Advertisement

Pathway discovery using transcriptomic profiles in adult-onset severe asthma

      Background

      Adult-onset severe asthma is characterized by highly symptomatic disease despite high-intensity asthma treatments. Understanding of the underlying pathways of this heterogeneous disease is needed for the development of targeted treatments. Gene set variation analysis is a statistical technique used to identify gene profiles in heterogeneous samples.

      Objective

      We sought to identify gene profiles associated with adult-onset severe asthma.

      Methods

      This was a cross-sectional, observational study in which adult patients with adult-onset of asthma (defined as starting at age ≥18 years) as compared with childhood-onset severe asthma (<18 years) were selected from the U-BIOPRED cohort. Gene expression was assessed on the total RNA of induced sputum (n = 83), nasal brushings (n = 41), and endobronchial brushings (n = 65) and biopsies (n = 47) (Affymetrix HT HG-U133+ PM). Gene set variation analysis was used to identify differentially enriched predefined gene signatures of leukocyte lineage, inflammatory and induced lung injury pathways.

      Results

      Significant differentially enriched gene signatures in patients with adult-onset as compared with childhood-onset severe asthma were identified in nasal brushings (5 signatures), sputum (3 signatures), and endobronchial brushings (6 signatures). Signatures associated with eosinophilic airway inflammation, mast cells, and group 3 innate lymphoid cells were more enriched in adult-onset severe asthma, whereas signatures associated with induced lung injury were less enriched in adult-onset severe asthma.

      Conclusions

      Adult-onset severe asthma is characterized by inflammatory pathways involving eosinophils, mast cells, and group 3 innate lymphoid cells. These pathways could represent useful targets for the treatment of adult-onset severe asthma.

      Graphical abstract

      Key words

      Abbreviations used:

      dES (Difference in enrichment scores), ES (Enrichment score), GSVA (Gene set variation analysis), ILC3 (Group 3 innate lymphoid cells)
      Severe asthma, affecting 3.6% of patients with asthma,
      • Hekking P.P.
      • Wener R.R.
      • Amelink M.
      • Zwinderman A.H.
      • Bouvy M.L.
      • Bel E.H.
      The prevalence of severe refractory asthma.
      is a heterogeneous disease that remains uncontrolled despite treatment with high-dose systemic corticosteroids, or the need of oral steroids or biologicals to achieve disease control.
      • Chung K.F.
      • Wenzel S.E.
      • Brozek J.L.
      • Bush A.
      • Castro M.
      • Sterk P.J.
      • et al.
      International ERS/ATS guidelines on definition, evaluation and treatment of severe asthma.
      This group accounts for a disproportional part of the economic and health burden of asthma, emphasizing the need for the development of novel treatments.
      • Accordini S.
      • Corsico A.G.
      • Braggion M.
      • Gerbase M.W.
      • Gislason D.
      • Gulsvik A.
      • et al.
      The cost of persistent asthma in Europe: an international population-based study in adults.
      A better insight in the underlying mechanisms of severe asthma and its phenotypes is therefore needed.
      • Chung K.F.
      Targeting the interleukin pathway in the treatment of asthma.
      Age of onset of asthma has been recognized as an important determinant of severe asthma phenotypes with the delineation of childhood-onset versus an adult-onset phenotype.
      • Amelink M.
      • de Groot J.C.
      • de Nijs S.B.
      • Lutter R.
      • Zwinderman A.H.
      • Sterk P.J.
      • et al.
      Severe adult-onset asthma: a distinct phenotype.
      Using cluster analysis of clinically available disease markers in mild to severe asthma populations has led to conflicting observations, with adult-onset asthma being either predominantly characterized by female sex and obesity, or by active airway inflammation, fixed airflow limitation, and male sex.
      • Haldar P.
      • Pavord I.D.
      • Shaw D.E.
      • Berry M.A.
      • Thomas M.
      • Brightling C.E.
      • et al.
      Cluster analysis and clinical asthma phenotypes.
      • Moore W.C.
      • Meyers D.A.
      • Wenzel S.E.
      • Teague W.G.
      • Li H.
      • Li X.
      • et al.
      Identification of asthma phenotypes using cluster analysis in the Severe Asthma Research Program.
      • Lefaudeux D.
      • De Meulder B.
      • Loza M.J.
      • Peffer N.
      • Rowe A.
      • Baribaud F.
      • et al.
      U-BIOPRED Study Group
      U-BIOPRED clinical adult asthma clusters linked to a subset of sputum omics.
      In addition, patients with adult-onset severe asthma not only were found to be highly symptomatic but were also more often nonatopic with the presence of eosinophilic airway inflammation or higher circulating neutrophil counts.
      • Amelink M.
      • de Groot J.C.
      • de Nijs S.B.
      • Lutter R.
      • Zwinderman A.H.
      • Sterk P.J.
      • et al.
      Severe adult-onset asthma: a distinct phenotype.
      • Miranda C.
      • Busacker A.
      • Balzar S.
      • Trudeau J.
      • Wenzel S.E.
      Distinguishing severe asthma phenotypes: role of age at onset and eosinophilic inflammation.
      Hence, adult-onset severe asthma could present with various bioclinical phenotypes.
      • Wenzel S.E.
      Asthma: defining of the persistent adult phenotypes.
      Contradictory underlying mechanisms have been suggested, including TH2-low inflammation on the one hand
      • Wenzel S.E.
      Asthma phenotypes: the evolution from clinical to molecular approaches.
      and TH2-high inflammation based on associations with specific IgE in a subgroup of adult-onset patients with severe asthma on the other.
      • Song W.J.
      • Sintobin I.
      • Sohn K.H.
      • Kang M.G.
      • Park H.K.
      • Jo E.J.
      • et al.
      Staphylococcal enterotoxin IgE sensitization in late-onset severe eosinophilic asthma in the elderly.
      Therefore, it is likely that multiple mechanistic pathways are involved in adult-onset severe asthma.
      We have examined the gene networks underlying adult-onset severe asthma to obtain molecular phenotypes. This approach would be useful to elucidate the relevant biological pathways that may be causing late-onset asthma. In heterogeneous samples such as adult-onset severe asthma, conventional single-gene expression comparison between groups may not capture subtle variations in (composite) underlying pathways. We used gene set variation analysis (GSVA), which is a statistical technique that enables the discovery of inflammatory and leukocyte lineage gene signatures by comparing combined enrichment scores (ESs) of established and predefined gene sets, especially in heterogeneous samples.
      • Hanzelmann S.
      • Castelo R.
      • Guinney J.
      GSVA: gene set variation analysis for microarray and RNA-seq data.
      We hypothesized that adult-onset severe asthma represents a complex phenotype, associated with specific airway transcriptomic profiles defined by GSVA. We aimed to discover and describe these profiles by studying the transcriptomic data obtained from nasal brushings, sputum and bronchial brushings, and biopsies collected in the adult severe asthma population of the multicenter pan-European U-BIOPRED cohort.
      • Shaw D.E.
      • Sousa A.R.
      • Fowler S.J.
      • Fleming L.J.
      • Roberts G.
      • Corfield J.
      • et al.
      Clinical and inflammatory characteristics of the European U-BIOPRED adult severe asthma cohort.

      Methods

       Design and subjects

      Details on the selection of patients and data collection have been published previously.
      • Shaw D.E.
      • Sousa A.R.
      • Fowler S.J.
      • Fleming L.J.
      • Roberts G.
      • Corfield J.
      • et al.
      Clinical and inflammatory characteristics of the European U-BIOPRED adult severe asthma cohort.
      In short, this was a cross-sectional, observational study using the baseline visits of the U-BIOPRED cohort from 16 clinical centers in 11 countries across Europe. A total of 421 adult patients (≥18 years) with severe asthma were included according to the Innovative Medicines Initiative (IMI) consensus statement.
      • Bel E.H.
      • Sousa A.
      • Fleming L.
      • Bush A.
      • Chung K.F.
      • Versnel J.
      • et al.
      Diagnosis and definition of severe refractory asthma: an international consensus statement from the Innovative Medicine Initiative (IMI).
      This comprised a confirmed diagnosis of asthma and either uncontrolled disease despite treatment with high-dose inhalation medication (≥1000 μg fluticasone equivalent) and a second controller, or the need of systemic oral corticosteroids or omalizumab to achieve asthma control, regardless of smoking history. The U-BIOPRED study was registered at ClinicalTrials.gov (identifier: NCT01976767) and was approved by all Medical Ethics Boards. All patients provided written informed consent.

