It has long been suspected that the inflammatory substrate underlying attacks of wheeze in preschool children is different from the substrate underlying attacks of atopic asthma in school-age children, even though viral colds are a common trigger. There is now a substantial body of evidence from epidemiological and airway sampling studies to support this view. The first convincing evidence that wheeze in preschool children is a different ‘phenotype’ of asthma was provided by the Tucson (USA) birth cohort study,1 which found that wheeze for many preschool children is not associated with atopy and that it resolves by school age. A recent Dutch birth cohort study of 3963 children supports these data, since only 178 (14%) of 1230 children with parent-reported preschool wheeze had a diagnosis of asthma at 7–8 years of age.2 Furthermore, in the few preschool children who have undergone invasive airway sampling, the typical pattern of chronic eosinophilic airway inflammation that is characteristic of atopic asthma has not been found.3 4

To date, few large trials have specifically targeted preschool wheeze, and clinicians have used therapies of proven efficacy in older asthmatic children, including a short course of oral steroids.5 Two recent randomised double-blind placebo-controlled trials (RCTs) of oral steroids in preschool wheeze performed by my research group, provide some evidence to inform practice. In the first study, preschool children (1–5 years of age) presenting to a hospital accident and emergency department were randomised to receive either a short course of oral steroids (20 mg once a day for 5 days) or placebo to be started by the parents at the first sign of their next wheeze attack.6 One hundred and twenty children had a further attack of wheeze at home and were given the trial medication. No difference was found in the 7-day parental-assessed respiratory symptom score (the primary outcome) between steroid and placebo treated children.6 In the second trial, 700 children presenting to hospital with an attack of preschool wheeze were randomised to receive to a 5-day course of oral prednisolone or placebo.7 The primary outcome was the duration of hospitalisation. Symptoms were also assessed using a validated symptom score, the Preschool Respiratory Assessment Measure (PRAM).7 No significant difference was found between the treatment groups for the duration of hospitalisation, or for the PRAM over the first 24 h.7 Furthermore, after discharge there was no significant difference in the parental-assessed 7-day symptom score, or use of inhaled bronchodilators.7 Taken together, these data suggest that a short course of oral steroids should not be given for attacks of preschool wheeze, but there are caveats. First, two other RCTs in acutely wheezy preschool children have reported a beneficial, albeit inconsistent, effect of oral steroids. Csonka et al8 in 230 children less than 3 years of age presenting to hospital with acute wheeze, reported a beneficial effect of oral steroid (2 mg per kg once a day) on the “need for additional medication” (the primary outcome), but no effect on discharge from hospital at 4 h. In 70 children aged 7–54 months, Tal et al9 reported that more children receiving intramuscular methylprednisolone were discharged at 3 h than those in the placebo group. Some benefit of prednisolone on duration of hospitalisation has also been reported in a subgroup of preschool children with wheeze triggered by specific viruses. However, these data are a result of post hoc secondary analyses of a subgroup of children 3 months to 16 years of age, where no overall beneficial effect of oral prednisone was found for the primary outcome.10,–,12 Although meta-analysis of these studies is needed, it is unlikely that a beneficial effect of systemic steroids will emerge in a pooled analysis. Another caveat is that there may be a small proportion of preschool children whose attacks of wheeze occur on a background of chronic, steroid-responsive, airway inflammation. In a highly selected group of 13 preschool children referred to the Royal Brompton Hospital with symptoms severe enough to justify bronchial biopsy (ie, severe recurrent doctor-diagnosed wheeze with at least three severe episodes lasting more than 3 days in the previous 6 months), Saglani et al13 found that the number of airway tissue eosinophils was above the upper range of controls in six children. It is therefore possible that a beneficial effect of oral steroids in a small subgroup with chronic airway eosinophilia is masked by the lack of response in the majority of children. In our recent RCTs we did attempt to examine response in children at higher risk of eosinophil activation. Prespecified subgroup analysis was carried out in children with either increased levels of eosinophil cationic protein6 or at increased risk of school-age asthma by history of four or more wheezing episodes and a parent with asthma, or physician-diagnosed eczema.7 No beneficial effect of oral prednisone in these putative ‘steroid-responsive’ subgroups was found.6 7 It remains unknown whether symptomatic preschool children who will develop atopic asthma based on preschool symptom pattern in later life are more responsive to oral steroids. The division of preschool wheeze into the ‘multi-trigger’ (wheeze with and between colds) and ‘episodic (viral)’ phenotypes by a recent European Respiratory Society Taskforce,14 is a commendable first attempt to tailor therapy to preschool wheeze phenotype. However, a previous attempt to predict the development of atopic asthma was inaccurate.15 In a retrospective analysis using a combination of eight clinical parameters from the Dutch cohort data, 34/2171 symptomatic children had a score of more than 35, a cut-off that is highly specific (99%) but is not sensitive (7%) for subsequent school-age asthma.2 The small number of preschool children indentified by this score means that to recruit 700 high-risk children into an RCT would require screening 25 000 symptomatic children. A final caveat is that our negative RCTs of oral steroids cannot be extrapolated to very high doses of aerosolised steroids. In preschool children with at least two episodes of wheeze, Ducharme et al16 assessed the efficacy of 3 mg inhaled beclometasone equivalent per day started at the onset of each cold and continued for up to 10 days for each episode over a 6–12-month period (n=129). Overall, 8% of viral colds led to ‘rescue’ oral steroids in the inhaled steroid group (used as a marker of severe wheeze) compared with 18% in the placebo group, but this was achieved at the expense of reduced. It is unclear why a beneficial effect of inhaled steroids was found in this RCT in a population that is not at high risk of developing atopic asthma.

Are there any other therapeutic options? Cysteinyl leukotriene activation occurs during attacks of preschool wheeze,17 and in an RCT of daily oral montelukast, Bisgaard et al18 found that montelukast reduces the rate of wheeze attacks by 31.9% compared with placebo. Outside a trial, it is debatable whether parents will continue to give oral montelukast in the long asymptomatic periods between attacks. An alternative is to ask parents to start oral montelukast at the first sign of a cold, then stop the medication after 10 days. This has shown promise in one RCT that recruited both preschool- and school-age children,19 but further data are needed before it can be recommended. A large trial (n=1300) funded in July 2009 by the National Institute of Health Research Efficacy and Mechanism Evaluation Programme,20 will hopefully address this question. Bearing in mind the potential long-term consequences on bone mineral density of repeated courses of oral steroids,21 my recommendation is that short burst therapy must not be given in the community for attacks of preschool viral wheeze. If short burst oral steroid therapy is to be given at all, it should only be considered in the minority of preschool children admitted to a hospital ward with clinical features strongly suggestive of atopic asthma (eg, a combination of multi-trigger wheeze, severe eczema, and a family history of atopic asthma that started in the preschool years), or with very severe bronchodilator-unresponsive wheeze.