Use of tocolytics: what is the benefit of gaining 48 hours for the fetus?
Prof GC Di Renzo, Department of Gynaecology and Obstetrics, Center of Reproductive and Perinatal Medicine, University of Perugia, Policlinico Monteluce, I-06122 Perugia, Italy. Email email@example.com
Preterm birth remains one of the serious problems in perinatal medicine and is associated with an increased risk of neonatal complications and long-term morbidity. Although each day that delivery is delayed between 22 and 28 weeks of gestation increases survival by 3%, since most spontaneous preterm labour occurs between 28 and 34 weeks of gestation, this is of secondary concern; the primary goal of delay is to improve the function of certain systems in the fetus and to balance the risks of a hostile intrauterine environment with the complications of extrauterine preterm life. Although there is a lack of definitive evidence that tocolytic drugs improve outcome following spontaneous preterm labour and preterm birth, there is ample evidence that tocolysis delays delivery for long enough to permit administration of a complete course of antepartum glucocorticoids and to facilitate in utero transfer to a tertiary care unit where neonatal care will be optimal. Both these measures have been associated with improved outcomes; antepartum glucocorticoids reduce the incidence of respiratory distress syndrome, intraventricular haemorrhage, periventricular leucomalacia and necrotising enterocolitis, and in utero transfer is associated with decreased morbidity and mortality and less hospital-based intervention compared with postnatal transportation. Consequently, women who are more likely to benefit from tocolysis are those at early gestational ages, those needing transfer to a hospital that can provide neonatal intensive care and those who have not yet received a full course of antepartum glucocorticosteroids. In these cases, delaying labour for at least 48 hours with drugs such as atosiban should be considered, since it offers clear advantages for the fetus.
In spite of its relatively low incidence in most western societies, preterm birth remains one of the serious problems in perinatal medicine. Preterm birth is associated with an increased risk of neonatal respiratory complications, white matter injury, intracranial haemorrhage, cerebral palsy, subnormal neuropsychological development and school performance.1
Delaying delivery may reduce the rate of long-term morbidity by facilitating the maturation of developing organs and systems. The benefits of administration of antepartum glucocorticosteroids (AGC) to reduce the incidence and severity of respiratory distress syndrome (RDS) may be exploited by delay.2 Delay may also permit transfer of the fetus in utero to a centre with neonatal intensive care unit (NICU) facilities.2 The myth that tocolytics should work only for 48 hours arose from the inaccurate interpretation of a meta-analysis on beta-agonists, which found that 48 hours was the only consistent finding among the several studies analysed to allow comparison.2 As a result, the advantage of ‘gaining 48 hours’ is an unsubstantiated goal which nowadays has been carved in the stone. It is important to bear in mind that each day of delay between 22 and 28 weeks of gestation increases survival by 3%, and during this gestational period, if needed, tocolysis can be prolonged beyond 48 hours. Since most spontaneous preterm labour occurs between 28 and 34 weeks of gestation, survival has become a secondary issue, and the primary goal of delay is to ameliorate the function of certain systems in the fetus and to try to understand whether tocolysis is indicated, balancing the risks of a hostile intrauterine environment with the complication of extrauterine preterm life.
A wide variety of agents have been advocated as suppressing uterine contractions. Those in current use include beta-agonists, calcium channel blockers, prostaglandin synthetase inhibitors, nitric oxide donors and oxytocin receptor antagonists. There is little reliable information about current clinical practice, but it is likely that ritodrine hydrochloride, a beta-agonist, remains the most widely used in Europe.
The primary aims of tocolytic therapy are to delay delivery to allow the administration of a complete course of AGC to reduce the incidence and severity of idiopathic neonatal RDS and, eventually, to arrange in utero transfer to a centre with NICU facilities.