       Data collection

      All data and samples were collected according to predefined standard operating procedures.
      • Shaw D.E.
      • Sousa A.R.
      • Fowler S.J.
      • Fleming L.J.
      • Roberts G.
      • Corfield J.
      • et al.
      Clinical and inflammatory characteristics of the European U-BIOPRED adult severe asthma cohort.
      Clinical data, such as age, sex, body mass index, age of onset of the disease, oral corticosteroid usage, and smoking history were collected by history taking. Control of the disease was assessed with the 7-item Asthma Control Questionnaire
      • Juniper E.F.
      • Bousquet J.
      • Abetz L.
      • Bateman E.D.
      Identifying ‘well-controlled’ and ‘not well-controlled’ asthma using the Asthma Control Questionnaire.
      and quality of life with the Asthma Quality of Life questionnaire.
      • Juniper E.F.
      • Buist A.S.
      • Cox F.M.
      • Ferrie P.J.
      • King D.R.
      Validation of a standardized version of the Asthma Quality of Life Questionnaire.
      Lung function was measured prebronchodilator and postbronchodilator according to standardized procedures.
      • Miller M.R.
      • Hankinson J.
      • Brusasco V.
      • Burgos F.
      • Casaburi R.
      • Coates A.
      • et al.
      Standardisation of spirometry.
      Atopy was defined as the presence of sensitization to 1 or more common aeroallergens, identified either by a positive skin prick test result (wheal diameter ≥3 mm) to an allergen extract or positive testing by analysis of allergen-specific IgE in serum (≥0.35 kU/L). Eosinophil and neutrophil counts were assessed in blood and induced sputum according to European Respiratory Society recommendations.
      • Paggiaro P.L.
      • Chanez P.
      • Holz O.
      • Ind P.W.
      • Djukanovic R.
      • Maestrelli P.
      • et al.
      Sputum induction.

       Adult-onset asthma

      Patients were qualified as adult-onset patients with severe asthma when they had their first diagnosis of asthma or onset of symptoms when they were 18 years or older, whereas childhood-onset severe asthma was defined as first symptoms on age less than 18 years.
      • Amelink M.
      • de Groot J.C.
      • de Nijs S.B.
      • Lutter R.
      • Zwinderman A.H.
      • Sterk P.J.
      • et al.
      Severe adult-onset asthma: a distinct phenotype.

       RNA sampling

      For details on sampling of the nasal brushings, sputum induction and bronchial brushings, and biopsies, see this article's Online Repository at www.jacionline.org.

       Microarray analyses of mRNA

      Gene expression was assessed on the total RNA from all available samples from nasal brushings (n = 41), sputum (n = 83), bronchial brushings (n = 65), or bronchial biopsies (n = 47) in U-BIOPRED severe asthma cohort. There was incomplete overlap between the various samples per patient (Fig 1). The Affymetrix HT HG-U133+ PM microarray platform (Affymetrix, Santa Clara, Calif) was used for analysis. Preprocessing and quality control were performed with multiarray average normalization (Almac, Craiganvond, United Kingdom). Subsequently, acquired CEL files were normalized, quality control was applied to exclude technical outliers (chip image analysis, Affymetrix GeneChip QC, RNA degradation analysis, distribution analysis, principal-component analysis, and correlation analysis), and renormalized using the robust multiarray method. Technical batch effects (eg, from microarray hybridization date/lot, RNA processing batch) were adjusted in the data matrices using linear modeling of batch (as random factor).
      Figure thumbnail gr1
      Fig 1Venn diagram indicating overlap of patients of which samples have been collected.

       Statistical analyses

       Group comparison of clinical variables

      Normally distributed variables were summarized by mean ± SD, skewed variables by median (interquartile range), and categorical variables by their frequencies and proportions. Group comparisons were done with independent t tests, Mann-Whitney U tests, and chi-square tests, as appropriate. Variables with a P value of less than .05 were considered significantly different.

       Gene set variation analysis

      GSVA is a statistical technique used for the exploration of the variation in underlying mechanisms between groups.
      • Hanzelmann S.
      • Castelo R.
      • Guinney J.
      GSVA: gene set variation analysis for microarray and RNA-seq data.
      A priori, 105 sets of genes were selected, on the basis of all available gene expression publications and data on airways disease obtained from in vivo and in vitro human sample studies and in vitro murine models. These included gene sets associated with the presence of asthma, leukocytes, and those associated with induced lung injury. The latter contained sets of genes that were identified after admission of poly(I:C) (an analogue of the double-stranded RNA that is produced by respiratory viruses during infection and therefore can be used as a model for exacerbations) and bleomycin (used in murine models in research of the course of lung fibrosis)
      • Peng R.
      • Sridhar S.
      • Tyagi G.
      • Phillips J.E.
      • Garrido R.
      • Harris P.
      • et al.
      Bleomycin induces molecular changes directly relevant to idiopathic pulmonary fibrosis: a model for “active” disease.
      (see Table E6 in this article's Online Repository at www.jacionline.org)

       ESs and false discovery

      ESs were calculated for each of these gene sets, on the basis of the expression of each of the genes within these sets, for each patient. Subsequently, mean ESs were calculated for each group (ie, adult-onset severe asthma and childhood-onset severe asthma). Generalized linear models, including correction for smoking status and corticosteroid usage, were used to compare ESs between the groups, with the estimate ranging from −1 to 1. To restrain false discovery, standard criteria were used by only considering those gene signatures that had differences of greater than or equal to 0.2 in ESs (dES) between the groups and P < .05 significantly differentially expressed and meaningful. This procedure follows the Microarray Consortium for Quality Control recommendations for the importance of applying group-difference thresholds.
      • Shi L.
      • Reid L.H.
      • Jones W.D.
      • Shippy R.
      • Warrington J.A.
      • Baker S.C.
      • et al.
      The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements.

