ANCS after immature fetal lung indices

A retrospective cohort study by Kamath-Rayne et al followed the different clinical treatments of women ≥34 weeks of gestation who had an amniocentesis for fetal lung maturity testing. One hundred two women whose fetal lung indices were immature were given ANCS before delivery. The ANCS-exposed infants (delivered at a mean gestation of 36 weeks) had higher rates of a composite adverse respiratory outcome, neonatal intensive care admission, hypoglycemia, and sepsis evaluation, compared with 76 infants who delivered expectantly (mean, 38 weeks of gestation) after immature fetal lung indices, and a third group of 184 infants who delivered after mature fetal lung indices (mean, 37 weeks of gestation). A second retrospective cohort study by Yinon et al evaluated 83 infants of women with immature fetal lung indices from 34–37 weeks of gestation who subsequently received betamethasone treatment vs 84 infants whose mothers did not, based on provider preference. There were no differences in RDS, transient tachypnea of the newborn infant (TTN), hypoglycemia, admission to a special care unit, or length of hospitalization between groups. The rates of respiratory support (continuous positive airway pressure or oxygen supplementation) and a composite respiratory outcome (RDS, TTN, or respiratory support) were lower in the treated group (8.4% vs 20% [ P =.03], 8.4% vs 21% [ P =.02], respectively). It is important to note that, for the study by Yinon et al, the average gestational age at delivery in the ANCS-exposed infants was later (37 vs 36 weeks of gestation) and the time between amniocentesis and delivery was slightly longer (5 days vs 4.6 days) than the Kamath-Rayne study.

A 2015 survey of 312 obstetricians and maternal-fetal medicine specialists showed that 44% of respondents use ANCS in patients >34 weeks of gestation. Many were using steroids in both late preterm and early term periods after immature fetal lung indices and waiting on the basis of the lung maturity test result and its expected rate of rise over time (unpublished data). Indeed, in a study by Shanks et al, ANCS administration after 34 weeks of gestation was associated with a higher mean weekly increase in a fluorescence polarization fetal lung maturity test than was no treatment, although the study was stopped early because of difficulty in patient recruitment and insufficient power to examine neonatal outcomes.

ANCS before elective cesarean delivery at term

Three studies have evaluated the use of ANCS before elective cesarean delivery at term. The Antenatal Steroids for Term Elective Cesarean Section (ASTECS) study, by Stutchfield et al, evaluated ANCS for 998 women who planned an elective cesarean delivery at ≥37 weeks of gestation. The primary outcome was admission to the special care unit with respiratory distress. The authors described the study as a multicenter pragmatic randomized trial; it was not blinded, and the criteria for admission to the different levels of intensive care were not clearly defined. Despite these weaknesses in study design, admission to the special care unit for respiratory distress decreased from 5.4% (24 patients) in the control group to 2.9% (11 patients) in the treatment group ( P =.02), and RDS was reduced between control and treatment groups (1.1% vs 0.2%), which was not statistically significant. Despite a decrease in admissions for respiratory distress in the ANCS-exposed group, there were similar numbers of admissions to the special care nursery in both groups (26 in treatment group; 32 in control group), although the authors note that the level of intensive care was less in the treatment group ( Table 1 ).

A nonblinded randomized trial by Ahmed et al analyzed the effect of 2 doses of antenatal dexamethasone given to a group of pregnant women at ≥37 weeks of gestation before elective cesarean delivery ( Table 1 ). Although the rates of respiratory distress overall were decreased in the treatment group (7.9% vs 23.2%; P <.01), the decrease was primarily from TTN (7% vs 19.6%). Similar to the study of Stutchfield et al, there were large absolute differences in rates of RDS (0.9% vs 3.6%; P =.40) and admission to the neonatal intensive care unit (NICU; 0.9% vs 6.3%; P =.07) that were associated with ANCS that were not statistically significant. The duration of admission to NICU was shorter in the treatment group (1.1 days compared with 3.8 days; P <.01).

Finally, Nada et al performed a randomized placebo-controlled trial of 3 doses of dexamethasone 48 hours before elective cesarean delivery at 38 to <39 weeks of gestation. In contrast to the other 2 studies, the treatment group had significantly decreased rates of admission to NICU overall. Similar to the study by Stutchfield et al, there was a significant decrease in NICU admission because of respiratory distress, likely mediated by the decrease in TTN in the treatment group compared with the control group (1.3% vs 3.4%; P =.01).

In summary, these 3 trials demonstrate a low incidence of respiratory-related adverse outcomes such as RDS after cesarean delivery at term, and the greatest effect of ANCS exposure in reduction of respiratory morbidity is related to TTN. Although not statistically significant, all 3 trials show a directionally similar reduction in RDS and NICU admission.