The secondary aim of tocolytic therapy is to delay delivery to reduce the perinatal mortality and morbidity associated with severe prematurity (22–29 weeks of gestation). Atosiban represents an advance in currently available tocolytics and should be considered a first-line tocolytic for the management of spontaneous preterm labour.2 Atosiban is licensed in Europe for treatment of spontaneous preterm labour. Duration of treatment should not exceed 48 hours, and the total dose given during a full course should preferably not exceed 330 mg of atosiban. At early gestational age with or without preterm prelabour rupture of membranes, the use of atosiban can be prolonged for a further few days without any significant adverse effects.2
As most contraindications are relative, tocolysis is preferable to the high risk of perinatal mortality or morbidity associated with very early gestational age. Maternal medical conditions, such as diabetes mellitus, could be negatively influenced by administration of tocolytic agents such as beta-agonists. Minor vaginal bleeding or spotting is often associated with spontaneous preterm labour and cervical changes but could also be because of placental abruption. If the cervix is dilated above 4–5 cm, tocolytic treatment can hardly be successful. If the pregnancy is between 22 and 28 weeks, prolongation of pregnancy by 1–2 weeks could markedly improve perinatal outcome, especially in multiple pregnancy.
If a tocolytic agent is used, beta-agonists such as ritodrine no longer seem the best choice. Alternatives such as atosiban appear to have comparable effectiveness in terms of delaying delivery from 48 hours up to 7 days and are associated with considerably fewer maternal and fetal adverse effects.2
Fetal lung immaturity is the principal contributor to neonatal mortality of preterm neonates; so the lung has been the primary focus of strategies to improve the survival of preterm neonates. If the lungs are immature, the infant will develop RDS, which still remains a leading cause of morbidity and mortality in preterm neonates.3 If untreated, around 25–30% of babies with RDS born before 28 weeks of gestation will die within 28 days of birth and another 25% will develop chronic lung disease, e.g. bronchopulmonary dysplasia, rendering affected babies dependent upon extra oxygen for long periods of life.3 A deficit of pulmonary surfactant is known to be involved in the pathophysiology of RDS, but other factors, such as the immaturity of lung structure, may play an additional role. Fetal lung maturity (FLM) is mainly the result of the enhancement (either quantitative or qualitative) of the synthesis and secretion of pulmonary surfactant and of structural changes of the lung parenchyma.
The findings by Liggins in 19694 that administration of AGC prevented the development of RDS in preterm lambs have stimulated an enormous deal of interest in the possibility of pharmacologically enhancing FLM. Many clinical studies have confirmed the findings of the first trial by Liggins and Howie in 19725 that antenatal administration of AGC to the mother is associated with a statistically significant reduction in the incidence of RDS.6,7
Betamethasone and dexamethasone are the preferred hormones used for AGC therapy. Both cross the placenta and are scarcely inactivated by placental enzymes, i.e. 11β-ol-dehydrogenase (11β-HSD), which convert active steroids into inactive 11-ketosteroids.6 Betamethasone crosses the placenta to the extent that fetal concentration is about one-third of that in maternal circulation. Betamethasone and dexamethasone have a fluoride substitute, greatly increasing glucocorticoid potency and giving negligible mineralocorticoid effect. Both drugs are given intramuscularly (IM), and this route is the most suited for clinical use because of the rapid absorption.6
AGC have also proven effective not only in reducing the incidence of RDS but also other complications, such as intraventricular haemorrhage (IVH), periventricular leucomalacia (PVL), retinopathy of prematurity, necrotising enterocolitis and patent ductus arteriosus.7−9 A meta-analysis comprising 18 trials including data on more than 3700 very preterm neonates showed that AGC administration was associated with a 50% reduction of RDS and that its effectiveness was more evident when delivery occurred after 24 hours and within 7 days of drug administration.10
AGC also significantly reduce the occurrence of neonatal death (odds ratio [OR] 0.6, 95% confidence interval [CI] 0.48–0.75).10 There is a significant reduction in IVH diagnosed both at autopsy (OR 0.29, 95% CI 0.14–0.61) and by ultrasound (OR 0.48, 95% CI 0.32–0.72). One single course of AGC may also reduce PVL and cerebral palsy. In a retrospective cohort study of 883 infants born at gestational age 24–31 weeks, betamethasone was associated with a reduced risk for cystic PVL when compared with those who did not receive AGC (OR 0.5, 95% CI 0.2–0.9).10
A complete course of AGC has also been found to be independently associated with a decreased risk for severe IVH in multiple preterm very low birthweight infants from multiple pregnancies.11 For every 11 fetuses treated with AGC, there will be one less case of RDS and a similar reduced need for postnatal surfactant treatment. There will be also one less death in the neonatal period for every 23 treated fetuses and a similar reduction in IVH. Since it is unlikely that further prospective controlled trials on antenatal corticosteroids will ever be performed, the results of the meta-analysis by Crowley7 represent the definitive evidence-based proof of the effects of AGC.
