Correspondence:
Manali Mukherjee MSc, PhD
Firestone Institute of Respiratory Health
St Joseph’s Healthcare Hamilton
L-314, 50 Charlton Avenue East
Hamilton, Ontario,
L8N 4A6, Canada
Tel. +1 905 522 1155 x37313
Email:
mukherj@mcmaster.ca
Key words: airway microbiome, IL-13, neutrophils, T2
inflammation, severe asthma
Conflict of interest: Dr. Mukherjee is supported by
investigator award from Canadian Institutes of Health Research and
Canadian Allergy, Asthma, and Immunology Foundation. She has received
grants from Methapharm Speciality Pharmaceuticals and honorarium from
AstraZeneca and GlaxoSmithKline outside the submitted manuscript. Dr.
Agache is Associate Editor of Allergy.
Word limit: 1011/1000 words
References: 9/9
To the editor,
With great interest we read the article by Azim and co-workers [1]
published in this issue of Allergy. To understand the inflammatory
component of the peripheral airways of severe asthmatics uncontrolled by
high dose inhaled corticosteroids (ICS) the authors examined the
bronchoalveolar lavage (BAL) in two well-characterised asthma
populations. In 78 severe asthma patients constituting the experimental
cohort, an increase in IL-13 BAL levels associated with increased
neutrophils without eosinophils, were reported. In the validation cohort
(n=18), the authors confirmed this “non-eosinophilic” phenotype and
further showed the presence of pathogenic bacteria, Moraxella
catarrhalis , Haemophilus sp and Streptococcus sp . The
significance of the study is three-fold. First, the identification of a
severe T2 asthma population with an IL-13 signature and airway
neutrophils that may be misclassified due to absence of the “classic”
biomarker – eosinophils. Second, it highlights potential alternate
cellular sources of IL-13 besides the T2 cells. Third, the association
of high BAL IL-13 with microbial dysbiosis in a low-eosinophilic airway
inflammation setting that indicates a novel pathogenetic role of IL-13,
further contributing to asthma severity.
Almost two decades back Ward et al reported a high variability in the
inflammatory BAL profile [2]. Therefore, the confirmation of a
neutrophilic airway cellularity in a follow-up validation cohort was
essential, and adds to the robustness of the study design. However, the
presence of neutrophils in the BAL may simply be an innate immune
response to the detected pathogenic bacteria or the ongoing effect of
the corticosteroid therapy, rather than a direct IL-13 effect. Both
confounders could have been easily addressed if the airway inflammation
was investigated longitudinally after a course of antibiotics. In
addition, the absence of eosinophils in the BAL could simply be a
numerical anomaly where the increased number of neutrophils masks the
eosinophil count. Again, this could have been checked by examining the
cellularity in BAL post-infection (i.e. after a course of antibiotics)
or by using a ratio between eosinophils and neutrophils instead of
absolute values.
Close inspection of the reported data reveals that an underlying
eosinophilic/T2 component in the high tertile IL-13 severe asthma group
cannot be ruled out altogether. First, the distribution on the scatter
plot (Figure 2 in Azim et al [1]) shows half of the patients
having >2% BAL eosinophils. In the replication cohort, the
BAL eosinophil% were higher in the high IL-13 group vs low (1.30±4.53
vs. 0.50±1.65) but multiple comparison tests have not been reported to
ratify a statistical difference. IL-5 levels were significantly higher
in the high IL-13 compared to the low tertile group (0.61 ± 1.53 vs 0.32
±0.35 pg/mL, p=0.001). IL-13 correlated with both BAL neutrophil%
(r=0.580, p<0.001) and BAL eosinophil%, (r=0.271, p=0.017),
even though the latter was a weaker association. Using routine sputum
cell differential, ~10% of severe asthmatics
consistently have a mixed granulocytic sputum, i.e. neutrophils
>65% together with borderline eosinophils in the range of
2-3%. These patients usually have ongoing airway infection that masks
their eosinophilia, as their eosinophilia re-emerges after the infection
is resolved [3]. A mixed granulocytic phenotype based on sputum
cytology has been reported both by the Airways Disease Endotyping for
Personalized Therapeutics (ADEPT) and Unbiased BIOmarkers in PREDiction
of respiratory disease outcome (U-BIOPRED) cohorts. The patient
population evaluated is similar to other observational studies where
7-10% of asthmatics have mixed phenotype [3], lowest lung function
[4], and similar dysbiosis by 16s sequencing [5]. Taken
together, the high tertile IL-13 “neutrophilic” group based on the BAL
analysis may be representative of the mixed granulocytic phenotype.
The debate continues on whether neutrophils contribute to the pathology
of severe asthma in the high tertile IL-13 group or whether they are
only present as a consequence of infection or the use of
corticosteroids. It must be pointed out that, targeting different
drivers of neutrophilic inflammation (e.g. IL-17, IL-23, CXCR2
[reviewed extensively in [6]]) had discouraging results in the
clinical trials, although patient selection was not optimal.
We next raise the question on whether increased IL-13 in the airway is
contributory or collateral to the observed infections. The IL-4/IL-13
pathway has been implicated in dampening the neutrophil response to an
infection by impairing IL-8 induced migration and neutrophil
extracellular trap formation [7]. LPS (infection mimic) induces
airway hyperresponsiveness and corticosteroid resistance via IL-13
dependent signalling of pulmonary macrophages [8]. Therefore,
increased IL-13 due to an underlying T2 pathology may impair host
defence. Of interest, increased sputum IL-13 levels predicts the
presence of airway autoantibodies in severe asthmatics with a mixed
granulocytic phenotype. Moreover, autoantibody-induced macrophage
dysfunction in the mixed phenotype can contribute to recurrent airway
infections [9]. That said, inherent immunodeficiencies or neutrophil
dysfunction were not ruled out in this study as other underlying reasons
for the observed airway infection in the replication cohort.
In summary, elevated levels of IL-13 in the peripheral airways with
increased neutrophils and low levels of eosinophils reveal an underlying
T2 pathway for the mixed granulocytic asthma phenotype, together with
alternate cellular sources of IL-13 besides eosinophils or the T2 cells,
such as alveolar macrophages or mast cells (Figure 1). The study also
adds a novel potential pathogenetic pathway (innate immune response
impairment with microbial dysbiosis) to IL-13’s “classic” contribution
to asthma severity, including but not limited to, airway smooth muscle
phenotype switch, mucus hypersecretion, eosinophil recruitment, and
epithelial activation (Figure 1). Pending further validation through
longitudinal studies, a therapeutic focus for this particular severe
asthma endotype is therefore IL-13, a cytokine that is truly pleiotropic
for its heterogeneity of cellular source and downstream functions.
Targeting the IL-13/IL-4 pathway using an anti-IL-4Ra monoclonal
antibody (dupilumab) was more successful than targeting IL-13 alone
[10]. This might be relevant for the population in question as IL-4
levels (though lowly detected) were correlated with IL-13 levels.
We conclude by congratulating the authors [1] for their distinct
observation of a severe asthma endotype with raised IL-13, neutrophilia
and dysbiosis in the peripheral airways. This population may benefit
from an IL-4/IL-13 targeted therapy, although the “classic” T2
selection biomarker – eosinophilia is not immediately noticeable. The
work also reinforces the value of using airway secretions, instead of
blood, to investigate asthma endotypes.
Author contributions : MM and AI wrote the manuscript.