Antenatal corticosteroids in the late preterm period

Older studies that included small numbers of infants at >34 weeks of gestation did not show a benefit of ANCS. Crowley’s meta-analysis in 1995 included 29 cases of RDS in 886 infants at >34 weeks of gestation, with no significant improvement in the incidence of RDS. Based on these numbers, Sinclair calculated a number-needed-to-treat of 145 to prevent 1 case of RDS. The Roberts and Dalziel meta-analysis included 189 infants treated from 35 to <37 weeks of gestation with the first dose of ANCS and showed no difference in rates of RDS. Since that meta-analysis was updated in 2010, further studies that have analyzed the use of ANCS in pregnancies at >34 weeks of gestation are discussed here. In 2010, Balci et al published a prospective clinical trial of 100 women at 34–36 weeks of gestation, one-half of whom were assigned randomly to receive a single dose of betamethasone at least 24 hours before delivery. Rates of RDS with admission to NICU were reported to be 4% in the treatment group vs 16% in the control group ( P =.046; odds ratio, 0.21 [95% confidence interval, 0.04–1.08]). TTN was not discussed.

Porto et al performed a randomized triple-blinded controlled trial of ANCS in Brazil that included 320 women at 34-36 weeks of gestation who were at risk of imminent premature delivery who were assigned randomly to a course of betamethasone vs placebo ( Table 2 ). The average gestational age of both groups at delivery was the same. When the treatment and control groups were compared, there were no differences in rates of neonatal intensive care admission, respiratory morbidity, RDS, TTN, hypoglycemia, or length of NICU stay.

A nonrandomized prospective cohort study in Lebanon by Ramadan et al compared the infants of women from 34 to <37 weeks of gestation who were at risk of imminent preterm delivery whose providers treated with antenatal betamethasone with a group whose providers did not ( Table 2 ). Of note, the control group had a higher number of babies at 36 to <37 weeks of gestation, and the treatment group had higher numbers of babies at 34 to <35 weeks of gestation. Similar to the Porto et al study, Ramadan et al noted no differences in rates of RDS, TTN, respiratory morbidity, NICU admission, or NICU stay. As opposed to the study of Porto et al, however, they detected statistically increased rates of hypoglycemia and suspected sepsis in the treated group.

In February 2016, the results of the large randomized controlled double-blind Antenatal Late Preterm Steroids Trial (ALPS) were published. Women with a singleton pregnancy and without previous exposure to ANCS who were at high risk of imminent delivery in the late preterm period (34–36 weeks 6 days) were assigned randomly to betamethasone 12 mg for 2 doses that were given 24 hours apart vs placebo. High risk of imminent delivery was defined strictly as preterm labor with intact membranes and at least 3 cm dilation or 75% cervical effacement or spontaneous rupture of membranes. There was no statistical difference in rates of RDS, the need for mechanical ventilation, or NICU admission between groups ( Table 2 ), although these were notably higher than rates seen in term infants ( Table 1 ). There was a statistical difference in the primary outcome, a composite endpoint that occurred within 72 hours of birth and consisted of 1 of the following events: the use of continuous positive airway pressure or high-flow nasal cannula for at least 2 consecutive hours, supplemental oxygen with a fraction of inspired oxygen of at least 0.30 for at least 4 continuous hours, extracorporeal membrane oxygenation, or mechanical ventilation. Rates of the primary outcome were 11.6% in the treatment group compared with 14.4% in the control group (relative risk, 0.80; 95% confidence interval, 0.66–0.97; P =.02). There was also a decreased need for surfactant in the ANCS group (1.8% vs 3.1%; P =.03). An unanticipated finding was an increased incidence of hypoglycemia, defined as a glucose level <40 mg/dL, in the ANCS group compared with the control group (24.0% vs 15.0%; relative risk, 1.60; 95% confidence interval, 1.37–1.87; P <.01). Although there was no discussion of number of blood glucose measurement, the nadir of these levels, maternal blood glucoses, nor interventions that were required for hypoglycemia, the infants with hypoglycemia were discharged an average of 2 days earlier than those without hypoglycemia, which suggests that the condition was “self-limiting.” The composite outcome of severe respiratory morbidity (continuous positive airway pressure or high-flow nasal cannula for at least 12 continuous hours, supplemental oxygen with a fraction of inspired oxygen of at least 0.30 for at least 24 continuous hours, extracorporeal membrane oxygenation, or mechanical ventilation, stillbirth, or neonatal death within 72 hours after delivery) was lower in the treatment group compared with the control group (8.1% vs 12.1%; relative risk, 0.67; 95% confidence interval, 0.53–0.84; P <.01), and the number-needed-to treat to prevent 1 case of the primary outcome was 35. No long-term follow-up data were presented.

The Society of Maternal-Fetal Medicine and American Congress of Obstetricians and Gynecologists have now recommended guidelines to adopt and implement the findings of the ALPS study into clinical practice. The recommendations include an adherence to the criteria for preterm labor to decrease the potential risk for overtreatment of women who ultimately would deliver at term. They acknowledge a lack of long-term neonatal outcome data and recommend standard guidelines for the assessment of monitoring of neonatal hypoglycemia in late preterm infants.

