to be an effect of low gestational age.