      Results

      Of all 421 adult patients with severe asthma in the U-BIOPRED cohort, 253 (60.1%) had adult-onset severe asthma and 158 (37.5%) had childhood-onset severe asthma. Data on age of onset were lacking in 10 (2.4%) patients and these were excluded in subsequent analysis. Adult-onset patients with severe asthma had a mean age of 56.2 ± 11.1 years and most (59.3%) were women (Table I). Furthermore, patients with adult-onset severe asthma had a significantly higher number of pack-years as compared with patients with childhood-onset severe asthma and included significantly more ex- and current smokers. Finally, the adult-onset patients were less often atopic and had higher absolute eosinophil blood counts and higher eosinophil percentage in sputum (Table I).
      Table ICharacteristics of patients with adult-onset severe asthma
      CharacteristicOnset <18 yOnset ≥18 yP value
      N158253
      Sex: female
      Number (percentage).
      107 (67.7%)150 (59.3%).09
      Age (y)
      Mean ± SD.
      44.7 ± 13.956.2 ± 11.1<.01
      BMI (kg/m2)
      Mean ± SD.
      29.92 ± 6.628.65 ± 6.1.05
      Age of onset (y)
      Median (interquartile range).
      5.0 (2.0-11.0)38.0 (28.0-48.0)<.01
      Smoking status
      Number (percentage).
      <.01
       Never smoker118 (74.7%)138 (54.5%)
       Ex-smoker25 (15.8%)88 (34.8%)
       Current smoker15 (9.5%)27 (10.7%)
      Pack-years (in ex/current smokers)
      Median (interquartile range).
      8.0 (1.7-17.8)14.0 (5.0-22.5).03
      OCS dose (in OCS users)
      Median (interquartile range).
      10.0 (7.9-20.0)10.0 (7.5-15.0).27
      Atopy—positive
      Number (percentage).
      124 (78.5%)146 (57.7%)<.01
      ACQ7
      Mean ± SD.
      2.9 ± 1.32.5 ± 1.2<.01
      Postbronchodilator FEV1 % pred
      Mean ± SD.
      76.0 ± 23.375.5 ± 20.0.82
      Blood eosinophils (×109/L)
      Median (interquartile range).
      0.20 (0.1-0.3)0.27 (0.1-0.5)<.01
      Blood neutrophils (×109/L)
      Mean ± SD.
      5.76 ± 2.75.36 ± 2.3.11
      Sputum eosinophils %
      Median (interquartile range).
      1.36 (0.2-4.8)5.3 (1.2-24.4)<.01
      Sputum neutrophils %
      Mean ± SD.
      55.9 ± 26.151.8 ± 26.1.31
      ACQ7, 7-Item Asthma Control Questionnaire; BMI, body mass index; OCS, oral corticosteroid. Boldface indicates P < .05.
      Number (percentage).
      Mean ± SD.
      Median (interquartile range).
      As expected, and in line with the patient characteristics of the complete U-BIOPRED severe asthma cohort (Table I), adult-onset patients with severe asthma in all 4 compartment subsets were significantly older than patients with severe childhood-onset asthma. In addition, eosinophil percentage in sputum was significantly higher in all subsets of patients with severe adult-onset asthma. Blood eosinophil count was only significantly higher in the subset of patients from whom sputum was collected, whereas it tended to be higher in the subsets of patients from whom bronchial biopsies were collected (Table II; see Tables E1-E4 in this article's Online Repository at www.jacionline.org).
      Table IICharacteristics of patients with adult-onset and childhood-onset severe asthma according to site of mRNA collection
      CharacteristicNasal brushingsSputumEndobronchial brushingsEndobronchial biopsies
      Onset <18 yOnset ≥18 yOnset <18 yOnset ≥18 yOnset <18 yOnset ≥18 yOnset <18 yOnset ≥18 y
      N2120325132332824
      Sex: female
      Number (percentage).
      14 (66.7%)8 (40.0%)18 (56.2%)31 (60.8%)19 (59.4%)13 (39.4%)17 (60.7%)11 (45.8%)
      Age (y)
      Mean ± SD.
      43.5 ± 14.4
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      58.4 ± 7.7
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      47.16 ± 11.90
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      57.45 ± 10.09
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      43.3 ± 13.9
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      55.8 ± 7.6
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      44.6 ± 13.0
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      57.3 ± 7.0
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      BMI (kg/m2)
      Mean ± SD.
      32.8 ± 7.130.7 ± 5.427.46 ± 5.5628.80 ± 4.9931.0 ± 6.629.4 ± 6.130.3 ± 6.728.6 ± 5.3
      Age of onset (y)
      Median (interquartile range).
      5.0 (2.0-9.0)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      42.0 (34.5-49.3)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      5.0 (3.0-10.3)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      39.0 (27.0-50.0)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      5.0 (2.0-9.0)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      43.0 (32.0-49.0)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      2.5 (2.0-6.0)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      43.0 (36.5-48.5)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      Smoking status
      Number (percentage).
       Never smoker15 (71.4%)10 (50.0%)25 (78.1%)22 (43.1%)22 (68.8)16 (48.5)19 (67.9%)14 (58.3%)
       Ex-smokers5 (23.8%)8 (40.0%)4 (12.5%)24 (47.1%)8 (25.0)13 (39.4)7 (25.0%)5 (20.8%)
       Current smokers1 (4.8%)2 (10.0%)3 (9.4%)5 (9.8%)2 (6.2)4 (12.1)2 (7.1%)5 (20.8%)
      Pack-years
      Median (interquartile range).
      12.6 (3.6-20.4)16.0 (5.5-38.8)15.0 (8.5-22.4)10.0 (2.3-17.0)4.1 (1.6-20.4)16.4 (5.5-33.0)3.0 (1.5-20.5)20.5 (16.1-30.8)
      Atopy—positive
      Number (percentage).
      18 (82.1%)12 (60.0%)29 (90.6%)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      27 (52.9%)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      25 (78.1%)21 (63.6%)23 (82.1%)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      12 (50.0%)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      OCS dose
      Median (interquartile range).
      20.0 (10.0-25.0)10.0 (7.6-10.0)10.0 (6.9-10.0)10.0 (7.5-15.0)10.0 (7.7-23.8)10.0 (7.5-11.3)10.0 (7.8-20.0)10.0 (7.6-10.0)
      ACQ7
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      2.07 ± 1.22.21 ± 1.02.9 ± 1.52.4 ± 1.22.3 (1.2)2.4 (1.1)2.3 ± 1.32.3 ± 1.1
      Postbronchodilator FEV1 % pred
      Mean ± SD.
      90.5 ± 20.3
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      76.7 ± 21.9
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      68.0 ± 24.372.1 ± 19.984.3 ± 19.475.0 ± 20.879.4 ± 19.578.2 ± 20.0
      Blood eosinophils
      Median (interquartile range).
      0.20 (0.10-0.22)0.21 (0.10-0.46)0.20 (0.14-0.30)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      0.37 (0.20-0.50)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      0.20 (0.10-0.30)0.20 (0.14-0.30)0.20 (0.10-0.30)0.25 (0.18-0.45)
      Blood neutrophils
      Mean ± SD.
      5.4 ± 2.45.1 ± 1.95.46 ± 2.745.41 ± 2.205.15 ± 2.595.64 ± 2.425.32 ± 2.645.31 ± 2.34
      Sputum eosinophils %
      Median (interquartile range).
      0.49 (0.15-1.70)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      8.52 (1.10-18.62)
      Number (percentage).
      2.32 (0.20-5.30)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      7.93 (1.23-25.92)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      0.58 (0.24-1.27)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      12.42 (1.58-24.61)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      0.71 (0.28-2.71)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      8.01 (3.32-20.20)
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      Sputum neutrophils %
      Mean ± SD.
      55.3 ± 22.254.2 ± 18.965.1 ± 25.454.9 ± 25.652.1 ± 23.455.3 ± 22.354.5 ± 22.648.7 ± 22.8
      ACQ7, 7-Item Asthma Control Questionnaire; BMI, body mass index; OCS, oral corticosteroid. Boldface values are significantly different between the groups (P < .05).
      Number (percentage).
      Mean ± SD.
      Significantly different between groups within sample (P < .05); for P values, see this article's Online Repository at www.jacionline.org.
      § Median (interquartile range).

       Significant different gene signatures

      In nasal brushings, 5 significantly different gene signatures were identified, in sputum 3, and in endobronchial brushings 6 (Table III; see Fig E1 in this article's Online Repository at www.jacionline.org).
      Table IIISummary of identified significantly different enriched gene signatures in patients with adult-onset severe asthma
      Gene signaturesGene signature associated withNasal brushingsSputumBronchial brushings
      Asthma gene signaturesAsthma—up
      In vivo model in human sample.
      +0.23
      Asthma—down
      In vivo model in human sample.
      −0.21
      Asthma (house dust mite rhesus–induced model)—up
      In vivo model in human sample.
      +0.21
      Fluticasone treatment in asthma—down
      In vivo model in human sample.
      +0.30
      Leukocyte gene signaturesMacrophage/GM-CSF—down
      In vitro model in human sample.
      +0.20
      Macrophage/GM-CSF/IFN-γ—up
      In vitro model in human sample.
      −0.20
      Eosinophil—up
      In vivo model in human sample.
      +0.30
      Mast cell—up
      In vivo model in human sample.
      +0.33+0.27
      ILC3—up
      In vivo model in murine model.
      +0.31
      Induced lung injury gene signaturesBleomycin-induced lung injury (day 14)—down
      In vivo model in murine model.
      +0.24
      Bleomycin-induced lung injury (day 21)—down
      In vivo model in murine model.
      +0.31
      Bleomycin-induced lung injury (day 28)—down
      In vivo model in murine model.
      +0.20
      Poly(I:C) induced inflammation—up
      In vivo model in murine model.
      −0.21
      Differences in mean gene signature ESs (dES) between adult- and childhood-onset severe asthma and childhood-onset severe asthma are shown (boldface indicates that values are lower in adult-onset severe asthma). ESs range from −1 to 1.
      In vivo model in human sample.
      In vitro model in human sample.
      In vivo model in murine model.