Challenges and consequences to antenatal corticosteroid use in late preterm infants

Neonatologists have long struggled to understand the complex relationship between glucocorticoids and neurodevelopmental impairment. Glucocorticoids can have beneficial effects on the developing brain. However, the gestational age at the time of the exposure, the duration of the exposure, the pharmacodynamics of the steroid, and genetic variation in glucocorticoid receptors in the brain all may modulate how the effects are manifested in the short and long term. There are steroid treatments with betamethasone and dexamethasone that are effective for the preterm infant, but the minimal doses have not been evaluated adequately.

Although a single course of ANCS appears to improve neurodevelopmental outcomes in infants born at <34 weeks of gestation, little is known about the long-term effects of exposure to ANCS after 34 weeks of gestation. This time period is critical for brain development because, at 34 weeks, the brain weight is only 65% of the term brain, and gyral and sulcal formation are still incomplete. From 34–40 weeks of gestation, cortical volume increases by 50% and 25% of the cerebellar development occurs. Stutchfield et al performed a follow-up assessment to the Antenatal Steroids for Term Elective Cesarean Section trial of betamethasone before an elective cesarean delivery at term; 93% of the original cohort could be contacted, but only 51% of them completed the questionnaires that served as the basis for the assessment. Although there were no differences in behavior and health for the subjects at 8–15 years of age, a higher incidence of being in the lowest quarter of academic ability was noted in the betamethasone group (17.7% vs 8.5%).

Several studies have examined the metabolic and hormonal effects of ANCS that were administered in the late preterm period. Hypoglycemia was noted in late preterm infants whose mothers were treated with ANCS in several studies. In a retrospective cohort study of 6675 preterm deliveries at 32–37 weeks of gestation at a single university hospital, the odds ratio of hypoglycemia in corticosteroid-exposed infants, after adjustment of gestational age, was 1.60 (95% confidence interval, 1.24–2.07). The hypoglycemia was increased in infants whose mothers had pregestational diabetes mellitus (adjusted odds ratio, 5.65; 95% confidence interval, 3.84–8.33). However, there were no statistically significant differences between groups when stratified by gestational age. A limitation of the study was the lack of maternal blood glucose levels. Sifianou et al analyzed cord blood in a convenience sample of 32 singleton newborn infants at >35 weeks of gestation whose mothers received a single 12-mg dose of betamethasone approximately 24 hours before planned cesarean delivery. This group was compared with 44 babies of comparable gestational age, sex, and nutritional status who were not exposed to ANCS. None of the mothers had pregestational or gestational diabetes mellitus. Cord blood levels of C-peptide and glucose were higher in ANCS-exposed fetuses, which indicates that these fetuses were hyperinsulinemic and thus at higher risk for neonatal hypoglycemia.

Finally, ANCS therapy in preterm infants may lead to cardiovascular and metabolic health problems in later life, although the reports, which were derived mostly from observational studies, are conflicting. For example, a cohort of 210 preterm survivors who were observed until age 14 years had higher systolic and diastolic blood pressures in adolescence, which could lead to clinical hypertension later in life. Potential mechanisms that were responsible for this observation have been identified in animal studies. A different cohort of 209 term born children who were exposed to ANCS in approximately week 30 of gestation had significantly increased cortisol reactivity to psychosocial stress when evaluated at 6–11 years of age, which suggests longer term effects of ANCS on the hypothalamus-pituitary-adrenal axis than previously thought. However, it is important to note that there are no data on long-term cardiovascular or metabolic data on late preterm or term infants.

Challenges for administration of ANCS

There are clear benefits to the administration of ANCS in preterm gestation when delivery occurs 24 hours to 7 days after treatment. However, ensuring that ANCS are given at the correct time, without adequate methods to predict the timing of preterm birth, is an ongoing challenge. Many perinatal collaboratives use treatment with ANCS as a marker of quality, and rates of ANCS have increased to approximately 90% for deliveries at 24–34 weeks of gestation. However, of concern, there has been a subsequent increase in suboptimal treatment <24 hours or >7 days before delivery) and questionably appropriate treatment (≥35 weeks exposed to ANCS). In all live births in Nova Scotia from 1998–2012, the rates of suboptimal treatment were 34% (odds ratio, 6.7; 95% confidence interval, 3.9–11.6) and questionably appropriate treatment of 1.7% (odds ratio, 7.5; 95% confidence interval, 4.9–11.3). In fact, in 2012, more than one-half the newborn infants whose mothers received ANCS were born at 35 weeks of gestation and thus were exposed unnecessarily. Another study demonstrated that, of 692 women who received ANCS at a single institution, 35.7% delivered at ≥34 weeks of gestation, and 17.9% remained pregnant beyond a week after ANCS and were still <34 weeks of gestation. A rescue dose of steroids is often used in these situations; however, multiple courses of ANCS are not recommended because of poor fetal head growth and increased risk of neurodevelopmental impairment by 5 years. To fully implement ANCS to the greatest benefit of the preterm pregnancies that are at risk of imminent delivery, improved methods of preterm birth prediction are needed. Furthermore, if ANCS treatment is extended from 34 to <37 weeks of gestation and many of these fetuses deliver at term, the effects are unknown.