Seminars in Fetal & Neonatal Medicine 22 (2017) 290e295

Contents lists available at ScienceDirect

Seminars in Fetal & Neonatal Medicine journal homepage: www.elsevier.com/locate/siny

Postnatal corticosteroids to prevent or treat bronchopulmonary dysplasia e Who might benefit? Lex W. Doyle a, b, c, d, *, Jeanie L.Y. Cheong a, b, c, d a

Newborn Services, The Royal Women's Hospital, Parkville, Australia Department of Obstetrics and Gynaecology, University of Melbourne, Parkville, Australia c Department of Paediatrics, University of Melbourne, Parkville, Australia d Clinical Sciences, Murdoch Children's Research Centre, Parkville, Australia b

a b s t r a c t Keywords: Bronchopulmonary dysplasia Corticosteroids Inhaled Systemic

Newborn infants born very preterm are at high risk of developing bronchopulmonary dysplasia, which is associated with not only mortality but also adverse long-term neurological and respiratory outcomes in survivors. Postnatal corticosteroids might reduce the risk of developing bronchopulmonary dysplasia, or reduce its severity. However, it is important to minimize exposure to the potentially harmful effects of corticosteroids, particularly on the developing brain. Systemic corticosteroids started after the first week of life have shown the most benefit in infants at highest risk of developing bronchopulmonary dysplasia, whereas inhaled corticosteroids have little effect in children with established lung disease. Systemic corticosteroids in the first week of life are not recommended, but inhaled corticosteroids, or corticosteroids instilled into the trachea using surfactant as a vehicle to distribute the corticosteroids through the lungs, offer promise with respect to prevention of bronchopulmonary dysplasia. © 2017 Elsevier Ltd. All rights reserved.

1. Introduction Corticosteroids to prevent or treat bronchopulmonary dysplasia (BPD) are controversial. Because of high rates of adverse outcomes, including gastrointestinal perforation in the short term, and cerebral palsy in the long term, many esteemed bodies, such as the American Pediatric Society, the Canadian Paediatric Society [1], and the European Association of Perinatal Medicine [2], have warned against the routine use of corticosteroids in high-risk infants. Although such warnings have resulted in a reduction in prescribing, corticosteroids remain a common treatment in very preterm infants, partly because BPD itself is associated with adverse longterm neurological and respiratory outcomes, and there are no alternative therapies known to be as effective in facilitating extubation in preterm children with severe lung disease. Clearly clinicians need guidance in balancing the risks versus benefits of corticosteroids in high-risk infants. Corticosteroids could be administered either directly to the lung via the tracheobronchial tree, or indirectly via the bloodstream,

* Corresponding author. Department of Obstetrics & Gynaecology, The Royal Women's Hospital, 20 Flemington Road, Parkville, Victoria, 3052, Australia. E-mail address: [email protected] (L.W. Doyle). http://dx.doi.org/10.1016/j.siny.2017.07.003 1744-165X/© 2017 Elsevier Ltd. All rights reserved.

through either parenteral or enteral administration. Direct administration to the lung is intuitively sensible, as it would limit exposure of non-lung organs to the potential harmful effects of corticosteroids. Corticosteroids can be aerosolized and added to the air the baby is breathing, but only a small proportion of the drug eventually reaches the lung. To overcome the problem of poor delivery of the drug, an alternative method is to add the corticosteroid to exogenous surfactant, then inject the combination into the trachea. Regardless of the mode of administration, it is important to know that the nett benefits of treatment with corticosteroids to prevent or treat BPD exceed the nett harms. The purpose of this review is to weigh the evidence for and against corticosteroids after birth to prevent or treat BPD. We consider different ways of administering corticosteroids, including inhaled, intratracheal, and systemic. Evidence from randomized controlled trials (RCTs), and syntheses of RCTs are relied upon the most to form our views. Although BPD is the major outcome of interest in this review, we consider competing risks of death, as well as other possible harms, including short-term complications and adverse long-term neurological sequelae, such as cerebral palsy. 2. Inhaled corticosteroids There are two reviews in the Cochrane Library on the topic of

