Steroids and Pre-school wheeze – the saga continues

There has been no class of drug more scrutinised over the last few years that the humble corticosteroid. ADRENAL said it wasn’t much use in adult septic shock. RECOVERY provided it’s redemption arc during COVID-19, and dexamethasone was the talk of the town once more. But, in paediatrics, we have been trying for a number of years to answer a specific steroid question – is it of any use in pre-school wheeze?

Pre-school wheeze has always appeared to be the odd man out of wheezing in children. The phenotypic models for asthma and bronchiolitis are relatively clear, but in the middle stands pre-school wheeze, a heterogeneous melange of several different factors lending themselves to a wheezy phenotype, that either exits like nothing has happened, or develops into an early asthmatic picture. With heterogenicity comes vast differences in practice – giving everyone steroids, only giving the severe cases steroids, only treating those with atopy with steroids, the list goes on. Thankfully, some clever research-type folks decided that this is fertile land for studying, and the latest Archives of Disease in Childhood paper from New Zealand aims to shed some light on the subject.

What do we know about this currently?

Prior to this publication, two papers formed the backbone of evidence to steroids in pre-school wheeze – Panickar et al in the NEJM in 2009, and Foster et al in The Lancet Respiratory Medicine in 2018. The original Panickar trial compared 5 days of prednisolone to placebo in wheezy children aged 10 months to 5 years, and concluded that prednisolone was not superior to placebo in treating pre-school wheeze. The critics focused on the lower age of included patients, as 10 month and up would have included a reasonable number of bronchiolitic patients that would not be steroid responsive, possibly biasing the results. None-the-less, Panickar’s study attacked our assumptions that pre-school wheeze should be treated blanketly with steroids, and steroid use swung out of favour in these patients. Then in 2018, Foster et al threw their hat in the cyclohexane ring with the PREVIEW trial. This single-centre study from a paediatric emergency department in Western Australia compared prednisolone to placebo in children aged 2-6. They concluded that steroids were superior to placebo in reducing length of stay slightly, and that this finding was independent of their pulmonary score at presentation. Interestingly, previous personal or family history of atopy, commonly cited as a reason to give steroids in the past, did not affect the outcomes of this study, suggesting that emphasis on this item of history has been overstated. Critics here outlined that the reduction in median LOS was just 2.4hrs – statistically, but perhaps not clinically as relevant, and that the pulmonary score used in the study is not validated in under 5s, which made up the majority of the study. Nonetheless, the PREVIEW trial cast some doubt on the previous evidence base that steroids were of no utility in pre-school wheeze, and so the wheel kept turning.

Tell me about this new ADC paper then?

Hot off the presses in this month’s ADC is a study from from New Zealand. They set out to cast some more light on this pertinent clinical question:

In patients aged 2-6 with respiratory symptoms and wheeze, is prednisolone equivalent to placebo in reducing severity of symptoms at 24 hours after intervention?

They set out to answer this by designing a multicentre, double-blind randomised, placebo controlled equivalence trial. They enrolled children between 24 months and 59 months (2-5 years) presenting to hospital with a wheeze associated with a respiratory illness.

Children excluded were those with wheeze related to non-respiratory causes, very mild initial disease (PRAM score <3), chronic cardiac conditions, chronic respiratory conditions (N.B. children with previous asthma or recurrent wheeze were not excluded), history of inhaled foreign body, current or past history of life-threatening asthma, contraindication to steroids, previous study enrolement, those that had taken corticosteroids in the past 7 days, and those unable to be followed up between 18-38 hours.

Once eligible, they received a standard therapy – in this case, 6 puffs of salbutamol via spacer every 20 mins for 1 hour, then bronchodilators as required from then on. They were approached for consent, and if they agreed, they were randomised to receive either 2mg/kg prednisolone once daily for three days, or a matched placebo.

Primary outcome for this study was reduction in PRAM score at 24 hours after administration of medication (for more about the PRAM score, see this link). Secondary outcomes were PRAM score at 4 hours, admission rate, length of ED and inpatient stay, amount of salbutamol given by 48 hours and 7 days, treatment with additional open-label prednisolone, time to return to normal activity, and adverse events, including use of IV medications, PICU admissions and representation to ED/primary care with respiratory illness within 7 days.

So what did they find?

Over the two year study period, they randomised 493 patients to the study. After withdrawal and post randomisation exclusions, 477 were analysed, 238 in the prednisolone arm, and 239 in the placebo arm. Primary outcome data was available for 393 of these patients, 195 from prednisolone arm and 198 from the placebo arm. There were no significant differences in changes in PRAM score between the two arms, and on subgroup analysis, there was no significant difference between those considered “salbutamol responsive”, those with a positive Asthma Predictive Index or based on the baseline PRAM score – i.e. the lack of primary outcome difference between arms was maintained even when looking at those presenting with more severe symptoms.

