Moderately early (7-14 days) postnatal corticosteroids for preventing chronic lung disease in preterm infants.

BACKGROUND: Corticosteroids have been used late in the neonatal period to treat chronic lung disease (CLD) in preterm babies, and early to try to prevent it. CLD is likely to be the result of persisting inflammation in the lung and the use of powerful anti-inflammatory drugs like dexamethasone has some rationale. Early use tends to be associated with increased adverse effects so that studies of moderately early treatment (7-14 days postnatal) might have the dual benefits of fewer side effects and onset of action before chronic inflammation is established. OBJECTIVES: To determine if moderately early (7-14 days) postnatal corticosteroid treatment vs control (placebo or nothing) is of benefit in the prevention and/or treatment of early chronic lung disease in the preterm infant. SEARCH STRATEGY: Randomised controlled trials of postnatal corticosteroid therapy were sought from the Oxford Database of Perinatal Trials, Cochrane Database of Controlled Trials, MEDLINE (1966 - October 2002), hand searching paediatric and perinatal journals, examining previous review articles and information received from practicing neonatologists. Authors of all studies were contacted, where possible, to confirm details of reported follow-up studies, or to obtain any information about long-term follow-up where none had been reported. SELECTION CRITERIA: Randomised controlled trials of postnatal corticosteroid treatment from 7-14 days of birth in high risk preterm infants were selected for this review. DATA COLLECTION AND ANALYSIS: Data regarding clinical outcomes including mortality, CLD (including late rescue with corticosteroids, or need for home oxygen therapy), death or CLD, failure to extubate, complications during the primary hospitalisation (including infection, hyperglycaemia, hypertension, hypertrophic cardiomyopathy, pneumothorax, severe intraventricular haemorrhage (IVH), necrotizing enterocolitis (NEC), gastrointestinal bleeding, and severe retinopathy of prematurity (ROP)), and long term outcome (including blindness, deafness, cerebral palsy and major neurosensory disability), were abstracted and analysed using RevMan 4.1. MAIN RESULTS: Seven studies enrolling a total of 669 participants were eligible for inclusion in this review. Moderately early steroid treatment (vs placebo or nothing) reduced mortality by 28 days, chronic lung disease at 28 days and 36 weeks, and death or chronic lung disease at 28 days or 36 weeks. Earlier extubation was facilitated. There was no significant effect on the rates of pneumothorax, severe ROP, or NEC. Adverse effects included hypertension, hyperglycaemia, gastrointestinal bleeding, hypertrophic cardiomyopathy and infection. Steroid-treated infants were less likely to need late rescue with dexamethasone. There were limited data from four studies of long term follow-up; these did not show evidence of an increase in adverse neurological outcomes. REVIEWER'S CONCLUSIONS: Moderately early corticosteroid therapy (started at 7-14 days) reduces neonatal mortality and CLD, but at the cost of important short term adverse effects. Limited evidence concerning long term effects is provided by the trials included in this review. The methodological quality of the studies determining the long-term outcome is limited in some cases, the children have been assessed predominantly before school age, and no study has been sufficiently powered to detect important adverse long-term neurosensory outcomes. Therefore, given the risk:benefit ratio of short-term effects and the limited long-term follow-up data, it seems appropriate to reserve moderately early corticosteroid treatment to infants who cannot be weaned from mechanical ventilation and to minimise the dose and duration of any course of therapy. More research is urgently needed, including long term follow-up of survivors included in previous and any future trials, before the benefits and risks of postnatal steroid treatment, including initiation at 7-14 days, can be reliably assessed (See DART study; Doyle 2000a).