       Asthma gene signatures

      In nasal brushings, a gene signature was identified that consists of genes that are downregulated following fluticasone treatment in asthma (see Table E5 in this article's Online Repository at www.jacionline.org, signature #12).
      • Xue J.
      • Schmidt S.V.
      • Sander J.
      • Draffehn A.
      • Krebs W.
      • Quester I.
      • et al.
      Transcriptome-based network analysis reveals a spectrum model of human macrophage activation.
      This signature had a significantly higher ES in patients with adult-onset versus childhood-onset severe asthma (dES = 0.30; P < .01) (Table III; Fig 2), indicating a different response to treatment with inhaled corticosteroid. In addition, in bronchial brushings, 3 different gene signatures that are associated with the presence of asthma were significantly differentially enriched in patients with adult-onset compared with childhood-onset severe asthma (Table III; Fig 3). First, a gene signature that is upregulated in association with asthma (Table E5, signature #9)
      • Woodruff P.G.
      • Boushey H.A.
      • Dolganov G.M.
      • Barker C.S.
      • Yang Y.H.
      • Donnelly S.
      • et al.
      Genome-wide profiling identifies epithelial cell genes associated with asthma and with treatment response to corticosteroids.
      was more enriched in adult-onset severe asthma (dES = 0.23; P = .02). Second, a corresponding set of genes that is downregulated in association with asthma (Table E5, signature #10)
      • Woodruff P.G.
      • Boushey H.A.
      • Dolganov G.M.
      • Barker C.S.
      • Yang Y.H.
      • Donnelly S.
      • et al.
      Genome-wide profiling identifies epithelial cell genes associated with asthma and with treatment response to corticosteroids.
      was suppressed in adult-onset asthma (dES = −0.21; P = .03), and, third, a gene signature associated with the presence of house dust mite–induced asthma (Table E5, signature #8)
      • Abbas A.R.
      • Jackman J.K.
      • Bullens S.L.
      • Davis S.M.
      • Choy D.F.
      • Fedorowicz G.
      • et al.
      Lung gene expression in a rhesus allergic asthma model correlates with physiologic parameters of disease and exhibits common and distinct pathways with human asthma and a mouse asthma model.
      was also more enriched in adult-onset severe asthma (dES = 0.21; P = .01). These data help validate the GSVA approach undertaken here.
      Figure thumbnail gr2
      Fig 2Gene signatures that have been identified to be differentially enriched in nasal brushings between patients with adult-onset severe asthma and childhood-onset severe asthma. ESs range from −1 to 1.
      Figure thumbnail gr3
      Fig 3Gene signatures that have been identified to be differentially enriched in endobronchial brushings between patients with adult-onset severe asthma and childhood-onset severe asthma. ESs range from −1 to 1.

       Leukocyte gene signatures

      Gene signatures associated with the presence or absence of leukocyte subtypes were differentially enriched in nasal brushings, sputum, and bronchial brushings in patients with adult-onset compared with childhood-onset severe asthma (Table III; Fig 2, Fig 3, Fig 4). In bronchial brushings, a gene signature associated with eosinophilia (Table E5, signature #56)
      • Leaker B.R.
      • Malkov V.A.
      • Mogg R.
      • Ruddy M.K.
      • Nicholson G.C.
      • Tan A.J.
      • et al.
      The nasal mucosal late allergic reaction to grass pollen involves type 2 inflammation (IL-5 and IL-13), the inflammasome (IL-1beta), and complement.
      was significantly more enriched (dES = 0.30; P < .01) in adult-onset compared with childhood-onset asthma. In addition, a set of genes specific for mast cells (Table E5, signature #52),
      • Leaker B.R.
      • Malkov V.A.
      • Mogg R.
      • Ruddy M.K.
      • Nicholson G.C.
      • Tan A.J.
      • et al.
      The nasal mucosal late allergic reaction to grass pollen involves type 2 inflammation (IL-5 and IL-13), the inflammasome (IL-1beta), and complement.
      was more enriched both in sputum (dES = 0.33; P = .01) and in endobronchial brushings (dES = 0.27; P = .02) of adult-onset patients with severe asthma (Table III; Figs 3 and 4). However, a wide scatter of ESs of this mast cell gene signature was observed among patients with adult-onset severe asthma especially in sputum (Fig 4). Furthermore, 2 gene signatures associated with the activation cascade of macrophages were differentially enriched in adult-onset asthma: first, a set of upregulated genes in macrophages stimulated with GM-CSF and IFN-γ (Table E5, signature #87),
      • Xue J.
      • Schmidt S.V.
      • Sander J.
      • Draffehn A.
      • Krebs W.
      • Quester I.
      • et al.
      Transcriptome-based network analysis reveals a spectrum model of human macrophage activation.
      which corresponded to M1-like macrophages, was less enriched in sputum of patients with adult-onset severe asthma (dES = −0.20; P < .01), whereas a signature of downregulated genes in GM-CSF–differentiated monocyte-derived macrophages (Table E5, signature #98)
      • Xue J.
      • Schmidt S.V.
      • Sander J.
      • Draffehn A.
      • Krebs W.
      • Quester I.
      • et al.
      Transcriptome-based network analysis reveals a spectrum model of human macrophage activation.
      was more enriched in nasal brushings (dES = 0.20; P = .01) of patients with adult-onset severe asthma, suggesting lower levels of macrophage activation. Finally, the gene signature associated with the presence of group 3 innate lymphoid cells (ILC3s) in a murine model (Table E5, signature #104)
      • Robinette M.L.
      • Fuchs A.
      • Cortez V.S.
      • Lee J.S.
      • Wang Y.
      • Durum S.K.
      • et al.
      Transcriptional programs define molecular characteristics of innate lymphoid cell classes and subsets.
      was significantly more enriched in nasal brushings (dES = 0.31; P < .01). Notably, 2 gene signatures associated with type 2-high inflammation in asthma tended to be significantly more enriched in bronchial brushings of patients with adult-onset severe asthma (ie, type 2-high inflammation in asthma [dES = 0.19; P < .01; Table E5, signature #13]
      • Choy D.F.
      • Modrek B.
      • Abbas A.R.
      • Kummerfeld S.
      • Clark H.F.
      • Wu L.C.
      • et al.
      Gene expression patterns of Th2 inflammation and intercellular communication in asthmatic airways.
      and IL-13 inflammation [dES = 0.18; P = .02; Table E5, signature #101]) (Fig 5).
      Figure thumbnail gr4
      Fig 4Gene signatures that have been identified to be differentially enriched in sputum between patients with adult-onset severe asthma and childhood-onset severe asthma. ESs range from −1 to 1.
      Figure thumbnail gr5
      Fig 5Type 2 signatures in bronchial brushings that are significantly borderline enriched in adult-onset severe asthma as compared with childhood-onset severe asthma. ESs range from −1 to 1.

       Induced lung inflammation gene signatures

      Finally, gene signatures consisting of a set of downregulated genes identified in a murine model after bleomycin-induced injury were significantly more enriched in nasal brushings and bronchial brushings of patients with adult-onset severe asthma (Table III; Table E5, signatures #10-12).
      • Woodruff P.G.
      • Boushey H.A.
      • Dolganov G.M.
      • Barker C.S.
      • Yang Y.H.
      • Donnelly S.
      • et al.
      Genome-wide profiling identifies epithelial cell genes associated with asthma and with treatment response to corticosteroids.
      Furthermore, a set of genes that is upregulated in mice after the administration of Poly(I:C) (Table E5, signature #13)
      • Harris P.
      • Sridhar S.
      • Peng R.
      • Phillips J.E.
      • Cohn R.G.
      • Burns L.
      • et al.
      Double-stranded RNA induces molecular and inflammatory signatures that are directly relevant to COPD.
      was found to be less enriched in the sputum of patients with adult-onset severe asthma (dES = −0.21; P = .01).