In a recent review, Jobe and Soll12 concluded that although these synthetic corticosteroids have almost identical structures, betamethasone is superior in preventing three major morbidities, i.e. RDS, IVH and neonatal death.
In a well performed study, betamethasone was found to reduce PVL, whereas dexamethasone tended to increase the risk (OR 1.5, 95% CI 0.8–2.9).13 In a prospective observational study of 201 preterm (24–34 weeks of gestation) singleton infants who received one or more antenatal doses of betamethasone or dexamethasone, multiple doses of dexamethasone were associated with an increased risk for PVL compared with betamethasone (OR 3.21, 95% CI 1.07–9.77). Furthermore, only betamethasone was associated with reduced mortality (OR 0.52, 95% CI 0.39–0.70) whereas dexamethasone was not (OR 0.89, 95% CI 0.60–1.32).13
The US National Institutes of Health (NIH) Consensus Conference has stated a general recommendation on the use of AGC for the induction of FLM in fetuses at risk of preterm birth.8 The optimal dosage to administer has been fixed as two doses of 12 mg of betamethasone given IM 24 hours apart or four doses of 6 mg of dexamethasone given IM every 12 hours. These therapeutic regimens result in about 75% of corticosteroid receptor occupancy, providing a near-maximal induction of antenatal corticosteroid-receptor-mediated response in fetal tissues.8
For infants born at 29–34 weeks of gestation, treatment with AGC clearly reduces the incidence of RDS and of overall mortality. While AGC do not clearly decrease the incidence of RDS in infants born at 24–28 weeks of gestation, they reduce its severity. More important, AGC clearly reduce mortality and the incidence of IVH in this group. All fetuses between 24 and 34 weeks of gestation should be considered candidates for this treatment, unless immediate delivery is imminent or AGC may have an adverse effect on the mother. The only absolute contraindications to the use of AGC are chorioamnionitis, peptic ulcer and tuberculosis. In infants born beyond 34 weeks of gestation, the risk of neonatal mortality, RDS and IVH is low. The use of AGC in mothers expected to deliver at more than 34 weeks is not recommended, unless there is evidence of pulmonary immaturity using amniotic fluid FLM tests.
Antenatal glucocorticosteroids: maternal adverse effects
Among maternal complications, pulmonary oedema has been reported particularly when AGC are used in combination with tocolytic agents. Since the mineralocorticoid action of betamethasone and dexamethasone is small, a potential role of AGC in sodium and water retention resulting in volume overload is considered negligible. Pulmonary oedema may develop as a consequence of the synergistic action with beta-agonist drugs, including (1) increased electrolyte and water retention in the lung, (2) an increase in mean pulmonary arterial pressure, (3) a disturbance of membrane properties and (4) decreased vascular resistance. The potential of pulmonary oedema seems to be particularly relevant in multiple pregnancy, in severely infected women or in any other circumstances in which fluids are administered injudiciously to the pregnant woman. Sporadic cases of maternal death have been reported in the 1980s and 1990s following the combined treatment of beta-agonists, especially fenoterol, with AGC. Overall, there is little evidence that one course of steroids alone causes major maternal adverse effects in healthy pregnant women. In addition, obstetricians should be cautious when AGC are administered to women with poorly controlled diabetes.
