Futuristic treatment approaches: beyond allergy and the
eosinophil
Increasingly, attention is turning to the role of chronic bacterial and
viral infection in preschool wheeze. In a study of 35 severe preschool
wheezers who underwent bronchoscopy and bronchoapveolar lavage (BAL) at
a time of clinical stability [55], 60% had a positive bacterial
culture or viral detection, and 26% had both. Unsupervised analysis
revealed two subgroups. One was positive for Moraxella
catarrhalis with marked BAL neutrophilia, the second was a mixed
microbiota picture. Although there was a tendency for EVW patients to be
in the Moraxella group, in general there was very poor agreement
between symptom patterns and BAL findings.
We also performed a larger analysis of 136 children aged 1-5 years, of
whom 105 had recurrent severe wheeze-RSW and 31 had non-wheeze
respiratory disorders (the best control group we could find, since
normal children cannot ethically undergo bronchoscopy [56]. We
measured peripheral blood leukocyte counts, and specific IgE to common
inhalant and food allergens. We defined allergic sensitization as
allergen-specific IgE ≥0.35 kUA/L to at least one allergen tested. All
children underwent a clinically indicated bronchoscopy, BAL, and
endobronchial biopsy. Bacterial culture, multiplex PCR to 20 viruses andMycoplasma were performed on BAL. Data were analysed by the
Partition Around Medoids algorithm coupled with Gower’s distance for
mixed data. Clinically, 30/105 of the severe wheeze patients had EVW,
and 44/105 as MTW; 28 patients could not be classified as either, again
underscoring the weakness of clinical phenotyping. Eight variables were
used to determine the clusters, namely blood and BAL neutrophil and
eosinophil counts, atopy, whether viral PCR and bacterial culture were
positive, and whether ICS had been prescribed (it was considered
unethical to stop treatment in these very fragile patients). We
identified four clusters on 134/136 children, which ore no relationship
to symptom pattern. All patients in cluster 1 were sensitized; they had
the highest blood eosinophils (mean=5.54%, SD=2.86%), the highest rate
of ICS use (91.7%), and moderate rates of bacterial culture positivity
(69.5%, especially Moraxella ) and viral detection (56.5%).
Cluster 2 was characterised by low BAL neutrophils (mean=9.44%,
SD=13.89%), and a low rate of positive bacteriology (17.1%) ad viral
detection (15.0%). All were prescribed ICS. In cluster 3 there was the
highest rate of positive bacterial cultures (H. Influenzae, Staph
Aureus, Strep. pneumoniae ) ad viral infection (96.8% & 86.7%,
respectively), and the highest level of bAL neutrophils (mean=31.7%,
SD=25.11%); 67.7% were prescribed ICS. Finally, no-one in cluster 4
was prescribed ICS, and most were non-atopic with persistent cough not
wheeze.
A number of things need to be considered when interpreting this first
preschool wheeze cluster analysis. This is a highly selected group of
children with really severe wheeze who have failed to respond to
therapy. There needs to be another validation cohort. We could not
ethically stop treatment. We do not know how well the families were
adherent to treatment or how much of the prescribed dose was actually
deposited in the lower airway. Hence the effect of any prior ICS
prescription on pathology, especially airway eosinophilia, cannot be
determined. We also do not know stability of phenotypes over time.
However, what this study does do is to turn the spotlight firmly on
infection, in at least some children. The relationship between disease
and infection is unclear. One hypothesis is that chronic infection
causes wheeze; another that infection is merely a marker of underlying
topical immunosuppression which is the underlying cause of wheeze. It is
also possible to hypothesise that ICS may be causing topical
immunosuppression and thus allowing infection to become chronic.
This study points to possible cluster-based treatments (Table 7). It
must be stressed that this is speculative, and the approach needs to be
subjected to testing with randomised controlled trials before it can be
recommended. However it is hoped that considering this will broaden the
reader’s perspective on the aetiology of preschool wheeze.