      Discussion

      This study shows that differentially enriched gene signatures can be identified in nasal brushings, sputum, and endobronchial brushings in patients with adult-onset severe asthma as compared with patients with childhood-onset severe asthma. These gene signatures suggest that multiple underlying pathways play a role in adult-onset severe asthma. Identified gene networks are associated with well-recognized inflammatory characterizations, such as eosinophilic airway inflammation, but also newly recognized inflammatory pathways, such as ILC3 pathways. Furthermore, the remarkably polarized differentiation of the gene signature associated with mast cells suggests that these inflammatory cells may be an important cell that contributes to adult-onset severe asthma.
      This study is the first to assess gene networks associated with adult-onset severe asthma. We have used GSVA, a technique that allows identification of gene signatures even in heterogeneous samples, with adjustments for major confounding factors such as smoking status and oral corticosteroid usage.
      • Hanzelmann S.
      • Castelo R.
      • Guinney J.
      GSVA: gene set variation analysis for microarray and RNA-seq data.
      This approach has identified newly recognized as well as previously described pathways
      • Amelink M.
      • de Groot J.C.
      • de Nijs S.B.
      • Lutter R.
      • Zwinderman A.H.
      • Sterk P.J.
      • et al.
      Severe adult-onset asthma: a distinct phenotype.
      • Miranda C.
      • Busacker A.
      • Balzar S.
      • Trudeau J.
      • Wenzel S.E.
      Distinguishing severe asthma phenotypes: role of age at onset and eosinophilic inflammation.
      • Amelink M.
      • de Nijs S.B.
      • Berger M.
      • Weersink E.J.
      • ten Brinke A.
      • Sterk P.J.
      • et al.
      Non-atopic males with adult onset asthma are at risk of persistent airflow limitation.
      • Amelink M.
      • de Nijs S.B.
      • de Groot J.C.
      • van Tilburg P.M.
      • van Spiegel P.I.
      • Krouwels F.H.
      • et al.
      Three phenotypes of adult-onset asthma.
      that distinguish adult-onset severe asthma from childhood-onset asthma. First, this included gene signatures in adult-onset severe asthma that are associated with eosinophilic airway inflammation. This corroborates with findings in patients with mild/moderate and severe asthma with onset during adulthood.
      • Haldar P.
      • Pavord I.D.
      • Shaw D.E.
      • Berry M.A.
      • Thomas M.
      • Brightling C.E.
      • et al.
      Cluster analysis and clinical asthma phenotypes.
      • Miranda C.
      • Busacker A.
      • Balzar S.
      • Trudeau J.
      • Wenzel S.E.
      Distinguishing severe asthma phenotypes: role of age at onset and eosinophilic inflammation.
      • Newby C.
      • Heaney L.G.
      • Menzies-Gow A.
      • Niven R.M.
      • Mansur A.
      • Bucknall C.
      • et al.
      Statistical cluster analysis of the British Thoracic Society Severe Refractory Asthma Registry: clinical outcomes and phenotype stability.
      Second, ILC3-associated gene signatures were identified. Previous studies have shown that ILC3s may play a role in subsets of asthma, such as obesity-associated asthma and neutrophilic asthma.
      • Kim S.H.
      • Sutherland E.R.
      • Gelfand E.W.
      Is there a link between obesity and asthma?.
      However, this is the first study to find an association between adult-onset (severe) asthma and ILC3. Third, our data are strongly pointing toward the presence of mast cells in a subset of patients with adult-onset severe asthma. A role for mast cells in asthma has been described before,
      • Wenzel S.E.
      • Balzar S.
      • Ampleford E.
      • Hawkins G.A.
      • Busse W.W.
      • Calhoun W.J.
      • et al.
      IL4R alpha mutations are associated with asthma exacerbations and mast cell/IgE expression.
      • Hinks T.S.
      • Zhou X.
      • Staples K.J.
      • Dimitrov B.D.
      • Manta A.
      • Petrossian T.
      • et al.
      Innate and adaptive T cells in asthmatic patients: relationship to severity and disease mechanisms.
      • Brightling C.E.
      • Bradding P.
      • Symon F.A.
      • Holgate S.T.
      • Wardlaw A.J.
      • Pavord I.D.
      Mast-cell infiltration of airway smooth muscle in asthma.
      but never before an association with adult-onset severe asthma in particular was shown.
      The strengths of this study are the following. In the U-BIOPRED study, processing and analysis of mRNA were performed centrally to provide optimal final data. By using a statistical technique (GSVA) that allows detection of underlying mechanisms, especially in heterogeneous samples, we were able to identify gene networks associated with pathways that have not been described before. This technique enabled detection of pathways, such as mast cell involvement, that could not be detected by conventional cell count methods in sputum. On the other hand, gene signatures associated with leukocytes that can be detected by these conventional techniques, such as eosinophils, were identified by GSVA, validating this technique. In addition, potential selection bias in this analysis was limited by correcting for smoking status and steroid usage. On the other hand, no correction for cell counts (eg, eosinophils) was applied, because this could result in underdetection of actual pathways involved in adult-onset severe asthma.
      Nevertheless, the study also has particular limitations. The categorizing of patients as adult-onset versus childhood-onset asthma was based on questions in the electronic Case Report From (eCRF), potentially including recall bias. This was inevitable in the present cross-sectional design. Moreover, even though we corrected for steroid usage, we cannot exclude differences in adherence to steroids between patients with adult-onset and childhood-onset asthma. The Medication Adherence Rating Scale questionnaire was obtained in all patients in this study, and did not differ between patients with adult-onset severe asthma and patients with childhood-onset severe asthma (P = .79). This may provide an indication that compliance was comparable between the groups, thereby not affecting the results. We have assessed gene signatures in different compartments from patients within the U-BIOPRED cohort that did not fully overlap (Fig 2). This may have caused incomplete concordance of identified gene signatures between the sample types (Table III). However, the fact that some of the gene signatures were identified in multiple tissues strengthen our observations and the discovery of potentially underlying mechanisms. Furthermore, GSVA is a bioinformatics technique that uses predefined sets of genes derived from previously published studies to form gene signatures. Obviously, these studies have had inevitable flaws in their experiments, which could not be taken into account in the present analysis. In addition, some of the gene signatures were derived from murine model studies, which can only be extrapolated very cautiously to human severe asthma. However, by a priori allowing all available gene signatures in the analysis, including those derived from murine models, potential identification of previously unrecognized underlying pathways was purposely enabled by recommended selection criteria of significant and meaningful signatures. It needs to be emphasized that murine models will certainly deviate from human disease and it remains to be established whether the murine signatures can be used to identify phenotypic differences in severe asthma. Different gene networks have been identified in the different airway compartments (eg, nasal brush and bronchial brush), implying variable disease mechanisms within patients with adult-onset severe asthma. However, the lack of complete overlap between samples makes it impossible to reject the hypothesis of “united airways” in this disease.