However, there were differences in secondary outcomes. Patients in the prednisolone arm were more likely to have a PRAM score of 0 at 24 hours compared to placebo. They were also less likely to receive further open-label steroids (1.3% vs 5.4%), and less likely to receive intravenous medications for wheeze (1.3% vs 4.6%). No other outcomes of significance were noted.

So what does it all mean?

After finishing the paper, we usually ask ourselves three questions about it:

Is the clinical question relevant to my practice?

Overwhelmingly, the answer is yes. Pre-school wheeze in non-pandemic years is a significant source of admission to acute paediatric services, and there is significant variability of practice with regards to who gets steroids, and when.

The primary outcome here was change in PRAM score between attendance and 24 hours. The subject of scoring systems in asthma can be divisive. Having an objective measurement of severity of wheeze is excellent for research, lowering the chances of bias, and reducing interpersonal variability in reporting wheeze improvement. Thing is, I don’t use a wheeze score in assessing improvement – and very few of us do, so, whilst the primary outcome is obviously of interest to me, I’m also pretty interested in the secondary outcomes here as well.

Is the benefit big enough?

The trial demonstrated that prednisolone was equivalent to placebo in reduction of baseline PRAM score at 24 hours after admission. However, there is more than meets the eye in this study. The vast majority of children presenting with acute wheeze in this study were better by 24 hours, with a median PRAM score of 0. So, given that data, measuring a primary outcome difference at 24 hours is likely not hugely helpful, as most kids are better by then, whether you’d given them 2kg of dexamethasone, or tickled their feet.

However, the secondary outcomes here are worth reading. Children given early prednisolone were less likely to receive intravenous medication compared to placebo (1.3% vs 4.6%), less likely to be admitted to hospital compared to placebo (23.5% vs 31.4%), and less likely to receive additional steroids at a later date (1.3% vs 5.4%). So, there are clearly benefits to giving steroids to some children with pre-school wheeze.

The difficulty is knowing which children benefit. Subgroup analysis showed no difference in primary outcome when selecting for those that were salbutamol responsive, those with a positive Asthma Predictive Index or those with more severe disease, which suggests that giving steroids to the more “asthma-y” appearing children, and those who come in very tight is seemingly not an evidence-based way to select who receives steroids and who doesn’t. This can be added to Foster et al’s trial data, which suggested that there was no difference in outcomes when selecting for children with an atopic history, another popular reason to give steroids, so we are a little in the dark on this one.

Where does this fit in the context of other research?

Again, this is a well designed study to allow meta-analysis and systematic review. It’s inclusion criteria are broadly similar to Foster et al’s trial, allowing further meta-interpretation of the data. It seems now that Panickar et al’s broad rejection of steroids in pre-school wheeze was perhaps a little hasty, aided by the lower age inclusion. However, this is not a resounding recommendation for blanket prescription of steroids to all wheezers either. This study confirms what we suspected from Foster et al’s study – amongst the heterogenous phenotypes that make up pre-school wheeze, there is a subgroup of children that will benefit from steroids, in a reduction in admission to hospital, and reduction in escalation of care. Whilst those effect sizes are modest, given the cost of steroids, harm profile, and harm/cost of admission and intravenous medication, this is a significant finding, and should prompt further research into what factors are most likely to affect a positive response to steroids.

It is important to note that this study used a dose of prednisolone of 2mg/kg OD for three days, which is twice as much as Foster et al used in their study.

Interestingly, the finding that most children with pre-school wheeze are better by 24 hours appears to be novel. The authors conclude that this supports the notion that 3 days of prednisolone might be overtreating the problem, and doses subsequent to the initial dose provide little additional therapeutic benefit. There is support for this in the literature, with a recent comparison of single-dose dexamethasone showing non-inferiority when compared to prednisolone.

Clinical bottom line

Pre-school wheeze is a heterogenous mixture of lots of different wheezy phenotypes. Within that bundle, there are a group of children that respond to steroids, and this results in a slightly lower admission rate, and slightly lower rates of intravenous medication. The knowledge that most children are better at 24hrs regardless of treatment likely adds power to the argument for single dose dexamethasone compared to 3 days of prednisolone, or at very least reducing the length of prednisolone treatment.

Absolutely selecting which children to give steroids to in this age group remains very challenging, as none of the traditional subgroups seem to show differences when selected for in either this study, nor Foster et al’s previous study, so the suggestion that all children with an atopic history should receive steroids, a commonly cited discriminator, is not supported in either of the recent studies.

Ultimately, this study underlines one thing: our diagnostic system for infants and children with wheeze remains poor, and further work into outlining the phenotypes that make up “the wheezy infant” are vital for optimising management of this common condition.

#PaperInAPic