L.W. Doyle, J.L.Y. Cheong / Seminars in Fetal & Neonatal Medicine 22 (2017) 290e295

inhaled corticosteroids; one on early use to prevent BPD [3], and the other considers trials when lung disease is more established [4]. 2.1. Early inhaled corticosteroids Shah et al. [3] synthesized the evidence concerning early (<2 weeks after birth) inhaled corticosteroids. For the outcome of death or BPD at 36 weeks, there were six trials that recruited 1285 participants. Fewer infants in the inhaled corticosteroid group had either died or had BPD at 36 weeks than controls: 35% (227/649) vs 40% (256/636), respectively [risk ratio (RR): 0.86; 95% confidence interval (CI): 0.75, 0.99; P ¼ 0.04]. In the inhaled corticosteroid group, there were fewer survivors with BPD at 36 weeks: 24% (131/ 544) vs 31% (171/554), respectively (RR: 0.76; 95% CI: 0.63, 0.93; P ¼ 0.005). There were no obvious differences between inhaled corticosteroid and comparison groups for complications such as patent ductus arteriosus (PDA), hyperglycemia, hypertension, sepsis, or gastrointestinal bleeding. The results of the systematic review are dominated by the trial of Bassler et al. [5]; prior to the inclusion of the Bassler et al. study there was little evidence of any effects of inhaled corticosteroids [6]. Bassler et al. recruited 863 infants born 23e27 weeks gestation who at <12 h required positive pressure support. Infants were randomly allocated to either budesonide or placebo, delivered via a spacer. The dose of budesonide was 400 mg 12-hourly for 14 days, then 200 mg 12-hourly until either 32 weeks postmenstrual age, or they were off respiratory support. In their study 40% (175/437) of infants in the budesonide group had died or had BPD by 36 weeks compared with 46% (194/419) in the placebo group (RR: 0.86; 95% CI: 0.75, 1.00; P ¼ 0.05). Interestingly, the improvement was influenced by a substantial reduction in the rate of BPD [budesonide 28% (101/363) vs placebo 38% (138/363); RR: 0.74; 95% CI: 0.60, 0.91; P ¼ 0.004] at the expense of a small increase in mortality by 36 weeks (17% (74/437) budesonide vs 14% (57/419) placebo; RR: 1.24; 95% CI: 0.91, 1.69); P ¼ 0.17). Moreover, approximately twothirds of the infants enrolled were intubated and ventilated at trial entry, among whom there was less effect of inhaled budesonide on the outcome of death or BPD (49% budesonide group vs 51% control), whereas among the remaining one-third who were not intubated on entry to the trial, there was a larger effect of inhaled budesonide on the outcome of death or BPD (21% budesonide group vs 36% control). This discrepancy does not make clinical sense, unless it suggests that the intubated babies had more established lung disease and that too much damage had already occurred to the lung prior to starting the inhaled corticosteroids. Bassler et al. also reported some other benefits for inhaled budesonide, including lower rates of PDA and reintubation. Interestingly almost one-third in both groups ultimately received systemic corticosteroids (29% budesonide vs 32% controls). Follow-up of the cohort is in progress, and may help to understand the effect of inhaled corticosteroids on the balance between the competing risks of death and BPD. Long-term outcomes have been reported for two trials in the systematic review; there was little evidence of effects of inhaled corticosteroids on rates of neurodevelopmental impairment or readmission to hospital for respiratory illnesses [7,8]. 2.2. Later inhaled corticosteroids Onland et al. [4] reviewed the evidence concerning later inhaled (7 days after birth) corticosteroids. There were not many trials (n ¼ 8) or participants (n ¼ 232) overall, and few studies reported important outcomes such as mortality or BPD, and hence there is little evidence on which to make any clinical recommendations concerning inhaled corticosteroids in infants with established lung disease.