      • Wagener A.H.
      • Zwinderman A.H.
      • Luiten S.
      • Fokkens W.J.
      • Bel E.H.
      • Sterk P.J.
      • et al.
      The impact of allergic rhinitis and asthma on human nasal and bronchial epithelial gene expression.
      Finally, the present study population did not include an independent test set or split-half analysis for validation purposes. The sample size of the biological specimens did not allow this. However, we took stringent measures to restrain false discovery by following the standardized Microarray Consortium for Quality Control recommendations for the importance of applying group-difference thresholds for gaining best reproducible results.
      • Shi L.
      • Reid L.H.
      • Jones W.D.
      • Shippy R.
      • Warrington J.A.
      • Baker S.C.
      • et al.
      The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements.
      This has resulted in a rejection of more than 70% of the gene signatures with a P value of less than .05 due to a dES of less than 0.2, as shown in Fig E1.
      On the basis of the nature of this discovery study, we can only speculate about the potential involvement of the identified pathways in adult-onset severe asthma. We have for the first time associated ILC3s with adult-onset severe asthma. The meaning of this result remains to be established, because of the recently observed plasticity of innate lymphoid cells,
      • Bernink J.H.
      • Krabbendam L.
      • Germar K.
      • de Jong E.
      • Gronke K.
      • Kofoed-Nielsen M.
      • et al.
      Interleukin-12 and -23 control plasticity of CD127(+) group 1 and group 3 innate lymphoid cells in the intestinal lamina propria.
      and IL-17–associated (promoted by ILC3)
      • Spits H.
      • Artis D.
      • Colonna M.
      • Diefenbach A.
      • Di Santo J.P.
      • Eberl G.
      • et al.
      Innate lymphoid cells–a proposal for uniform nomenclature.
      gene signatures were not identified to be significantly associated with adult-onset severe asthma. Our results, however, are strongly indicative of mast cell involvement in a subset of adult-onset severe asthma, which goes beyond a general role for mast cells in refractory asthma.
      • Brightling C.E.
      • Bradding P.
      • Symon F.A.
      • Holgate S.T.
      • Wardlaw A.J.
      • Pavord I.D.
      Mast-cell infiltration of airway smooth muscle in asthma.
      This may fit in with recent observations by Wang et al
      • Wang G.
      • Baines K.J.
      • Fu J.J.
      • Wood L.G.
      • Simpson J.L.
      • McDonald V.M.
      • et al.
      Sputum mast cell subtypes relate to eosinophilia and corticosteroid response in asthma.
      who showed that mast cell phenotypes as measured by RNA expression profiles are associated with clinical expression of the disease. Elucidating the role of mast cells in adult-onset severe asthma could result in novel targets for subgroups of this difficult-to-treat disease. Notably, both eosinophils and mast cells are characterized by IL-33 receptors,
      • Dyer K.D.
      • Percopo C.M.
      • Rosenberg H.F.
      IL-33 promotes eosinophilia in vivo and antagonizes IL-5-dependent eosinophil hematopoiesis ex vivo.
      • Junttila I.S.
      • Watson C.
      • Kummola L.
      • Chen X.
      • Hu-Li J.
      • Guo L.
      • et al.
      Efficient cytokine-induced IL-13 production by mast cells requires both IL-33 and IL-3.
      which have been associated with asthma and progression of the disease.
      • Li R.
      • Yang G.
      • Yang R.
      • Peng X.
      • Li J.
      Interleukin-33 and receptor ST2 as indicators in patients with asthma: a meta-analysis.
      This generates the hypothesis of IL-33 pathways to be involved in adult-onset severe asthma in particular.
      We did not identify induced lung injury and/or fibrosis-associated gene signatures in adult-onset but rather in childhood-onset severe asthma. It has been hypothesized that the epithelial vulnerability and impaired injury repair are important drivers in the onset and course of asthma.
      • Grainge C.L.
      • Davies D.E.
      Epithelial injury and repair in airways diseases.
      Low enrichment of these gene signatures, which we identified in adult-onset severe asthma, may result in impaired injury and repair, which could lead to increased airway inflammation and remodeling. This would corroborate with recent findings of a clinical study in severe asthma in which eosinophilic inflammation and remodeling were associated with late-onset disease.
      • Folliet L.
      • Freymond N.
      • Pacheco Y.
      • Devouassoux G.
      Early- and late-onsets of severe asthma are associated with divergent phenotypes of the disease.
      Finally, no significantly differentially enriched gene signatures were identified in bronchial biopsies. This may seem unexpected, but could be underlined by the heterogeneity of cell types within the bronchial biopsies precluding detection of gene expression profiles, in contrast to the more homogeneous samples from the nasal and bronchial brushings and sputum.
      What is the clinical relevance of our data? Unraveling the underlying mechanisms of severe asthma and its subtypes is needed to improve treatment of severe asthma.
      • Chung K.F.
      Asthma phenotyping: a necessity for improved therapeutic precision and new targeted therapies.
      This particularly holds for adult-onset severe asthma, which often does not respond to regular treatment options such as steroids. Targeted treatments are becoming available for blocking IL-4, IL-5,
      • Bel E.H.
      • Wenzel S.E.
      • Thompson P.J.
      • Prazma C.M.
      • Keene O.N.
      • Yancey S.W.
      • et al.
      Oral glucocorticoid-sparing effect of mepolizumab in eosinophilic asthma.
      and IL-13 of the T2 pathways.
      • Chung K.F.
      Targeting the interleukin pathway in the treatment of asthma.
      • Parulekar A.D.
      • Diamant Z.
      • Hanania N.A.
      Role of T2 inflammation biomarkers in severe asthma.
      Our results of this explorative study are suggestive that other pathways could be examined for targeted treatment in patients with adult-onset severe asthma, including mast cells in a subset of these patients, ILC3, and IL-33.
      In conclusion, adult-onset severe asthma not only represents a clinically heterogeneous phenotype but also is associated with multiple gene expression profiles that indicate complex underlying pathways involving eosinophils, ILC3s, and mast cells. These point toward possible new targets that could represent targeted treatments for subgroups of adult-onset severe asthma.
      Key messages
      • Expression of gene signatures is significantly different in adult-onset severe asthma as compared with childhood-onset severe asthma, indicating distinct underlying mechanisms.
      • Gene profiles identified in adult-onset severe asthma include those associated with eosinophilia, mast cells, and ILC3.
