Dexamethasone for BPD

Doyle LW, Ehrenkranz RA, Halliday HL. Dexamethasone treatment after the first week of life for bronchopulmonary dysplasia in preterm infants: a systematic review. Neonatology 2010;98(4):289-296. Pubmed, Free full text article

Background: Dexamethasone has powerful anti-inflammatory effects and has been used to treat established bronchopulmonary dysplasia (BPD), but it is uncertain whether the benefits outweigh the risks of treatment.

Objectives: To determine the effect of late (>7 days) postnatal dexamethasone treatment compared with control (placebo or nothing) to prevent or treat BPD in the preterm infant.

Methods: Randomised controlled trials (RCTs) of late postnatal dexamethasone therapy to treat or prevent BPD were sought using methods of the Cochrane Collaboration. Data regarding clinical outcomes including mortality, BPD, death or BPD, complications during the primary hospitalisation, and long-term outcome were abstracted and analysed using RevMan 5. Results: 19 RCTs enrolling 1,345 participants were eligible for this review. Late dexamethasone treatment reduced neonatal mortality, but not later mortality. Benefits of late dexamethasone included reductions in failure to extubate, BPD and the combined outcome of death or BPD. There were clear short-term complications, including hyperglycaemia and hypertension, but not intestinal perforation. Trends of an increase in cerebral palsy or abnormal neurological examination were partly offset by a trend in the opposite direction in death before late follow-up.

Conclusions: The benefits of late dexamethasone may not outweigh actual or potential adverse effects. Given the evidence of both benefits and harms of treatment, and the limitations of the evidence at present, it appears prudent to reserve the use of late dexamethasone to infants who cannot be weaned from mechanical ventilation, and to minimise the dose and duration of any course of treatment.

Comments: Although a very effective treatment to improve the pulmonary function of preterm infants with evolving or established BPD, use of dexamethasone has become constrained as a result of concerns regarding impact on long-term neurodevelopment. A statement issued by the AAP and CPS in 2002 warned against using dexamethasone and stated that, "Outside the context of a randomized trial, the use of corticosteroids should be limited to exceptional clinical circumstances (eg. an infant on maximal ventilatory and oxygen support). In those circumstances, parents should be fully informed about the known short and long term risks and agree to treatment." To many, this statement was excessively harsh, especially with the recognition that the studies showing increased risk of long term sequelae used either very early treatment and/or very high starting doses with long duration of treatment. Nevertheless, the statement had the desired effect of reducing the percentage of preterm infants exposed to dexamethasone with only a marginal increase in BPD, perhaps because of a concomitant improvement in respiratory support practices.

This report outlines the highlights of a recent systematic review looking at the impact of dexamethasone for BPD when treatment begins after the first week of life. The full report is available as part of the Cochrane Collaboration (link here). By eliminating the studies using dexamethasone for prevention of BPD (ie. starting in the first day or 2 of life) and adding newer trials, the balance seems to shift more in favour of there being an overall beneficial effect of dexamethasone treatment. There are still trends suggesting that there might be an increased risk of neurodevelopment problems with dex treatment but there is certainly no definitive evidence from these studies that dex treatment causes CP or cognitive problems. There was a wide variation in dosage and duration of treatment in these trials and a suggestion that regimens with lower cumulative doses may be just as effective with less risk of impacting neurodevelopment. I remember Roberta Ballard putting up a slide at Hot Topics when the issues about long-term outcome were first raised - the slide said simply, "It's the dose, stupid".

The authors of this systematic review continue to be cautious since most studies did not have a pure placebo group (ie. many babies in the placebo groups received open label dex sometime later on in their hospital course) and most were not powered to show differences in long-term outcome. I believe that the recommendations are reasonable and suggest that we can "loosen the reigns" on dexamethasone somewhat. I have been personally concerned that the extreme restrictions on the use of dex that followed the AAP/CPS statement have disadvantaged many infants as we have had to wait until they are on "maximal ventilatory and oxygen support" lest we make ourselves medicolegally vulnerable.

I would be interested in what others have been doing and whether practice is shifting towards somewhat more liberal postnatal steroid usage?

What do you tell parents about the possible risk to their baby if you think he/she should be considered for dex treatment?

Thanks Mike. This is a great topic, and I hope it generates some thoughtful discussion. We have pretty much abandoned the use of dexamethasone for these patients for fear of liong-term neurodevelopmental sequelae. Instead, for those preemies who get "stuck" on the ventilator we sometimes resort ot trying low-dose hydrocortisone treatment (2-5 mg/kg/dose, 2-3x/day for 7-10 days). While this usually works, I'm still concerned about the possible long-term sequellae. I'd really like to hear the opinoins of those in our Respiratory Care focus group. What do you recommend in terms of steroid treatment for premature infants who are stuck on a ventilator and not weaning?

Andy Kairalla MD
Editor

I find the conclusions of this new meta-analysis to be a bit more aimed to how steroids are being used. I wrote a comment after the 2002 statement [See 2-056 for Dr Burchfield's earlier comment. ED] concerning the most likely survival benefit of steroids in the sick neonate, how many patients were randomized who were not that sick, and how we should not abdicate our responsibilities as physicians to make tough decisions. I believe in and practice keeping parents abreast of the clinical situation and treatment plans. I do not feel this is a decision that requires parental consent. If a parent chooses one of the two options and the baby does poorly, we have assigned them a level of remorse that should have never been given to them. In his comment, Dr Dunn mentions a lack of purity in the control groups in the studies in that many of these patients received steroids off-study. This cannot be minimized. Prior to this meta-analysis, 40% of the control patients received dexamethasone. Why? Most likely, the clinician felt that the patient was so ill that he/she was not going to allow the baby to continue to deteriorate without knowing the patient received steroids. How many of these patients survived due to this decision, and if those patients had not received steroids and died, how much sooner would we have see a survival benefit in the treatment group? The concern for side effects is real, although transient hyperglycemia and hypertension does not is a small trade-off for survival. The neurological side effects remain a concern, but appear to be more in question in this population. I am not advocating a return to the practices of the 1990's, but am glad that this softer statement for use of steroids for infants with severe respiratory disease consistent with BPD is in place.

UserName: David J Burchfield, MD
Institution: University of Florida
telephone: 352-273-8985
email: burchdj@peds.ufl.edu