Efficacy of inhaled salbutamol with and without prednisolone for first acute rhinovirus‐induced wheezing episode

Acute rhinovirus‐induced wheezing is common in young children and may respond to systemic corticosteroid. There are no trials on the efficacy of inhaled beta2‐agonist in this clinical scenario.


| INTRODUC TI ON
Bronchiolitis affects up to 20%-30% of children during first two years of life and is the most common reason for hospitalization in children. 1 Respiratory syncytial virus and rhinovirus (RV) are the most common etiologic agents. The former is associated with typical bronchiolitis syndrome whereas RV aetiology starts to dominate at little older age, typically after 12 months of age, and is particularly interesting since it is closely linked to presence of wheezing and has been recognized as a major risk factor of atopic asthma up to the age of 13 years. [2][3][4][5][6] Interestingly, previous post hoc studies have found out that RV-affected first-time wheezing children, especially those with high virus genome loads, appear to respond to oral prednisolone Results: Median age of subjects was 13 months, 32% were sensitized and 22% had doctor-diagnosed eczema. In the duration of hospitalization, salbutamol high-dose/ placebo versus salbutamol on-demand/placebo groups did not differ (p = .12). In the occurrence of and time to relapse within 2 months, a significant group × treatment interaction was observed (both p = .02), such that high-dose group had less and longer time to relapses than on-demand group in prednisolone arm (both p < .05), but no difference was detected in placebo arm (both p > .26).

Conclusions:
In young, hospitalized children with first episode of rhinovirus-induced wheezing, high-dose inhaled salbutamol may interact with oral prednisolone.
However, further trials are warranted.

K E Y W O R D S
Asthma, beta 2 -agonist, bronchiolitis, corticosteroid, paediatrics, virus

Key Message
• We evaluated high-dose nebulized versus on-demand inhaled salbutamol in children with first rhinovirusinduced wheeze.
• There was an interaction between high-dose salbutamol and prednisolone, but hospitalization duration did not differ.
• Combined prednisolone and high-dose salbutamol for rhinovirus-induced wheeze in young children need further investigation. High-dose salbutamol n=35 On-demand salbutamol n=60 Prednisolone n=19 Placebo n=16 Prednisolone n=30 Placebo n=30

G R A P H I C A L A B S T R A C T
In young, hospitalized children with first episode of rhinovirus-induced wheezing, high-dose salbutamol may interact with oral prednisolone. However, further trials are warranted.
in terms of less recurrent wheezing and less need for asthma control medication before school age. [7][8][9] The potential explanations for the increased risk of asthma include low interferon responses (i.e., impaired viral defence), early airway inflammation (i.e., broken epithelial barrier, T helper 2 polarized immune responses), biased hostmicrobiome and genetic variation at the 17q21 and CDHR3 loci. 1,10,11 Early systemic corticosteroid treatment is likely to down-regulate early inflammatory events.  Table S1, and selection of trials are shown in Figure S1). 18 Moreover, conclusive evidence of their efficacy in subgroups of bronchiolitis remains unavailable.
Although bronchiolitis is usually discussed as a single disease, it is a heterogeneous condition, which implies that the treatment should be administered on a more personalized basis, as stated in a recent EAACI Task Force paper. 1 Since RV-induced first wheezing illness has many asthma-like features, 1,2,6,10 we hypothesized that it would respond to beta 2 -agonists. Therefore, our aim was to investigate post hoc the short-term efficacy of beta 2 -agonist with and without oral corticosteroid in the first acute RV-induced wheezing episode in children aged less than 2 years.

| Subjects
The study population came from two randomized controlled tri-

| Study protocol
The need for hospitalization was decided by an on-duty study physician independent of study.
The recruitment to the study was done by study physician. At study entry, the guardian filled in a standard questionnaire on host and environmental risk factors for asthma. Then the child was physically examined by study physician, a nasopharyngeal aspirate sample was obtained for viral diagnostics using a standardized procedure, 20 and a baseline blood sample was drawn. The children were randomized to be given either oral prednisolone or a placebo; at study entry in Vinku, and after positive RV PCR test in Vinku2. 9,19 Both studies used identical sampling and follow-up protocols including daily symptom diaries for the first two months as well as scheduled follow-up visits at 2 weeks and 2 months by the study physicians. In addition, the guardian was asked to bring the child to the study phy-