291

3. Intratracheal administration of corticosteroids There is only one RCT that has assessed intratracheal administration of corticosteroids. Yeh et al. [9] recruited 265 infants with birthweight <1500 g, with a chest X-ray consistent with severe respiratory distress syndrome, and who were intubated and ventilated, requiring 50% oxygen, and who were <4 h after birth. Infants were randomly allocated to budesonide 0.25 mg/kg mixed with 100 mg/kg surfactant or to surfactant only. The surfactant was Survanta® (AbbVie, Inc., North Chicago, IL, USA). Doses of study medication were given every 8 h until the oxygen requirement was <30%. Death or BPD at 36 weeks was less frequent in the budesonide group (42%; 55/131) than in the controls (66%; 89/134) (RR: 0.58; 95% CI: 0.44, 0.77; P < 0.001). The rates of BPD alone at 36 weeks were 29% (38/131) in the budesonide group vs 50% (67/134) in the control group (RR: 0.70; 95% CI: 0.58, 0.86; P < 0.001), whereas death alone at 36 weeks was not reduced as much: 13% (17/131) budesonide vs 16% (22/134) controls (RR: 0.96; 95% CI: 0.87, 1.06; P ¼ 0.54). Follow-up of survivors to 30 months of age is in progress, but thus far there have been no substantial differences observed in rates of neurosensory impairment or in Bayley Scale scores [9]. This study has treatment effects that are large and would be very important if replicated. 4. Systemic corticosteroids There are two Cochrane reviews in which systemic corticosteroids are compared with controls, one focusing on treatment starting early (<8 days of age) [10], the other in which treatment starts later (>7 days of age) [11]. 4.1. Early (<8 days of age) systemic corticosteroids Of the corticosteroids available for systemic treatment, randomized trials have only been reported where dexamethasone or hydrocortisone were the primary drugs of interest. There are 20 early randomized controlled trials in which outcomes in infants treated with dexamethasone have been compared with controls, and nine trials in which infants treated with hydrocortisone have been compared with controls [10]. 4.1.1. Early dexamethasone There are numerous benefits from early dexamethasone, including lower rates of BPD at both 28 days and 36 weeks, and regarding the combined outcomes of BPD at both ages with death, lower rates of failure to extubate, PDA and retinopathy of prematurity (ROP), and less need for later systemic corticosteroids (Table 1). These benefits are counteracted by a number of harms, most notably more gastrointestinal perforation in the short term, and more cerebral palsy in the long term (Table 1). On balance, the benefits of systemic dexamethasone starting in the first week of life do not outweigh the harms, particularly cerebral palsy, and dexamethasone in the first week of life is not recommended to prevent BPD. 4.1.2. Early hydrocortisone In the Cochrane review there are few effects of early hydrocortisone in the doses used in the RCTs [10]. However, since the Cochrane review was published in 2014 there has been a large randomized trial of early hydrocortisone reported by Baud et al., 2016 [12]. They recruited 523 inborn infants 24e27 weeks gestation in the first 24 h after birth who were considered likely to survive the immediate newborn period. Infants were randomly allocated to either 8.5 mg/kg of hydrocortisone or placebo for a total of 10 days. Of the infants treated with hydrocortisone, 60% (153/

292

L.W. Doyle, J.L.Y. Cheong / Seminars in Fetal & Neonatal Medicine 22 (2017) 290e295

Table 1 Some benefits and harms from early systemic dexamethasone. Outcome Benefits BPD at 28 days Death or BPD at 28 days BPD at 36 weeks Death or BPD at 36 weeks Failure to extubate by day 7 PDA Severe ROP Late rescue with corticosteroids Harms Hyperglycemia Hypertension Gastrointestinal perforation Gastrointestinal hemorrhage Cerebral palsy Death or cerebral palsy

No. of studies

Dexamethasone

Control

RR (95% CI)

P-value

16 14 15 15 7 17 8 10

40.6% 65.1% 19.8% 43.1% 37.1% 30.5% 11.8% 35.3%

48.6% 72.1% 28.3% 49.8% 50.4% 40.3% 15.4% 49.8%

(629/1293) (820/1138) (350/1235) (615/1233) (236/468) (539/1339) (115/745) (489/981)

0.85 0.91 0.70 0.87 0.75 0.76 0.77 0.72

(0.78, (0.86, (0.61, (0.80, (0.65, (0.69, (0.60, (0.65,

0.92) 0.96) 0.81) 0.94) 0.86) 0.84) 0.99) 0.80)

<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.043 <0.001

12 10 9 10 7 7

42.8% (458/1067) 21.6% (211/978) 7.2% (70/968) 11.2% (98/876) 15.9% (75/472) 44.9% (212/472)

31.7% (335/1050) 11.9% (115/965) 4.1% (40/968) 5.9% (50/849) 8.9% (40/449) 38.3% (172/449)

1.35 1.84 1.73 1.87 1.75 1.17

(1.21, (1.53, (1.20, (1.35, (1.20, (1.00,

1.49) 2.21) 2.51) 2.58) 2.55) 1.37)

<0.001 <0.001 0004 <0.001 0.004 0.045

(539/1328) (752/1155) (247/1249) (538/1248) (181/488) (417/1367) (90/762) (353/993)

RR, risk ratio; CI, confidence interval; BPD, bronchopulmonary dysplasia; PDA, patent ductus arteriosus; ROP, retinopathy of prematurity.

