Use of Antenatal and Postnatal Steroids1

There is conclusive evidence that antenatal corticosteroid treatment reduces the risk of respiratory distress syndrome (RDS), mortality, and intraventricular hemorrhage (IVH) in preterm infants.⁹ Acceleration of lung development by corticosteroids results in increased tissue and alveolar surfactant and structural changes that enhance the response to postnatal administration of surfactant.^4,6

In 1994, the National Institutes of Health (NIH) Consensus Development Conference recommended that a course of antenatal corticosteroids be given to pregnant women who were at risk for preterm delivery within the next seven days.²⁴ This was based on the observation that the reduction in RDS was greatest in babies delivered between 24 hours and seven days of treatment.⁸ This recommendation led to the practice of giving multiple courses of treatment if the pregnancy persisted longer than one week, although there was no supporting evidence for this approach from human trials.

Evidence from animal studies and limited human data suggests that excessive antenatal corticosteroid exposure may have adverse effects on fetal growth and neurodevelopment.^1,10,22,28 In a systematic review of 19 randomized controlled trials performed in animals, repeated courses of antenatal corticosteroids improved lung function in the eight trials that reported that outcome¹ (Table 1.1). However, in nine of eleven studies that evaluated fetal growth, growth restriction resulted from repeated courses. In addition, multiple courses were associated with adverse effects on nervous system or brain function or growth in all seven studies that examined these outcomes.

Among observational studies in humans comparing single and repeated courses of corticosteroids, the majority showed no difference in outcomes of RDS, mortality, and IVH.²⁰ Only one randomized controlled trial in humans comparing multiple to single courses of corticosteroids has been reported to date. Among women who were at high risk of preterm delivery, no difference between a single course and weekly courses of corticosteroids until 34 weeks gestation or delivery was found in the primary outcome of composite neonatal morbidity, consisting of RDS, BPD, severe IVH, periventricular leukomalacia,2 proven sepsis, necrotizing enterocolitis, or death (22.5 versus 28%; RR 0.80, 95% CI 0.59 to 1.10).¹² This trial was terminated before completion of the planned enrollment because of marginal differences between the groups in the interim analysis and concerns about the safety of multiple exposures. In 2000, the second NIH consensus conference on antenatal corticosteroids concluded that there were insufficient data from clinical trials regarding efficacy and safety and recommended against the routine use of repeated courses.³

The use of antenatal corticosteroids has not reduced the risk of BPD, also known as chronic lung disease (CLD), which occurs in 23 to 26 percent of very low birth weight infants.^18,19 A major factor in the development of BPD is inflammation that may result from increased oxygen concentration, mechanical injury, or infection.¹⁶ Low serum cortisol concentrations, poor responses to corticotropin, and increased levels of cortisol precursors have been associated with an increased risk of BPD, suggesting that infants who develop BPD have relative adrenal insufficiency that may result in an exaggerated inflammatory response to lung injury.^15,17,26 Because of their anti-inflammatory effect, as well as the suggestion of relative adrenal insufficiency in affected infants, corticosteroids, predominantly dexamethasone, have been widely used to treat or prevent BPD in preterm infants.

Multiple trials using a variety of doses and durations of therapy have shown a beneficial pulmonary effect from postnatal corticosteroids, usually dexamethasone, given soon after birth. Systematic reviews and meta-analyses of 21 trials of early treatment (less than 96 hours after birth) and 7 trials of moderately early treatment (at 7 to14 days of age) showed that corticosteroids significantly reduced the risk of CLD and of CLD or death at 28 postnatal days and 36 postmenstrual weeks, facilitated earlier extubation, and reduced the later use of corticosteroids.^13,14 However, no effect on mortality was seen.

Serious short-term and long-term complications appear to outweigh the beneficial pulmonary effects of postnatal corticosteroid treatment in most circumstances (Table 1.2). Acute adverse effects include hyperglycemia, systemic hypertension, and gastrointestinal bleeding and perforation. Intermediate complications include infection, cardiomyopathy, adrenal suppression, and growth failure. Late effects include neurodevelopmental abnormalities, including cerebral palsy.