| Outcomes
Three predefined primary outcomes for this current study were: (1) time until ready for hospital discharge, (2) occurrence of a new physician-confirmed wheezing episode during the 2-month follow-up, as well as (3) time to a new physician-confirmed wheezing episode during the 2-month follow-up. 9,19 Original primary outcome of Vinku study was time until ready for hospital discharge. 19 Original primary outcomes of Vinku2 study were (1) the occurrence of and time to a new physician-confirmed wheezing episode within 2 months after discharge, (2) number of physician-confirmed wheezing episodes during the first 12 months after discharge, (3) diagnoses of asthma during the full study period. 9 The assessment of ready for discharge was based on a clinical scoring, as previously Note: Values are shown as mean (SD), median (interquartile range) or number (%).
Data were analysed by two-sample t-test, Mann-Whitney U-test, χ 2 test or Fisher exact test.
The predefined secondary outcomes for the current analysis included the duration of cough, duration of wheezing, occurrence of a new physician-confirmed wheezing episode within 2-month follow-up as inpatient, and number of bronchodilator puffs during the 2 weeks after discharge, as well as time to a new physicianconfirmed wheezing episode. Original secondary outcomes of Vinku study were (1) oxygen saturation during hospital stay, (2) wheeze and cough during 2 weeks after discharge from the hospital, (3) readmission to the outpatient clinic or hospital for recurrent wheezing during a two-month period after discharge and (4) blood eosinophil counts at discharge. 19 Original secondary outcomes of Vinku2 study included occurrence of symptoms (cough, wheeze, noisy breathing, breathlessness, rhinitis) as well as exact number of bronchodilator puffs used up to 2 months after discharge from the hospital. 9 For the current analysis, study follow-up was limited to 2 months.

| Definitions
Wheezing refers to expiratory breathing difficulty with bilateral high pitched sounds during expiration. Wheezing episodes accompanied by RV detection by PCR were called RV-induced wheezing episodes which are a subgroup of bronchiolitis. 9 Atopy was defined as positive immunoglobulin (Ig) E antibody (≥.35 kU/L) to any of the fol-

| Laboratory data
Nasopharyngeal aspirate or swab samples were stored at +4° until analysed within 3 days after collection. Nucleic acids were either extracted with a commercial nucleic acid extraction kit (High Pure Viral Nucleic Acid Kit, Roche diagnostics) or using NucliSens EasyMag automated extractor (bioMerieux). Nucleic acids were stored at −70°C if not analysed immediately.
An in-house PCR test was used to detect rhinoviruses. The primers were derived from the highly conserved 5' noncoding region of Picornavirus genome and they virtually detect all rhino-and enterovirus genotypes. 21,22 The forward primers (positive strand) used were 5′-CGGCCCCTGAATGCGGCTAA-3′, and reverse primers (negative strand) were 5′-CGCCCCCTGAATGCGGCTAA-3′.
An RT-PCR hybridization method was used in the Vinku study, as previously described in detail. 22 7,9,20,24 In addition to viruses mentioned ahead, coronaviruses (229E, NL63, OC43 and HKU1) were analysed by PCR in Vinku 2 study. 9,24 Human bocavirus (HBoV) was analysed using PCR and serology as previously described. 24,25 The methodology of these has been previously described. 7,9,19 Blood eosinophil count and serum levels of allergen-specific IgE were analysed by the routine diagnostics of the Central Laboratory of Turku University Hospital. Serum 25-hydroxyvitamin D measurements were done by liquid chromatography-tandem mass spectrometry at Massachusetts General Hospital (Boston, USA).

| Statistics
Differences in baseline characteristics between groups were ana-

F I G U R E 2
The vast majority (32/35, 91%) of post-hospitalization recurrences were confirmed at the study clinic.

| Secondary outcomes
No significant interactions were detected in secondary outcomes.
High-dose group had shorter duration of cough than on-demand group (treatment adjusted main effect of salbutamol group p < .001) ( Figure 5).
Prednisolone treatment arm had less new physician-confirmed wheezing episodes as inpatient than placebo arm (salbutamol group adjusted main effect of treatment p = .03) ( Figure 3B) ( Table 2).

| Adjusted analyses for interactions
The interaction between group and treatment on new physician-  (Table S3).

| Adverse events
No clinically significant (severe or serious) adverse events were reported.

