Antenatal Phenobarbital and Neonatal Intracranial Hemorrhage

Source: Shankaran S, et al. The effect of antenatal phenobarbital therapy on neonatal intraventricular hemorrhage in preterm infants. N Engl J Med 1997;337:466-471.

Intracranial hemorrhage (ich) constitutes a sig-nificant cause of neurological morbidity in infants born at 33 weeks or less. About 20-30% of infants born at or below 33 weeks sustain this complication. Neonatal ICH has classically been "graded" according to a four-level scale:

Grade I—Germinal matrix hemorrhage only

Grade II—Intraventricular hemorrhage without hydrocephalus

Grade III—Intraventricular extension with acute hydrocephalus

Grade IV—Intraparenchymal extension

Only 10-15% of infants (or 2-4% of all < 34-week premature infants) with ICH sustain Grade III or IV bleeds, which are associated with substantially poorer outcomes than those of lower grade. Nevertheless, considering the lifetime consequences of higher grade ICH, any reasonable strategy to reduce its incidence would likely spare a number of individuals from handicapping neurological disabilities.

The pathophysiology of ICH is complex. Some cases of ICH may be attributable to the hyperperfusion that follows a period of ischemia in the immature brain during which cerebral flow autoregulation is not fully developed. Thus, factors that predispose to ischemia, such as placental insufficiency and sepsis, increase the risk of ICH. Some recent studies have examined the effects of various agents administered just before or just after birth on the incidence of ICH; benefit has been suggested for both indomethacin and corticosteroids.

Several recent studies have also examined phenobarbital. The sedative effect of phenobarbital might reduce blood pressure and cerebral blood flow fluctuations in the fetus if administered antenatally. At least four other studies (each of which included < 150 subjects) had suggested a potential benefit of perinatal phenobarbital (see, for example, Kaempf, et al. J Pediatr 1990;117:933-938, and references cited therein), and confirmed by meta-analysis. This motivated Shankaran et al to conduct a large multi-center, randomized, double-blinded, placebo-controlled trial on the effect of antenatal phenobarbital on neonatal ICH.

Shankaran et al examined 610 women (bearing 668 infants) who were 24-33 weeks pregnant and for whom there was an obstetrical indication for delivery within the 24 hours following recruitment into the study. Women then received a single dose of phenobarbital (10 mg/kg) intravenously or saline and additional oral maintenance of phenobarbital (or placebo) if they had not delivered within 24 hours of entry into the study. Phenobarbital was discontinued in those women who had not delivered by 33 weeks gestation.

Shankaran et al found an incidence of ICH of 23% in the phenobarbital-treated group (n = 344 infants) and 24% in the saline/placebo group (n = 324).


Although mothers treated with antenatal phenobarbital were more sedated and their babies had lower one-minute APGAR scores, this study found no significant differences in the two study groups with respect to Bayley II Mental Development or Psychomotor scores at 18-22 months, or the incidence of cerebral palsy. There is no evidence that antenatal phenobarbital, as administered here, is harmful in the short or long term.

While antenatal phenobarbital may be safe, the large study of Shankaran et al suggests that, unlike several previous studies, antenatal phenobarbital provides no significant benefit in reducing the incidence of ICH. This therapy cannot be recommended based on the best available data. —rt