Antenatal glucocorticosteroids: fetal effects
Antenatal glucocorticosteroids result in a variety of fetal effects that are unrelated to the process of acceleration of FLM.14 They affect fetal behaviour. It is known that AGC treatment abolishes diurnal rhythms in fetal hormones. Betamethasone produces a reduction of fetal body and breathing movements and fetal heart rate (FHR) variability. Such effects are transient and return to normal on discontinuation of therapy. Conversely, dexamethasone causes an increase in FHR variation. That these effects are gestational age dependent has been shown in a study by Mulder et al.15 in fetuses exposed to betamethasone at 29–34 weeks of gestation, where a decrease in FHR on day 1 (indicative of baroreceptor reflex) and reduced breathing activity and prolonged episodes of quiescence with a concomitant decrease in body movements on days 1 and 2 were elicited. These changes were not observed if betamethasone administration occurred earlier, at 26–28 weeks of gestation.
The observed reductions in FHR variation and in body and breathing movements following betamethasone treatment might be interpreted as a sign of fetal hypoxemia, but there were no striking changes in Doppler waveform patterns in uterine artery, umbilical artery or in other blood vessels in normally grown fetuses following AGC.14 This being the case, delivery on the basis of changes in biophysical variables 2 or 3 days after betamethasone administration should be considered an error of judgement. Advantages of improved maturation might offset the risks of iatrogenic preterm birth.16 In the past, a number of fetuses have been delivered unnecessarily as a result of suspected fetal distress following administration of AGC. In case of uncertainty about fetal conditions, Doppler waveform patterns of fetal vessels may be recorded and/or FHR monitored, with special emphasis to the occurrence of FHR decelerations.16
AGC exert a direct and potent vasodilatory effect on human umbilical artery resistance in vitro, providing an explanation for the previously unexplained vascular effects associated with administration of AGC. These effects are of considerable interest, especially in growth-restricted fetuses. These agents are known to have a vasodilatory effect on the human umbilical artery in vitro and to improve umbilical vascular dynamics transiently in many fetuses with absent end-diastolic flow.14
In human pregnancy, uterine and umbilical arterial (UA) blood flow velocity waveform patterns detected using Doppler ultrasound are considered to reflect placental resistance. It has been shown that UA blood flow velocity and pulsatility do not change following betamethasone treatment in normal or growth-restricted fetuses, provided that UA blood flow velocity waveform is normal before betamethasone or dexamethasone administration.16−18 Other studies have shown a return of end-diastolic flow in the umbilical artery following betamethasone treatment in human pregnancies complicated by absent end-diastolic flow.19 Simchen et al.20 have identified a small subset of fetuses with intrauterine growth restriction who rather than responding to steroids with a decrease in vascular resistance, develop increased resistance. This led to poor outcomes, presumably the result of an increase in systemic mean arterial pressure leading to an incremental increase in the after-load of very sick fetuses with impending cardiac failure.20 These studies suggest that betamethasone has vasodilatory effects on the placental bed in the presence of high vascular resistance.17 However, the dynamic changes of UA blood flow may be a reason for the difficulties in detecting effects of AGC treatment under clinical conditions. The long time interval between betamethasone treatment and Doppler examinations under clinical conditions may explain the difficulties in demonstrating consistent effects of betamethasone on the umbilical circulation.
The dynamics of FHR and, consequently, of cardiac output and UA changes in response to betamethasone are probably the result of multifactorial AGC effects.16 It is conceivable that fetal blood pressure increases after each betamethasone injection and activates a baroreflex response that leads to the initial FHR decrease. An FHR increase following an initial bradycardia has also been shown in sheep after maternal injection of 12 mg of dexamethasone.21 Fetal injection of betamethasone just before delivery also led to an increased heart rate and cardiac output in a newborn lamb.22 A delayed FHR increase has also been reported following infusion of dexamethasone or betamethasone directly to the near-term sheep fetus.23
Antenatal betamethasone exposure increases fetoplacental perfusion, and glucocorticoid-induced fetal growth restriction is not likely to be caused by reduction of placental perfusion. The dynamic betamethasone effects on fetal and fetoplacental circulation are mainly determined by changes in fetal cardiac output and not reflected in changes of UA waveforms indices.17 With relevance to the human fetus, the results show the limitations of conclusions based on UA velocity measurements and Doppler indices regarding placental perfusion.
