Antenatal glucocorticoid treatment: are we doing harm to term babies?

Abstract

Antenatal glucocorticoids are given to 7–10% of women in Europe and North America at threat of preterm labor. It is undisputed that this therapy significantly reduces neonatal morbidity and mortality in infants born before 34 wk gestation (1, 2). Based on outcomes from multiple clinical trials, current obstetric practice is to administer either a single 2-d course of betamethasone or dexamethasone (i.e. two 12-mg doses of betamethasone im at 24-h intervals or four 6-mg doses of dexamethasone im at 12-h intervals) to women in preterm labor between 24 and 34 wk gestation (3). The use of repeat courses of antenatal glucocorticoids is more controversial, although some evidence suggests that a repeat course, if the mother does not deliver but remains at risk of preterm labor, may yield neonatal benefit (2). However, because early clinical diagnosis of preterm labor is difficult and there are no reliable laboratory tests, there are a substantial number of women who receive antenatal glucocorticoid treatment yet subsequently deliver babies at full term, i.e. after 37 wk gestation. Little is known about the potential risks or benefits of synthetic glucocorticoid exposure to these infants. It is now well recognized that exposure to an adverse in utero environment during development has long-term effects on physiology and later risk of adult disease (4). Overexposure of the developing fetus to excess glucocorticoids is a key mechanism thought to underlie such “developmental programming” (4). Numerous preclinical studies have shown that administration of dexamethasone to pregnant mothers leads to long-term changes in the offspring, notably permanently altering cardiometabolic and neuropsychiatric functions as well as hypothalamicpituitary-adrenal (HPA) axis activity (4). The latter may contribute to some of the broader pathogenesis associated with experimental programming. Both betamethasone and dexamethasone have high affinity for glucocorticoid receptors but are poor substrates for the placental enzyme 11hydroxysteroid dehydrogenase type 2, which normally acts as a partial barrier, protecting the fetus from glucocorticoids by breaking down endogenous cortisol into inactive cortisone (4). These synthetic glucocorticoids readily cross the placenta, potentially exposing the fetus to excess glucocorticoids. In animal models, the developing HPA axis appears particularly vulnerable to excess glucocorticoids (4), yet the evidence for detrimental effects of antenatal glucocorticoids on human HPA axis activity has been less clear-cut. This is partly because most studies have included preterm infants, who often have coexisting medical complications and are exposed to stressors and painful procedures while in the neonatal intensive care unit, factors that can per se alter HPA axis function (5). Likewise, the type, dose, and timing of administration of glucocorticoid therapy often differ between studies, complicating interpretation. Furthermore, studies have included limited assessments of the offspring HPA axis, often with a single measurement of cortisol levels. Given these caveats, the limited existing data, which include preterm babies in whom the HPA axis may be immature, suggest that among infants exposed to antenatal glucocorticoids, baseline cortisol levels are suppressed during the first postnatal week, and cortisol responses to the painful stress of the heel-stick procedure are suppressed for up to 8 wk (6), suggesting some persistent suppressive effects on HPA axis regulation, but hardly the programming of increased (or reduced) HPA activity seen throughout the lifespan in animal models.