| DISCUSS ION
Rhinovirus-induced early wheezing has been recognized a major risk factor for subsequent asthma in many studies. [4][5][6] This post hoc analysis is the first to investigate and to demonstrate the efficacy of beta 2-agonist medication in the first RV-induced early wheezing episode. In predefined primary outcomes, we were able to show that (1) high-dose regularly administered beta 2 -agonist was more effective compared to on-demand administered salbutamol in terms of less relapses within two months. (2) We also observed a significant interaction between the treatments indicating that effect of salbutamol treatment was further improved with concomi-  Most previous trials investigating the efficacy of beta 2 -agonists in children with bronchiolitis and/or first wheeze did not demonstrate clinical benefit (Table S1). These discouraging results have led most clinical guidelines do not recommend the use of beta 2 -agonists in bronchiolitis. [12][13][14][15][16][17] Only one recent study has investigated subgroups, namely comparing atopic and non-atopic children and found better clinical efficacy in atopic compared to non-atopic children. 26 This finding is in line with our study. In recent years, we have learned to 7 years. 4,[7][8][9]19 Due to study design, the current trial partly repeats these findings regarding the efficacy of prednisolone, but interestingly, results suggests that most beneficial response will be achieved with the combination of nebulized salbutamol and prednisolone. While many previous trials as, for example, in Cochrane review suggest that oral corticosteroids have no role in treatment of bronchiolitis, they fail to separate different bronchiolitis phenotypes. 28 In agreement, other RCT studies have also shown that best treatment response may be achieved by combining systemic corticosteroid with adrenergic agonists in children with bronchiolitis. 29,30 Moreover, in our data the short-term main outcome, time until discharge, supports the possibility of clinical benefit of highdose beta 2 -agonist treatment (loose analysis), the longer-term primary outcomes, relapse and time to relapse due to wheezing within 2 months, support the benefit of high-dose salbutamol as well as oral corticosteroid and the 'medium-term' secondary outcome, duration of cough within 2 weeks supports the benefit of both. In addition, in our data, a discrepancy was found when comparing different relapse types (in-and outpatient relapses). This inconsistency was found only in salbutamol on-demand/prednisolone groups, and it indicates that sole high-dose beta 2 -agonist without prednisolone might not necessarily be optimal approach, because of the increase in relapses as shown in Figure 3. Exact mechanism is not clear, but it may be due to increased bronchial hyperreactivity. 31  The strengths of the current study include careful characterization of the subjects and same detailed prospective follow-up in both original trials. The study design of combining data from two previous prednisolone intervention trials provided greater statistical power. However, the beta 2 -agonist treatment regimens differed between the two studies. There are also weaknesses. Statistical power analysis for the salbutamol intervention was not done, and rather small sample size did not permit optimal analyses in multivariable model. However, the significant interactions persisted in adjusted analyses, and results of many outcomes were in line arguing against false-positive finding. Moreover, the results may not be generalizable to outpatients, since all our subjects were enrolled in the hospital ward; the sample size was too small to permit meaningful analysis of inpatient versus outpatient interactions.
Also, because the differences in study protocols (Vinku2 study had prospective RCT design whereas Vinku study had post hoc design) duration of hospitalization was not fully comparable in regards with prednisolone treatment (B2 on-demand/prednisolone group from Vinku2 study) since prednisolone was administered only after positive RV PCR-finding in Vinku2 study which caused a delay in drug administration compared to other study groups. In other outcomes, the 45 h time delay in the administration of prednisolone is statistically taken to into account by the group effect (study effect). One can question whether these were truly the first wheezy episodes.
To address this issue, we checked medical records and interviewed a parent using a standard questionnaire to confirm that there was no previous wheezing episode. Although some patients received additional medication after discharge during 2-month follow-up period, no differences were found between the use of salbutamol or antibiotics. Use of inhaled and oral corticosteroids seemed to be linked to relapse outcomes (data are not shown).
Although RVs are the most commonly detected respiratory viruses during symptomatic infections, they are also the most commonly found respiratory viruses in asymptomatic children. However, previous studies have found that RV reinfections are in nearly all cases caused by different RV genotype, and that time period for a new RV infection is relative short. 34 Due to this finding we can assume positive RT-PCR sample shows an incipient, ongoing or passing RV infection which can be symptomatic or asymptomatic.
In summary, while previous studies have shown that beta 2agonists are not efficacious for the treatment of bronchiolitis, our data suggest that high-dose salbutamol and prednisolone may be beneficial for a subgroup of young children with an initial severe wheezing episode caused by RV. It is probable that some of these children have early asthma-like airway inflammation, which may explain the observed efficacy. Our data support the use of a short course of systemic corticosteroid and at least a therapeutic trial with high-dose short-acting beta 2 -agonists in these selected, RV-induced cases of bronchiolitis. However, our results are partly inconsistent and should be considered as hypothesis-generating. Prospective clinical trials are warranted. More broadly speaking, bronchiolitis is a heterogeneous condition 1,2,10 and our data suggest that its treatment should be administered on a more personalized basis than is recommended by current clinical guidelines.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest in connection with this paper.

AUTH O R CO NTR I B UTI O N S
The study protocol and manuscript were written by the investiga-

DATA AVA I L A B I L I T Y S TAT E M E N T
Research data are not shared.