      U-BIOPRED authors: Adcock IM, National Heart and Lung Institute, Imperial College, London, UK; Ahmed H, European Institute for Systems Biology and Medicine, CNRS-ENS-UCBL-INSERM, Lyon, France; Auffray C, European Institute for Systems Biology and Medicine, CNRS-ENS-UCBL-INSERM, Lyon, France; Bakke P, Department of Clinical Science, University of Bergen, Bergen, Norway; Bansal AT, Acclarogen Ltd, St. John's Innovation Centre, Cambridge, UK; Baribaud F, Janssen R&D, Spring House, Pa; Bates S, Respiratory Therapeutic Unit, GSK, London, UK; Bel EH, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands; Bigler J, Previously Amgen Inc; Bisgaard H, COPSAC, Copenhagen Prospective Studies on Asthma in Childhood, Herlev and Gentofte Hospital, University of Copenhagen, Copenhagen, Denmark; Boedigheimer MJ, Amgen Inc, Thousand Oaks, Calif; Bønnelykke K, COPSAC, Copenhagen Prospective Studies on Asthma in Childhood, Herlev and Gentofte Hospital, University of Copenhagen, Copenhagen, Denmark; Brandsma J, University of Southampton, Southampton, UK; Brinkman P, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands; Bucchioni E, Chiesi Pharmaceuticals SPA, Parma, Italy; Burg D, Centre for Proteomic Research, Institute for Life Sciences, University of Southampton, Southampton, UK; Bush A, National Heart and Lung Institute, Imperial College, London, UK, Royal Brompton and Harefield NHS trust, UK; Caruso M, Department of Clinical and Experimental Medicine, University of Catania, Catania, Italy; Chaiboonchoe A, European Institute for Systems Biology and Medicine, CNRS-ENS-UCBL-INSERM, Lyon, France; Chanez P, Assistance publique des Hôpitaux de Marseille - Clinique des bronches, allergies et sommeil, Aix Marseille Université, Marseille, France; Chung FK, National Heart and Lung Institute, Imperial College, London, UK; Compton CH, Respiratory Therapeutic Unit, GSK, London, UK; Corfield J, Areteva R&D, Nottingham, UK; D'Amico A, University of Rome ‘Tor Vergatá, Rome Italy; Dahlen SE, Centre for Allergy Research, Karolinska Institutet, Stockholm, Sweden; De Meulder B, European Institute for Systems Biology and Medicine, CNRS-ENS-UCBL-INSERM, Lyon, France; Djukanovic R, NIHR Southampton Respiratory Biomedical Research Unit and Clinical and Experimental Sciences, Southampton, UK; Erpenbeck VJ, Translational Medicine, Respiratory Profiling, Novartis Institutes for Biomedical Research, Basel, Switzerland; Erzen D, Boehringer Ingelheim Pharma GmbH & Co. KG; Biberach, Germany; Fichtner K, Boehringer Ingelheim Pharma GmbH & Co. KG, Biberach, Germany; Fitch N, BioSci Consulting, Maasmechelen, Belgium; Fleming LJ, National Heart and Lung Institute, Imperial College, London, UK, Royal Brompton and Harefield NHS trust, UK; Formaggio E, previously CROMSOURCE, Verona, Italy; Fowler SJ, Centre for Respiratory Medicine and Allergy, Institute of Inflammation and Repair, University of Manchester and University Hospital of South Manchester, Manchester Academic Health Sciences Centre, Manchester, UK; Frey U, University Children's Hospital, Basel, Switzerland; Gahlemann M, Boehringer Ingelheim (Schweiz) GmbH, Basel, Switzerland; Geiser T, Department of Respiratory Medicine, University Hospital Bern, Switzerland; Guo Y, Data Science Institute, Imperial College, London, UK; Hashimoto S, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands; Haughney J, International Primary Care Respiratory Group, Aberdeen, UK; Hedlin G, Department of Women's and Children's Health & Centre for Allergy Research, Karolinska Institutet, Stockholm, Sweden; Hekking PW, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands; Higenbottam T, Allergy Therapeutics, West Sussex, UK; Hohlfeld JM, Fraunhofer Institute for Toxicology and Experimental Medicine, Hannover, Germany; Holweg C, Respiratory and Allergy Diseases, Genentech, San Francisco, Calif; Horváth I, Semmelweis University, Budapest, Hungary; Howarth P, NIHR Southampton Respiratory Biomedical Research Unit, Clinical and Experimental Sciences and Human Development and Health, Southampton, UK; James AJ, Centre for Allergy Research, Karolinska Institutet, Stockholm, Sweden; Knowles R, Arachos Pharma, Stevenge, UK; Knox AJ, Respiratory Research Unit, University of Nottingham, Nottingham, UK; Krug N, Fraunhofer Institute for Toxicology and Experimental Medicine, Hannover, Germany; Lefaudeux D, European Institute for Systems Biology and Medicine, CNRS-ENS-UCBL-INSERM, Lyon, France; Loza MJ, Janssen R&D, Spring House, Pa; Lutter R, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands; Manta A, Roche Diagnostics GmbH, Mannheim, Germany; Masefield S, European Lung Foundation, Sheffield, UK; Matthews JG, Respiratory and Allergy Diseases, Genentech, San Francisco, Calif; Mazein A, European Institute for Systems Biology and Medicine, CNRS-ENS-UCBL-INSERM, Lyon, France; Meiser A, Data Science Institute, Imperial College, London, UK; Middelveld RJM, Centre for Allergy Research, Karolinska Institutet, Stockholm, Sweden; Miralpeix M, Almirall, Barcelona, Spain; Montuschi P, Università Cattolica del Sacro Cuore, Milan, Italy; Mores N, Università Cattolica del Sacro Cuore, Milan, Italy; Murray CS, Centre for Respiratory Medicine and Allergy, Institute of Inflammation and Repair, University of Manchester and University Hospital of South Manchester, Manchester Academic Health Sciences Centre, Manchester, UK; Musial J, Department of Medicine, Jagiellonian University Medical College, Krakow, Poland; Myles D, Respiratory Therapeutic Unit, GSK, London, UK; Pahus L, Assistance publique des Hôpitaux de Marseille, Clinique des bronches, allergies et sommeil, Espace Éthique Méditerranéen, Aix-Marseille Université, Marseille, France; Pandis I, Data Science Institute, Imperial College, London, UK; Pavlidis S, National Heart and Lung Institute, Imperial College, London, UK; Powel P, European Lung Foundation, Sheffield, UK; Praticò G, CROMSOURCE, Verona, Italy; Puig Valls M, CROMSOURCE, Barcelona, Spain; Rao N, Janssen R&D, Spring House, Pa; Riley J, Respiratory Therapeutic Unit, GSK, London, UK; Roberts A, Asthma UK, London, UK; Roberts G, NIHR Southampton Respiratory Biomedical Research Unit, Clinical and Experimental Sciences and Human Development and Health, Southampton, UK; Rowe A, Janssen R&D, Spring House, Pa; Sandström T, Department of Public Health and Clinical Medicine, Umeå University, Umeå, Sweden; Seibold W, Boehringer Ingelheim Pharma GmbH, Biberach, Germany; Selby A, NIHR Southampton Respiratory Biomedical Research Unit, Clinical and Experimental Sciences and Human Development and Health, Southampton, UK; Shaw DE, Respiratory Research Unit, University of Nottingham, UK; Sigmund R, Boehringer Ingelheim Pharma GmbH & Co KG, Biberach, Germany; Singer F, University Children's Hospital, Zurich, Switzerland; Skipp PJ, Centre for Proteomic Research, Institute for Life Sciences, University of Southampton, Southampton, UK; Sousa AR, Respiratory Therapeutic Unit, GSK, London, UK; Sterk PJ, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands; Sun K, Data Science Institute, Imperial College, London, UK; Thornton B, MSD, Kenilworth, NJ; van Aalderen WM, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands; van Geest M, AstraZeneca, Mölndal, Sweden; Vestbo J, Centre for Respiratory Medicine and Allergy, Institute of Inflammation and Repair, University of Manchester and University Hospital of South Manchester, Manchester Academic Health Sciences Centre, Manchester, UK; Vissing NH, COPSAC, Copenhagen Prospective Studies on Asthma in Childhood, Herlev and Gentofte Hospital, University of Copenhagen, Copenhagen, Denmark; Wagener AH, Academic Medical Center Amsterdam, Amsterdam, The Netherlands; Wagers SS, BioSci Consulting, Maasmechelen, Belgium; Weiszhart Z, Semmelweis University, Budapest, Hungary; Wheelock CE, Centre for Allergy Research, Karolinska Institutet, Stockholm, Sweden; and Wilson SJ, Histochemistry Research Unit, Faculty of Medicine, University of Southampton, Southampton, UK.

