Does Neonatal Phenobarbital Treatment Affect Testicular and Adrenal Functions and Steroid Binding in Plasma in Infancy?*

Because of its enzyme-inducing properties, which are thought to explain its lowering effect on neonatal hyperbilirubinemia, phenobarbital (PB) has been widely used as prophylactic treatment in neonatal jaundice. Experimental data suggesting that PB might have various effects on endocrine metabolism or regulation prompted us to investigate whether PB had some adverse effects on testicular or adrenal function in human neonates. A group (I) of 12 prematures given a single im injection of 10 mg/kg PB on the day of birth were studied longitudinally for 6 months and compared with untreated, premature (group II), and fullterm (group III) infants. Plasma testosterone (T), Δ4-androstenedione (Δ4), 17α-hydroxyprogesterone, and cortisol (F) were similar in the 3 groups at birth and before PB therapy. A postnatal rise in T and 17α-hydroxyprogesterone levels occurred in the 3 groups. Study of the temporal patterns showed that peak T levels were lower, and their decline beyond 2 months of age was more rapid, in group I than in group II. However, peak T levels were delayed in both groups compared with group III. Temporal patterns and mean levels of Δ4 and F, their binding to plasma proteins, and their unbound plasma concentrations were comparable in the 3 groups. In contrast, the postnatal rise in T binding was markedly delayed in group I, whereas it was identical in groups II and III. Unbound T levels were greatly elevated in group I in comparison to control groups. It is suggested that prophylactic PB treatment does not affect testicular or adrenal secretions in infancy. PB did not displace T, Δ4, or F from their binding sites in plasma in vitro. It is thus possible that PB selectively inhibits the synthesis of T-estradiolbinding globulin, but a competition of the metabolites of the drug on T-estradiol-binding globulin-binding sites cannot be excluded. Increased unbound T levels may have, as a result, an enhanced MCR of T, thus explaining the temporal patterns observed. The long term effect, if any, of the exposure to high levels of unbound T for the first 2 months of life in these infants treated by PB is unknown.