Antenatal steroid administration for accelerating fetal lung maturation has been in use for more than 30 years; yet, controversy still exists regarding potential long-term adverse effects. Studies in animals have shown that administration of AGC during pregnancy alters renal expression of several key regulatory molecules at different developmental stages.24,25 These alterations are followed in most cases with the development of hypertension in the adult offspring. These observations have been substantiated by epidemiological studies in human beings showing an increase in hypertensive disorders in adults treated with AGC during intrauterine life.26
In sheep, administration of natural or synthetic glucocorticoids at 28 days of gestation is associated with the development of high blood pressure and a decrease in nephron number.25,27 The development of adult hypertension may stem from a derangement of fetal kidney development, and it has been reported that a single course of betamethasone given at a gestational age and dose regimen equivalent to that currently used in clinical practice induced a 25% decrease in nephron number in late gestation.28,29
In conclusion, fetal adverse effects in the short term appear to be easily detectable and not worrisome, and obstetricians and mothers should be aware of them in order to avoid inappropriate emergency delivery. At long-term follow up, it is still unclear how many adverse effects can be attributed to AGC, and some may appear in adult life. This should be weighed against the fact that the avoidance of RDS and IVH by AGC undoubtedly decreases neonatal morbidity and therefore adult morbidity. A summary of current recommendations on the use of AGC is outlined in Table 1. The advantage of gaining 2–3 days in cases of spontaneous preterm labour is clear.
Table 1. Summary of current recommendations on the use of antenatal glucocorticosteroids. Reproduced with permission from Di Renzo GC, et al. J Perinat Med 2006.2
|Administration of one single course of ACG is the most important treatment to prevent RDS and brain injury and increase survival that can be provided by the obstetrician to women at risk of preterm delivery at 24–34 weeks of gestation|
|Based on observational clinical and animal studies, betamethasone is preferable to dexamethasone|
|There is no direct evidence that tocolytic treatment per se might affect the risk of perinatal brain injury or adverse neurological outcome|
In everyday clinical practice, it is not uncommon that delivery occurs before the completion of the course of AGC, i.e. two doses of 12 mg of betamethasone 24 hours apart. To establish whether beneficial effects may still ensue, an incomplete course of antenatal corticosteroids was associated with reduction in the need for vasopressors, the rate of IVH and neonatal death in preterm neonates in one study.30 One dose of AGC given 4–24 hours before delivery was clinically comparable in terms of reduction of IVH and mortality with the recommended schedule of the NIH in surfactant-treated preterm infants.30−32
Transfer in utero
Antenatal transfer guarantees a significantly better neonatal outcome with respect to severe neonatal morbidity than postnatal transport and compares favourably with inborn admissions.33,34 In cases of spontaneous preterm labour at 24–34 weeks of gestation admitted to an obstetric unit not linked to a NICU, it is of overwhelming importance to inhibit uterine contractions (unless absolutely contraindicated) and arrange transfer in utero to the nearest referral unit. The stabilisation of the mother and the fetus is essential before transport, and modifications of homeostasis due to transport can be evaluated by a simple scoring system.35,36 In a few cases, transport can be carried out by air with specially equipped helicopters or aeroplanes.37 Sophisticated neonatal transport has improved the safety of transporting preterm neonates, but specialised neonatal transport and advanced neonatology techniques have not removed the significant advantage of decreased morbidity, mortality and length of hospital intervention resulting from in utero transport.38,39
As soon as the diagnosis of spontaneous preterm labour has been made,2 it is recommended that neonatologists involved in management decisions are informed to ensure that a NICU cot is available on site or that an in utero transfer to a centre with NICU facilities may be arranged.2 Despite this and the possibility of treatment with supplementary surfactant to neonates, RDS still remains a major health problem. This is also due to the fact that not all preterm births have been exposed to therapeutic doses of AGC and that RDS is not solely the consequence of lack of surfactant.32
In the absence of clear evidence that tocolytic drugs improve outcome following spontaneous preterm labour and preterm birth, it is reasonable not to use them. Women who are more likely to benefit from tocolysis are those at a very early gestational age, those needing transfer to a hospital that can provide neonatal intensive care or those who have not received a full course of AGC to promote FLM. For these women, tocolytic drugs should be considered for at least 48 hours, which can offer a clear advantage for the fetus.40