Does the intrauterine growth-restricted fetus benefit from antenatal glucocorticoids?

In Australia, 8% of the 280,000 annual births are preterm; specifically, prior to 37 weeks completed gestation [1]. Preterm infants represent 75% of all neonatal deaths in Australia, with the vast majority of these deaths due to pulmonary disease [2]. The costs of caring for preterm infants are enormous – US$5.8 billion in the USA, representing 57% of neonatal care in that country [3]. The cost to support a single infant born at 25 weeks is estimated to be at US$250,000 [4]. Since Liggins’ proposal that glucocorticoids (GCs), including cortisol, β-methasone and dexamethasone, specifically induce pulmonary surfactant synthesis [5], it has been shown that synthetic GCs reduce the incidence of respiratory distress syndrome by approximately 50% in preterm infants [6]. Interestingly, not all babies may benefit from antenatal GC; for example, the growth-restricted fetus. Although the size of this subset of preterm babies is not clear, 7% of Australian babies are born intrauterine growth restricted (IUGR) [1]. This editorial discusses the evidence for and against the use of antenatal GC, particularly in relation to the complex cardiovascular and respiratory physiology of the IUGR infant.

Benefits of antenatal GC treatment

The use of a single course of antenatal GCs, either dexamethasone or β-methasone, is recommended for women at risk of preterm labor between 24 and 34 weeks, and is associated with positive outcomes for the infant in that they reduce the incidence of neonatal complications associated with preterm birth; for example, neonatal respiratory distress [6], intraventricular hemorrhage [7] and neonatal death [7–9]. Experimental studies show an improvement in fetal lung mechanics after very short treatment-to-delivery times [10], with concomitant alterations in lung structure [11]. Antenatal GC is most effective when administered between 7 days and 24 h before delivery [8]. As 50% of women in threatened preterm labor do not deliver within 7–14 days of initial antenatal GC [12], it has, in previous years, been common practice to repeat the course of antenatal GC at weekly intervals until the time of delivery [13–15]. The latest Cochrane review on antenatal GCs in women at risk of preterm labor concludes that a single course of antenatal GC is recommended to accelerate fetal lung maturation [16,17]. However, further research is required concerning the optimal dose-to-delivery interval, optimal corticosteroid to use and the effects of multiple pregnancies, as well as to confirm the long-term effects into adulthood, especially of multiple doses [16]. A recent, randomized controlled trial shows that multiple doses every 14 days do not improve survival in preterm babies but do decrease fetal growth [18].

Antenatal GCs alter lung structure, as they result in a thinning of the alveolar walls, a higher proportion of alveolar ducts and a lower alveolar wall fraction in preterm sheep [19]. These structural indices result in a higher average alveolar volume and improved lung function within 2 days after GCs [19]. Improved lung function is measured by improved lung compliance [20,21], greater pulmonary surfactant stability (i.e., sustained ability to lower surface tension) and the ability of the surfactant to improve lung compliance in surfactant-deficient animals [20]. These effects occurred after a single treatment 48 h before delivery [20,21]. On the other hand, increased surfactant lipid content and saturation, as well as an increased expression of surfactant proteins (SPs) in alveolar (i.e., lavageable) surfactant, is only manifest after repeated doses over a prolonged period of time (7–21 days) [19,21]. In sheep, repetitive doses at 7-day intervals improve fetal lung function cumulatively when delivered preterm [22].

Costs of antenatal GC treatment

Despite the benefits of antenatal GC treatment for neonatal lung function, antenatal GCs are also associated with several negative birth outcomes, such as adrenal suppression, neonatal sepsis, and increased risk of neonatal and maternal infections [8,23]. Of greatest concern is that antenatal GCs reduce fetal growth, resulting in both lower body weight, but also in a smaller and less mature brain [12,23]. In fact, these children remain smaller at school age [24]. In addition, animal studies show that antenatal GCs reduce cerebral blood flow due to increased cerebral vascular resistance with a decrease in oxygen delivery [25]. Moreover, there are deleterious effects on cerebral myelination [26] and significant alterations in the function of the hypothalamic–pituitary–adrenal axis [27,28]. These changes may affect later neurodevelopment [29]. There is an increase in umbilical blood flow in response to β-methasone [30]. In pregnant ewes infused for 48 h with β-methasone, fetal cerebral blood flow and oxygen delivery are reduced due to increased cerebrovascular resistance. This may account for the decreased brain growth observed at term in human fetuses following either single or repeated GCs [31]. GCs act to increase vascular resistance through a variety of mechanisms, including enhanced sensitivity to vasoconstrictors and decreased sensitivity to vasodilators [32].

Effects of antenatal GC treatment in the IUGR fetus

There are conflicting data on whether antenatal GCs are [9] or are not [7] associated with a reduction in the complications associated with preterm delivery in IUGR fetuses. A large study of 19,759 very low-birthweight neonates found that antenatal GCs lowered the risk of respiratory distress, intraventricular hemorrhage and death in both normal and growth-restricted fetuses [9]. On the other hand, a study of 1148 neonates found that there was no difference in the incidence of respiratory distress, intraventricular hemorrhage or necrotizing enterocolitis in growth-restricted fetuses whether they were treated with antenatal GCs or not [7]. The already higher circulating plasma cortisol concentrations in the IUGR fetus may have already affected lung growth and surfactant maturation such that exposure to higher doses may have no or a detrimental effect. In addition, the effects of GCs are widespread, not limited to the lung, and as such, actions of GCs at other sites may be deleterious for the IUGR fetus. The complexity of the interactions between the factors that regulate SP synthesis is exemplified in the case of the IUGR fetus.