      Supplementary data

      References

        • Hekking P.P.
        • Wener R.R.
        • Amelink M.
        • Zwinderman A.H.
        • Bouvy M.L.
        • Bel E.H.
        The prevalence of severe refractory asthma.
        J Allergy Clin Immunol. 2015; 135: 896-902
        • Chung K.F.
        • Wenzel S.E.
        • Brozek J.L.
        • Bush A.
        • Castro M.
        • Sterk P.J.
        • et al.
        International ERS/ATS guidelines on definition, evaluation and treatment of severe asthma.
        Eur Respir J. 2014; 43: 343-373
        • Accordini S.
        • Corsico A.G.
        • Braggion M.
        • Gerbase M.W.
        • Gislason D.
        • Gulsvik A.
        • et al.
        The cost of persistent asthma in Europe: an international population-based study in adults.
        Int Arch Allergy Immunol. 2013; 160: 93-101
        • Chung K.F.
        Targeting the interleukin pathway in the treatment of asthma.
        Lancet. 2015; 386: 1086-1096
        • Amelink M.
        • de Groot J.C.
        • de Nijs S.B.
        • Lutter R.
        • Zwinderman A.H.
        • Sterk P.J.
        • et al.
        Severe adult-onset asthma: a distinct phenotype.
        J Allergy Clin Immunol. 2013; 132: 336-341
        • Haldar P.
        • Pavord I.D.
        • Shaw D.E.
        • Berry M.A.
        • Thomas M.
        • Brightling C.E.
        • et al.
        Cluster analysis and clinical asthma phenotypes.
        Am J Respir Crit Care Med. 2008; 178: 218-224
        • Moore W.C.
        • Meyers D.A.
        • Wenzel S.E.
        • Teague W.G.
        • Li H.
        • Li X.
        • et al.
        Identification of asthma phenotypes using cluster analysis in the Severe Asthma Research Program.
        Am J Respir Crit Care Med. 2010; 181: 315-323
        • Lefaudeux D.
        • De Meulder B.
        • Loza M.J.
        • Peffer N.
        • Rowe A.
        • Baribaud F.
        • et al.
        • U-BIOPRED Study Group
        U-BIOPRED clinical adult asthma clusters linked to a subset of sputum omics.
        J Allergy Clin Immunol. 2017; 139: 1797-1807
        • Miranda C.
        • Busacker A.
        • Balzar S.
        • Trudeau J.
        • Wenzel S.E.
        Distinguishing severe asthma phenotypes: role of age at onset and eosinophilic inflammation.
        J Allergy Clin Immunol. 2004; 113: 101-108
        • Wenzel S.E.
        Asthma: defining of the persistent adult phenotypes.
        Lancet. 2006; 368: 804-813
        • Wenzel S.E.
        Asthma phenotypes: the evolution from clinical to molecular approaches.
        Nat Med. 2012; 18: 716-725
        • Song W.J.
        • Sintobin I.
        • Sohn K.H.
        • Kang M.G.
        • Park H.K.
        • Jo E.J.
        • et al.
        Staphylococcal enterotoxin IgE sensitization in late-onset severe eosinophilic asthma in the elderly.
        Clin Exp Allergy. 2016; 46: 411-421
        • Hanzelmann S.
        • Castelo R.
        • Guinney J.
        GSVA: gene set variation analysis for microarray and RNA-seq data.
        BMC Bioinformatics. 2013; 14: 7
        • Shaw D.E.
        • Sousa A.R.
        • Fowler S.J.
        • Fleming L.J.
        • Roberts G.
        • Corfield J.
        • et al.
        Clinical and inflammatory characteristics of the European U-BIOPRED adult severe asthma cohort.
        Eur Respir J. 2015; 46: 1308-1321
        • Bel E.H.
        • Sousa A.
        • Fleming L.
        • Bush A.
        • Chung K.F.
        • Versnel J.
        • et al.
        Diagnosis and definition of severe refractory asthma: an international consensus statement from the Innovative Medicine Initiative (IMI).
        Thorax. 2011; 66: 910-917
        • Juniper E.F.
        • Bousquet J.
        • Abetz L.
        • Bateman E.D.
        Identifying ‘well-controlled’ and ‘not well-controlled’ asthma using the Asthma Control Questionnaire.
        Respir Med. 2006; 100: 616-621
        • Juniper E.F.
        • Buist A.S.
        • Cox F.M.
        • Ferrie P.J.
        • King D.R.
        Validation of a standardized version of the Asthma Quality of Life Questionnaire.
        Chest. 1999; 115: 1265-1270
        • Miller M.R.
        • Hankinson J.
        • Brusasco V.
        • Burgos F.
        • Casaburi R.
        • Coates A.
        • et al.
        Standardisation of spirometry.
        Eur Respir J. 2005; 26: 319-338
        • Paggiaro P.L.
        • Chanez P.
        • Holz O.
        • Ind P.W.
        • Djukanovic R.
        • Maestrelli P.
        • et al.
        Sputum induction.
        Eur Respir J Suppl. 2002; 37: 3s-8s
        • Peng R.
        • Sridhar S.
        • Tyagi G.
        • Phillips J.E.
        • Garrido R.
        • Harris P.
        • et al.
        Bleomycin induces molecular changes directly relevant to idiopathic pulmonary fibrosis: a model for “active” disease.
        PLoS One. 2013; 8: e59348
        • Shi L.
        • Reid L.H.
        • Jones W.D.
        • Shippy R.
        • Warrington J.A.
        • Baker S.C.
        • et al.
        The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements.
        Nat Biotechnol. 2006; 24: 1151-1161
        • Xue J.
        • Schmidt S.V.
        • Sander J.
        • Draffehn A.
        • Krebs W.
        • Quester I.
        • et al.
        Transcriptome-based network analysis reveals a spectrum model of human macrophage activation.
        Immunity. 2014; 40: 274-288
        • Woodruff P.G.
        • Boushey H.A.
        • Dolganov G.M.
        • Barker C.S.
        • Yang Y.H.
        • Donnelly S.
        • et al.
        Genome-wide profiling identifies epithelial cell genes associated with asthma and with treatment response to corticosteroids.
        Proc Natl Acad Sci U S A. 2007; 104: 15858-15863
        • Abbas A.R.
        • Jackman J.K.
        • Bullens S.L.
        • Davis S.M.
        • Choy D.F.
        • Fedorowicz G.
        • et al.
        Lung gene expression in a rhesus allergic asthma model correlates with physiologic parameters of disease and exhibits common and distinct pathways with human asthma and a mouse asthma model.
        Am J Pathol. 2011; 179: 1667-1680
        • Leaker B.R.
        • Malkov V.A.
        • Mogg R.
        • Ruddy M.K.
        • Nicholson G.C.
        • Tan A.J.
        • et al.
        The nasal mucosal late allergic reaction to grass pollen involves type 2 inflammation (IL-5 and IL-13), the inflammasome (IL-1beta), and complement.
        Mucosal Immunol. 2017; 10: 408-420
        • Robinette M.L.
        • Fuchs A.
        • Cortez V.S.
        • Lee J.S.
        • Wang Y.
        • Durum S.K.
        • et al.
        Transcriptional programs define molecular characteristics of innate lymphoid cell classes and subsets.
        Nat Immunol. 2015; 16: 306-317
        • Choy D.F.
        • Modrek B.
        • Abbas A.R.
        • Kummerfeld S.
        • Clark H.F.
        • Wu L.C.
        • et al.
        Gene expression patterns of Th2 inflammation and intercellular communication in asthmatic airways.
        J Immunol. 2011; 186: 1861-1869
        • Harris P.
        • Sridhar S.
        • Peng R.
        • Phillips J.E.
        • Cohn R.G.
        • Burns L.
        • et al.
        Double-stranded RNA induces molecular and inflammatory signatures that are directly relevant to COPD.
        Mucosal Immunol. 2013; 6: 474-484
        • Amelink M.
        • de Nijs S.B.
        • Berger M.
        • Weersink E.J.
        • ten Brinke A.
        • Sterk P.J.
        • et al.
        Non-atopic males with adult onset asthma are at risk of persistent airflow limitation.
        Clin Exp Allergy. 2012; 42: 769-774
        • Amelink M.
        • de Nijs S.B.
        • de Groot J.C.
        • van Tilburg P.M.
        • van Spiegel P.I.
        • Krouwels F.H.
        • et al.
        Three phenotypes of adult-onset asthma.
        Allergy. 2013; 68: 674-680
        • Newby C.
        • Heaney L.G.
        • Menzies-Gow A.
        • Niven R.M.
        • Mansur A.
        • Bucknall C.
        • et al.
        Statistical cluster analysis of the British Thoracic Society Severe Refractory Asthma Registry: clinical outcomes and phenotype stability.
        PLoS One. 2014; 9: e102987
        • Kim S.H.
        • Sutherland E.R.
        • Gelfand E.W.
        Is there a link between obesity and asthma?.
        Allergy Asthma Immunol Res. 2014; 6: 189-195
        • Wenzel S.E.
        • Balzar S.
        • Ampleford E.
        • Hawkins G.A.
        • Busse W.W.
        • Calhoun W.J.
        • et al.
        IL4R alpha mutations are associated with asthma exacerbations and mast cell/IgE expression.
        Am J Respir Crit Care Med. 2007; 175: 570-576
        • Hinks T.S.
        • Zhou X.
        • Staples K.J.
        • Dimitrov B.D.
        • Manta A.
        • Petrossian T.
        • et al.
        Innate and adaptive T cells in asthmatic patients: relationship to severity and disease mechanisms.
        J Allergy Clin Immunol. 2015; 136: 323-333
        • Brightling C.E.
        • Bradding P.
        • Symon F.A.
        • Holgate S.T.
        • Wardlaw A.J.
        • Pavord I.D.
        Mast-cell infiltration of airway smooth muscle in asthma.
        N Engl J Med. 2002; 346: 1699-1705
        • Wagener A.H.
        • Zwinderman A.H.
        • Luiten S.
        • Fokkens W.J.
        • Bel E.H.
        • Sterk P.J.
        • et al.
        The impact of allergic rhinitis and asthma on human nasal and bronchial epithelial gene expression.
        PLoS One. 2013; 8: e80257
        • Bernink J.H.
        • Krabbendam L.
        • Germar K.
        • de Jong E.
        • Gronke K.
        • Kofoed-Nielsen M.
        • et al.
        Interleukin-12 and -23 control plasticity of CD127(+) group 1 and group 3 innate lymphoid cells in the intestinal lamina propria.
        Immunity. 2015; 43: 146-160
        • Spits H.
        • Artis D.
        • Colonna M.
        • Diefenbach A.
        • Di Santo J.P.
        • Eberl G.
        • et al.
        Innate lymphoid cells–a proposal for uniform nomenclature.
        Nat Rev Immunol. 2013; 13: 145-149
        • Wang G.
        • Baines K.J.
        • Fu J.J.
        • Wood L.G.
        • Simpson J.L.
        • McDonald V.M.
        • et al.
        Sputum mast cell subtypes relate to eosinophilia and corticosteroid response in asthma.
        Eur Respir J. 2016; 47: 1123-1133
        • Dyer K.D.
        • Percopo C.M.
        • Rosenberg H.F.
        IL-33 promotes eosinophilia in vivo and antagonizes IL-5-dependent eosinophil hematopoiesis ex vivo.
        Immunol Lett. 2013; 150: 41-47
        • Junttila I.S.
        • Watson C.
        • Kummola L.
        • Chen X.
        • Hu-Li J.
        • Guo L.
        • et al.
        Efficient cytokine-induced IL-13 production by mast cells requires both IL-33 and IL-3.
        J Allergy Clin Immunol. 2013; 132: 704-712.e10
        • Li R.
        • Yang G.
        • Yang R.
        • Peng X.
        • Li J.
        Interleukin-33 and receptor ST2 as indicators in patients with asthma: a meta-analysis.
        Int J Clin Exp Med. 2015; 8: 14935-14943
        • Grainge C.L.
        • Davies D.E.
        Epithelial injury and repair in airways diseases.
        Chest. 2013; 144: 1906-1912
        • Folliet L.
        • Freymond N.
        • Pacheco Y.
        • Devouassoux G.
        Early- and late-onsets of severe asthma are associated with divergent phenotypes of the disease.
        Eur Respir J. 2015; 46: PA5009https://doi.org/10.1183/13993003.congress-2015.PA5009
        • Chung K.F.
        Asthma phenotyping: a necessity for improved therapeutic precision and new targeted therapies.
        J Intern Med. 2016; 279: 192-204
        • Bel E.H.
        • Wenzel S.E.
        • Thompson P.J.
        • Prazma C.M.
        • Keene O.N.
        • Yancey S.W.
        • et al.
        Oral glucocorticoid-sparing effect of mepolizumab in eosinophilic asthma.
        N Engl J Med. 2014; 371: 1189-1197
        • Parulekar A.D.
        • Diamant Z.
        • Hanania N.A.
        Role of T2 inflammation biomarkers in severe asthma.
        Curr Opin Pulm Med. 2016; 22: 59-68