255) survived free of BPD at 36 weeks compared with 51% (136/ 266) in the placebo group (RR: 1.48; 95% CI; 1.02, 2.16; P ¼ 0.04). Both death and BPD were approximately equally affected by hydrocortisone; the mortality and BPD rates at 36 weeks, respectively, were 18% and 22% in the hydrocortisone group, and 23% and 26% in the placebo group. Rates of gastrointestinal perforation were 5% in the hydrocortisone group and 4% in the placebo group. A brain magnetic resonance imaging (MRI) scan was obtained on 85% of survivors at term-equivalent age e there was no difference in the rates of MRI abnormalities in the two groups (30% in both groups). Adding data from the Baud study to the Cochrane review, there are still few obvious short-term benefits, although the reduction in the combined rate of death or BPD at 36 weeks with hydrocortisone is on the borderline of significance. However, the rate of gastrointestinal perforation remains significantly higher (Table 2). Results of follow-up to a median age of 22 months, corrected for prematurity, have recently been reported from the Baud et al. study [13]. Of the 523 who were randomly assigned, 406 survived to 22 months, and 93% (379) were evaluated (194 in the hydrocortisone group, 185 in the placebo group). Rates of neurodevelopmental impairment were similar in both groups (27% hydrocortisone; 30% placebo), as were rates of cerebral palsy (6% hydrocortisone; 5% placebo) and mean developmental scores (91.7 hydrocortisone; 91.4 placebo). Combining all studies with long-term outcomes, rates of cerebral palsy or death combined with cerebral palsy were not substantially different between hydrocortisone and control groups (Table 2). On balance, the marginal benefit of systemic hydrocortisone starting in the first week of life to prevent BPD probably does not outweigh the potential harm from gastrointestinal perforation, but more long-term follow-up data on neurological function and respiratory health may help to tip the balance. Currently, hydrocortisone in the first week of life is not recommended to prevent BPD.

4.2. Late (>7 days of age) systemic corticosteroids All of the reported randomized trials in which treatment started after the first week of life have used dexamethasone, either entirely or mostly [11]. There are two current RCTs in progress testing hydrocortisone after the first week of life, but there are no results to date. The first trial is being conducted by Onland et al. and is known as the SToP-BPD study [14]. They will recruit 400 infants either <1250 g birthweight or <30 weeks gestational age who are ventilator dependent between 7 and 14 days after birth. The total dose of hydrocortisone is 72.5 mg/kg over 22 days. The primary outcome is mortality or BPD at 36 weeks. Open-label treatment with corticosteroids will be permitted. The second trial is being conducted within the NICHD network [15]. They will recruit 800 infants born <30 weeks gestational age who are intubated between 14 and 28 days after birth. They will give 18 mg/kg of hydrocortisone over 10 days. Major outcomes are mortality or moderateesevere BPD at 36 weeks, and survival free of moderateesevere disability at 22e26 months corrected age. As with early dexamethasone use, there are some clear benefits from late dexamethasone, but also some harms, albeit all short term, with little evidence for long-term harm (Table 3). The major benefits include substantial reductions in the rates of BPD, and of death or BPD, lower rates of extubation failure, and less need for rescue corticosteroids and for home oxygen therapy. The harms are short term, including higher blood glucose and blood pressure, and more severe ROP. Although there is a higher rate of children having abnormal neurological examinations at follow-up, this is based on only four RCTs and a total of 200 infants randomized. By contrast, there is little effect on a diagnosis of cerebral palsy, with or without mortality, which is based on 15 RCTs and 855 infants randomized (Table 3). Moreover, the higher rate of severe ROP in the dexamethasone group does not translate into more children having

Table 2 Some outcomes from early systemic hydrocortisone. Outcome

No. of studies

Hydrocortisone

Control

RR (95% CI)

P-value

Death at 36 weeks BPD at 36 weeks Death or BPD at 36 weeks Gastrointestinal perforation Severe IVH Cerebral palsy Death or cerebral palsy

7 7 8 7 9 6 6

16.7% (109/654) 35.3% (231/654) 52.3% (351/670) 8.1% (44/545) 14.4% (97/674) 7.6% (35/459) 25.6% (133/520)

19.3% (129/669) 38.1% (255/669) 57.5% (395/687) 4.7% (26/559) 15.2% (105/693) 7.3% (33/451) 29.5% (157/532)

0.84 0.92 0.91 1.70 0.95 1.02 0.86

0.22 0.25 0.05 0.02 0.71 0.93 0.15

RR, risk ratio; CI, confidence interval; BPD, bronchopulmonary dysplasia; IVH, intraventricular hemorrhage.