Adverse events are frequently associated with early postnatal dexaethaasone use even when relatively low doses are given. 3This was illustrated by the trial conducted by the National Institute of Child Health and Human Development Neonatal Research Network.²⁵ Infants with a birth weight of 501 to 1000 gms who were mechanically ventilated before 12 hr of age were randomly assigned to treatment with dexamethasone or placebo starting on the first postnatal day, and then tapered over a 10-day period. The initial dexamethasone dose was 0.15mg/kg per day, which is equivalent to approximately five times the cortisol replacement dose, and less than one-third of the dose used in most other trials or in clinical practice. A total of 220 infants were enrolled before the trial was stopped because of an unanticipated and unacceptable rate of spontaneous gastrointestinal perforations in the dexamethasone group.

The rate of CLD or death at 36 weeks postmenstrual age, the primary outcome, was not significantly different in the dexamethasone or placebo groups (63 versus 69%, relative risk 0.92, 95% confidence intervals 0.76 to 1.11).²⁵ Hypertension, defined as a systolic blood pressure greater than 80 mmHg, occurred significantly more often in the dexamethasone group (27 versus 4%) than in the controls. Hypoglycemia, indicated by the need for insulin treatment, was also more frequent in the dexamethasone-treated infants (23 versus 12%). These adverse effects, observed even at the lower doses used in this study, may result from relatively high serum levels of dexamethasone due to a prolonged half-life of the drug in extremely immature infants.^7,21

During the first 14 days after birth, spontaneous intestinal perforation, the adverse event that led to termination of the trial, occurred in more infants who received dexamethasone than placebo (13 versus 4%).²⁵ Perforation was associated with indomethacin treatment in the first 24 hours after birth. A possible mechanism is the inhibition by corticosteroids and indomethacin at two points in the synthesis of prostaglandins, thus interfering with their role in maintaining gastrointestinal integrity.^11,29

Similar to most other studies, early dexamethasone administration reduced the rate of later clinical open-label use of dexamethasone.²⁵4However, the high rate of exposure to open-label dexamethasone in both groups (34 and 51% in dexamethasone and placebo groups, respectively) makes the effects of early treatment on long-term outcomes difficult to interpret.

Dexamethasone also adversely affected growth.²⁵ At 36 postmenstrual weeks, weight and head circumference were less in the dexamethasone group compared to placebo, even though more infants in the placebo group received open-label dexamethasone.²⁵ These extremely immature infants may be especially susceptible to the catabolic effects of dexamethasone given soon after birth.

There is increasing concern that postnatal treatment with corticosteroids has adverse effects on brain growth and neurodevelopment.^5,23,28 In an observational study, cerebral cortical gray matter volume assessed by quantitative magnetic resonance imaging was substantially less in premature infants who were treated with dexamethasone, compared to those who never received the drug.²³ Although infants who were treated tended to be more immature and ill, the impairment of brain growth is consistent with experimental studies.

Long-term follow-up data after early dexamethasone treatment are limited, although they suggest an increased risk of developmental delay, cerebral palsy, and abnormal neurological examination.^5,13 Although the few available studies are heterogeneous, lack complete follow-up, and often have a substantial rate of open-label steroid use, the risk of damage to the central nervous system by dexamethasone exposure is biologically plausible. Perinatal administration of corticosteroids has long been recognized to have adverse effects on brain development in animals. These effects included immediate impairment of brain cell division, differentiation, myelination, and electrophysiological reactions and later disturbances of circadian rhythm and behavior.²⁷

In summary, more research is needed on the effects of perinatal corticosteroid treatment, especially long-term neurodevelopment. Currently, antenatal administration of a single course of corticosteroids is recommended when preterm delivery is anticipated. Repeated courses should not be used routinely.³ Because of concerns about adverse effects, the Committees on Fetus and Newborn of the American Academy of Pediatrics and the Canadian Paediatric Society recommend against the routine use of postnatal dexamethasone to prevent or treat BPD in preterm infants.²