Impact of IUGR, hypoxemia & cortisol on the surfactant system

Fetal sheep models have been utilized to investigate the effects of IUGR on the lung and surfactant system and antenatal GCs, either separately or additively, with conflicting but interesting mechanistic results. IUGR is most commonly caused by placental insufficiency leading to chronic hypoxemia and hypoglycemia [33,34]. The fetus responds to the reduced substrate supply with a range of adaptations. Neuroendocrine adaptations include a reduction in plasma IGF-1 concentration [35] and an augmentation of the normal prepartum surge in cortisol [36–38]. There are also changes in lung structure in the IUGR fetus, which include structural deficits to both the alveolar cells and the lung parenchyma [39,40].

One study attempted to address the impact of hypoxemia directly, without the complications of nutrient restriction and/or hypoglycemia. In this case, only older fetuses responded to hypoxemia with an increase in SP mRNA expression correlating with the prepartum cortisol surge [41], a result very similar to that observed with the uteroplacental embolization model of IUGR by Gagnon et al.[42]. On the other hand, in an umbilical cord-occlusion model [43], there was a significant reduction in SP-A, -B and -C mRNA expression between 130 and 133 days [43]. This decrease in SP mRNA expression occurred despite an increase in cortisol at the end of the 4-day experimental period. The authors argue that their model causes a much more severe hypoxemia (partial pressure of oxygen in the blood decreased by 70% as opposed to 30–50% in other studies), and is associated with both hypercapnia and a transient acidosis [43].

In contrast, we have used the carunclectomy model to induce chronic fetal hypoxemia and hypoglycemia across the last 4 weeks of gestation in the sheep [36,44,45]. In this model, we have shown that fetal hypoxemia decreases gene and protein expression of SP-A, -B and -C [46]. Hypoxia stabilizes a potent transcription factor, HIF-1α, which is essential for the differentiation of type II alveolar epithelial cells and SP expression [47]. However, it appears that chronic, as opposed to acute hypoxia, have differing effects on the stability of the HIF-1 complex [48]. This may explain differences in SP expression in the uteroplacental embolization and carunclectomy sheep models of IUGR. It is therefore possible that the different causes of IUGR and/or hypoxemia induce different fetal adaptations that lead to different changes in SP expression. Clearly the timing of the hypoxemic insult relative to gestational age, the duration of hypoxemia and the impact on plasma cortisol or exogenous GCs are all crucial in eliciting the SP or SP mRNA expression response, and hence lung maturation.

Impact of IUGR & antenatal GCs on the cardiovascular system

The fetus adapts to a decreased substrate supply by slowing its growth, as well as undergoing important cardiovascular adaptations [44,45]. In particular, there is a redistribution of blood flow to the brain, adrenals and heart at the expense of the peripheral organs in order to survive in a suboptimal environment. This results in relative brain sparing [45,49–51]. However, the important issue is that antenatal GCs cause cardiovascular changes in normally grown fetuses, including decreased fetal cerebral blood flow [25], decreased brain growth and increased blood pressure, that are contrary to the cardiovascular adaptations that the IUGR fetus must make to survive. Hence, there is a possibility that antenatal GC in IUGR infants may compromise cardiovascular development. In fact, in a recent study in a single umbilical artery ligation sheep model of IUGR, β-methasone caused an equivalent fall in carotid artery blood flow in IUGR fetuses, but a larger rebound increase that was not observed in control fetuses, and ultimately led to increased cerebral oxidative damage in IUGR fetuses [52]. In addition, studies in humans suggest that IUGR fetuses with absent end-diastolic blood flow have no change in middle cerebral artery blood flow in the 48 h after β-methasone treatment [53].

Conclusion

The ability of the IUGR fetus to spare brain growth is key to surviving in an intrauterine environment with a reduced substrate supply. Exposure to GCs, which alter the fetus’ ability to regulate blood pressure and, thus, blood flow, are likely to cause further decreases in fetal body and brain growth. Evidence from sheep models of IUGR suggest that the timing, duration and severity of placental insufficiency may be key indicators in the fetal surfactant response to a reduced substrate supply and the effectiveness of antenatal GCs in improving lung function and, thus, reducing the incidence of neonatal morbidity due to respiratory distress syndrome. The impact of single versus multiple doses of antenatal GC in the IUGR fetus is, for the aforementioned reasons, an even greater concern. These questions must be interrogated with randomized controlled studies to determine both short- and long-term outcomes in infants, and with mechanistic studies in animal models of human IUGR.

Financial & competing interests disclosure

Sandra Orgeig acknowledges research funding from the Australian Research Council (DP0666086). Janna L Morrison was supported by fellowships from the Heart Foundation (PF 03A 1283 and CR 07A 3328), and the National Health and Medical Research Council (Biomedical CDA 511341). Janna L Morrison acknowledges research funding from the National Health and Medical Research Council of Australia (456418 and 456421). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.