(0.63, (0.80, (0.83, (1.07, (0.74, (0.64, (0.71,

1.11) 1.06) 1.00) 2.70) 1.23) 1.62) 1.05)

L.W. Doyle, J.L.Y. Cheong / Seminars in Fetal & Neonatal Medicine 22 (2017) 290e295

293

Table 3 Some benefits and harms from late systemic dexamethasone. Outcome Benefits Death at 28 days BPD at 28 days Death or BPD at 28 days BPD at 36 weeks Death or BPD at 36 weeks Failure to extubate by day 7 Late rescue with corticosteroids Discharge home on oxygen Harms Hyperglycemia Glycosuria Hypertension Cardiomyopathy Severe ROP Abnormal neurological examination Cerebral palsy Death or cerebral palsy

No. of studies

Dexamethasone

Control

RR (95% CI)

P-value

8 6 5 9 9 10 13 7

5.5% (18/327) 75.2% (230/306) 77.9% (215/276) 47.1% (128/272) 58.5% (159/272) 52.0% (128/246) 15.0% (81/539) 19.5% (59/303)

10.6% 87.1% 93.7% 63.1% 77.6% 80.5% 31.8% 27.3%

(35/329) (276/317) (269/287) (166/263) (204/263) (202/251 (177/557) (84/308)

0.49 0.87 0.84 0.76 0.76 0.64 0.47 0.71

(0.28, (0.81, (0.78, (0.66, (0.68, (0.56, (0.38, (0.54,

0.85) 0.94) 0.89) 0.88) 0.85) 0.74) 0.59) 0.94)

0.01 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.002

16 2 14 4 12 4 15 15

28.5% (181/635) 79.2% (19/24) 9.9% (58/588) 19.3% (23/119) 24.2% (93/385) 27.2% (28/103) 13.4% (58/432) 30.1% (130/432)

20.1% (128/636) 8.3% (2/24) 4.9% (29/587) 6.7% (8/119) 23.8% (65/273) 16.5% (16/97) 12.1% (51/423) 31.2% (132/423)

1.50 8.03 2.12 2.76 1.38 1.81 1.12 0.95

(1.25, (2.43, (1.45, (1.33, (1.07, (1.05, (0.79, (0.77,

1.80) 26.5) 3.10) 5.74) 1.79) 3.11) 1.60) 1.16)

<0.001 <0.001 <0.001 0.006 0.01 0.033 0.51 0.58

RR, risk ratio; CI, confidence interval; BPD, bronchopulmonary dysplasia; PDA, patent ductus arteriosus; ROP, retinopathy of prematurity.

blindness at follow-up [dexamethasone 2.2% (8/365) vs control 2.8% (10/355); RR: 0.78; 95% CI: 0.35, 1.73; P ¼ 0.54].

4.2.1. Is there a higher-risk group who might benefit from late systemic dexamethasone? This question was first addressed in 2005 in a meta-regression analysis of the risk difference in death or cerebral palsy regressed against the risk of BPD [16]. The conclusion of that study was that if the risk of BPD was >50%, then one could consider using systemic dexamethasone. Since the initial analysis, more data have become available and the meta-regression was updated in 2014, with the conclusions largely unchanged e the confidence intervals were a little narrower and the regression line was slightly shifted downwards (Fig. 1) [17]. The biggest practical problem is predicting the risk for an individual infant to develop BPD. Prediction equations have been produced [18], but local data for the risk of BPD should preferably be used for individual centers. In practice, dexamethasone seems to be withheld until infants are several weeks or more after birth, are ventilator dependent via an endotracheal tube, and need considerable inspired oxygen concentrations, typically >40%. However, with the advent of more non-invasive forms of assisted

Fig. 1. Risk difference (RD, %) for death or cerebral palsy (CP) among all participants vs rate of bronchopulmonary dysplasia (BPD, %) in the control group. Each study is shown by a circle whose area is proportional to that study's weight. Regression line and its 95% confidence interval are shown. CS, corticosteroids (Adapted from Doyle et al. [17].).

ventilation, including nasal CPAP and high-flow intranasal gas delivery, some non-intubated infants are being commenced on systemic corticosteroids in an attempt to avoid intubation or BPD, but with no data from RCTs to support such practices. 5. Systemic versus inhaled corticosteroids There are two reviews in the Cochrane Library on the topic of systemic versus inhaled corticosteroids; one on early use to prevent BPD [19], and the other comprises trials in which lung disease is more established [20]. 5.1. Early systemic vs inhaled corticosteroids For the prevention of BPD, there are only two studies included in the review [19]. Groneck et al. [21] recruited 16 infants born weighing <1200 g, who were receiving mechanical ventilation after 3 days of age. Infants were alternately allocated to inhaled beclomethasone 1.5 mg/day from day 3 to day 28, or to systemic dexamethasone 0.5 mg/kg/day from day 11 to day 13, and subsequently tapered over 10e28 days. The trial stopped early after six out of seven infants in the inhaled group developed BPD. Halliday et al. [22] enrolled infants who required mechanical ventilation before 72 h after birth who were either <30 weeks gestation and in >30% oxygen, or who were 30e31 weeks gestation and in >50% oxygen. They were randomly allocated to four groups: (i) early (<72 h) dexamethasone, n ¼ 135; (ii) early (<72 h) inhaled budesonide, n ¼ 143; (iii) delayed (>15 days) selective dexamethasone, n ¼ 150; or (iv) delayed (>15 days) selective inhaled budesonide, n ¼ 142. The total dexamethasone dose was 2.7 mg/kg over 12 days. The inhaled budesonide dose was 0.8 mg/day if the infant weighed <1000 g, or 1.2 mg/day if the infant weighed 1000e1500 g, and was given for up to 12 days if the infant remained intubated. The combined rate of BPD at 28 days from the two trials was 61% (92/150) in the inhaled group, compared with 51% (73/144) in the dexamethasone group (RR: 1.21; 95% CI: 0.98, 1.48; P ¼ 0.072). The rates of death or BPD at 28 days from the two trials combined were 78% (117/150) in the inhaled group and 74% (106/144) in the dexamethasone group (RR: 1.05; 95% CI: 0.93, 1.20; P ¼ 0.43). Only the Halliday trial reported BPD at 36 weeks, which was 34% (49/ 143) in the inhaled group, compared with 24% (32/135) in the dexamethasone group (RR: 1.45; 95% CI: 0.99, 2.11; P ¼ 0.056), and the rates of death or BPD at 36 weeks were 57% (82/143) in the

294

L.W. Doyle, J.L.Y. Cheong / Seminars in Fetal & Neonatal Medicine 22 (2017) 290e295

inhaled group and 53% (71/135) in the dexamethasone group (RR: 1.09; 95% CI: 0.88, 1.35; P ¼ 0.43). For other outcomes, inhaled corticosteroids were associated with longer durations of assisted ventilation (mean increase: 3.9 days; 95% CI: 0.2, 7.6; P ¼ 0.037) and of oxygen therapy (mean increase: 11.1 days; 95% CI: 2.0, 20.2; P ¼ 0.017). On a positive note, inhaled corticosteroids were associated with less hyperglycemia. 5.2. Delayed systemic versus inhaled corticosteroids There are three studies included in the review when treatment was started when lung disease was more established [20]. In the study of Halliday et al. [22], of those randomly allocated to delayed treatment, 60/150 allocated to delayed (>15 days) selective dexamethasone were ultimately treated, as were 54/142 allocated to delayed (>15 days) selective inhaled budesonide. Suchomski et al. [23] recruited infants <2001 g birthweight who were receiving mechanical ventilation and oxygen at age 14 days. They were randomly allocated to three groups: (i) inhaled beclomethasone 400 mg/day, n ¼ 26; (ii) inhaled beclomethasone 800 mg/day, n ¼ 25; or (iii) systemic dexamethasone 42-day course, n ¼ 27. Rozycki et al. [24] enrolled infants <1500 g birthweight or <31 weeks gestation who were on mechanical ventilation and oxygen at 14 days of age. They were randomly allocated to four groups: (i) aerosol placebo, systemic dexamethasone, n ¼ 15; (ii) high-dose beclomethasone (2.4e3.7 mg/kg/day), systemic placebo, n ¼ 16; (iii) medium-dose beclomethasone (1.0e1.8 mg/kg/day), systemic placebo, n ¼ 15; or (iv) low-dose beclomethasone (0.5e0.7 mg/kg/d), systemic placebo, n ¼ 15. There were no substantial differences between treatment groups in rates of BPD at 36 weeks, or death or BPD at 36 weeks [20]. There were no other obvious benefits or harms [20]. 6. Conclusions The collective evidence concerning the risks versus benefits of inhaled versus systemic corticosteroids is weak, with few studies and small sample sizes, lack of rigour, such as random allocation in some trials, and all studies allowed open-label use of corticosteroids. In addition, the complexity of some of the study designs, with multiple levels of dosing and timing, does not help with interpretation. On balance, there is little to support the efficacy of inhaled corticosteroids over systemic corticosteroids when started in the first week of life in reducing lung disease in the short term. Similarly, there is little evidence to support inhaled corticosteroids over systemic corticosteroids when treatment is started in infants with established lung disease after the first few weeks of life. Systemic corticosteroids started after the first week of life have shown the most benefit in infants at highest risk of developing BPD, whereas inhaled corticosteroids alone have little effect in children with established lung disease. Systemic corticosteroids in the first week of life are not recommended, but inhaled corticosteroids, or corticosteroids instilled into the trachea using surfactant as a vehicle to distribute the corticosteroids through the lungs, offer promise with respect to prevention of BPD. Practice points  Systemic dexamethasone started after the first week of life has shown the most benefit in infants at highest risk of developing bronchopulmonary dysplasia.  Systemic corticosteroids in the first week of life are not recommended.  There is currently not enough evidence to support inhaled corticosteroids over systemic corticosteroids at any age.

Research directions  Inhaled corticosteroids, or corticosteroids instilled into the trachea using surfactant as a vehicle to distribute the corticosteroids through the lungs, soon after birth are promising in preventing BPD, but more research is required, particularly on the long-term effects of treatment on the brain and lung.  Results of current RCTs comparing corticosteroids other than dexamethasone to treat ventilator-dependent infants after the first week of life are awaited.

Conflict of interest statement None declared. Funding sources Supported in part by Project Grant No. 108700, Program Grant No. 606789, and Centre of Research Excellence No. 1060733 from the National Health and Medical Research Council (NHMRC) of Australia. The funding sources had no role in the collection, analysis or interpretation of data, or in the writing of the manuscript. References [1] American Academy of Pediatrics. Committee on fetus and newborn, and canadian paediatric society, fetus and newborn committee. Postnatal corticosteroids to treat or prevent chronic lung disease in preterm infants. Pediatrics 2002;109:330e8. [2] Halliday HL. Guidelines on neonatal steroids. Prenat Neonat Med 2001;6: 371e3. [3] Shah VS, Ohlsson A, Halliday HL, Dunn M. Early administration of inhaled corticosteroids for preventing chronic lung disease in very low birth weight preterm neonates. Cochrane Database Syst Rev 2017;1:CD001969. [4] Onland W, Offringa M, van Kaam A. Late (7 days) inhalation corticosteroids to reduce bronchopulmonary dysplasia in preterm infants. Cochrane Database Syst Rev 2012;4:CD002311. [5] Bassler D, Plavka R, Shinwell ES, Hallman M, Jarreau PH, Carnielli V, et al. Early inhaled budesonide for the prevention of bronchopulmonary dysplasia. N Engl J Med 2015;373:1497e506. [6] Shah VS, Ohlsson A, Halliday HL, Dunn M. Early administration of inhaled corticosteroids for preventing chronic lung disease in ventilated very low birth weight preterm neonates. Cochrane Database Syst Rev 2012;5: CD001969. [7] Jangaard KA, Stinson DA, Allen AC, Vincer MJ. Early prophylactic inhaled beclomethasone in infants less than 1250 g for the prevention of chronic lung disease. Paediatr Child Health 2002;7:13e9. [8] Nakamura T, Yonemoto N, Nakayama M, Hirano S, Aotani H, Kusuda S, et al. Early inhaled steroid use in extremely low birthweight infants: a randomised controlled trial. Archs Dis Childh Fetal Neonatal Ed 2016;101(6):F552e6. [9] Yeh TF, Chen CM, Wu SY, Husan Z, Li TC, Hsieh WS, et al. Intratracheal administration of budesonide/surfactant to prevent bronchopulmonary dysplasia. Am J Respir Crit Care Med 2016;193:86e95. [10] Doyle LW, Ehrenkranz RA, Halliday HL. Early (<8 days) postnatal corticosteroids for preventing chronic lung disease in preterm infants. Cochrane Database Syst Rev 2014;5:CD001146. [11] Doyle LW, Ehrenkranz RA, Halliday HL. Late (>7 days) postnatal corticosteroids for chronic lung disease in preterm infants. Cochrane Database Syst Rev 2014;5:CD001145. [12] Baud O, Maury L, Lebail F, Ramful D, El Moussawi F, Nicaise C, et al. Effect of early low-dose hydrocortisone on survival without bronchopulmonary dysplasia in extremely preterm infants (PREMILOC): a double-blind, placebocontrolled, multicentre, randomised trial. Lancet 2016;387:1827e36. [13] Baud O, Trousson LC, Biran V, Leroy E, Mohamed D, Alberti C, for the Premiloc Trial Group. Association between early low-dose hydrocortisone therapy in extremely preterm neonates and neurodevelopmental outcomes at 2 years of age. JAMA 2017;317:1329e37. [14] Onland W, Offringa M, Cools F, De Jaegere AP, Rademaker K, Blom H, et al. Systemic hydrocortisone to prevent bronchopulmonary dysplasia in preterm infants (the SToP-BPD study); a multicenter randomized placebo controlled trial. BMC Pediatr 2011;11:102. [15] ClinicalTrialsgov. A randomized controlled trial of the effect of hydrocortisone on survival without bronchopulmonary dysplasia and on neurodevelopmental outcomes at 22e26 months of age in intubated infants <30 weeks gestation age. 2011. https://clinicaltrials.gov/ct2/show/NCT01353313? term¼watterbergþANDþhydrocortisone&rank¼1].

L.W. Doyle, J.L.Y. Cheong / Seminars in Fetal & Neonatal Medicine 22 (2017) 290e295 [16] Doyle LW, Halliday HL, Ehrenkranz RA, Davis PG, Sinclair JC. Impact of postnatal systemic corticosteroids on mortality and cerebral palsy in preterm infants: effect modification by risk for chronic lung disease. Pediatrics 2005;115:655e61. [17] Doyle LW, Halliday HL, Ehrenkranz RA, Davis PG, Sinclair JC. An update on the impact of postnatal systemic corticosteroids on mortality and cerebral palsy in preterm infants: effect modification by risk of bronchopulmonary dysplasia. J Pediatr 2014;165:1258e60. [18] Laughon MM, Langer JC, Bose CL, Smith PB, Ambalavanan N, Kennedy KA, et al. Prediction of bronchopulmonary dysplasia by postnatal age in extremely premature infants. Am J Respir Crit Care Med 2011;183:1715e22. [19] Shah SS, Ohlsson A, Halliday HL, Shah VS. Inhaled versus systemic corticosteroids for preventing chronic lung disease in ventilated very low birth weight preterm neonates. Cochrane Database Syst Rev 2012;5:CD002058. [20] Shah SS, Ohlsson A, Halliday HL, Shah VS. Inhaled versus systemic corticosteroids for the treatment of chronic lung disease in ventilated very low birth

295

weight preterm infants. Cochrane Database Syst Rev 2012;5:CD002057. [21] Groneck P, Goetze-Speer B, Speer CP. Effects of inhaled beclomethasone compared to systemic dexamethasone on lung inflammation in preterm infants at risk of chronic lung disease. Pediatr Pulmonol 1999;27:383e7. [22] Halliday HL, Patterson CC, Halahakoon CW. A multicenter, randomized open study of early corticosteroid treatment (OSECT) in preterm infants with respiratory illness: comparison of early and late treatment and of dexamethasone and inhaled budesonide. Pediatrics 2001;107:232e40. [23] Suchomski SJ, Cummings JJ. A randomized trial of inhaled versus intravenous steroids in ventilator-dependent preterm infants. J Perinatol 2002;22: 196e203. [24] Rozycki HJ, Byron PR, Elliott GR, Carroll T, Gutcher GR. Randomized controlled trial of three different doses of aerosol beclomethasone versus systemic dexamethasone to promote extubation in ventilated premature infants. Pediatr Pulmonol 2003;35:375e83.