Multiview cluster analysis identifies variable corticosteroid response phenotypes in severe asthma

Wei Wu, Seojin Bang, Eugene R. Bleecker, Mario Castro, Loren Denlinger, Serpil C. Erzurum, John V. Fahy, Anne M. Fitzpatrick, Benjamin M. Gaston, Annette T. Hastie, Elliot Israel, Nizar N. Jarjour, Bruce D. Levy, David Mauger, Deborah A. Meyers, Wendy C. Moore, Michael Peters, Brenda R. Phillips, Wanda Phipatanakul, Ronald L. Sorkness & 1 others Sally E. Wenzel

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Rationale: Corticosteroids (CSs) are the most effective asthma therapy, but responses are heterogeneous and systemic CSs lead to long-term side effects. Therefore, an improved understanding of the contributing factors in CS responses could enhance precision management. Although several factors have been associated with CS responsiveness, no integrated/cluster approach has yet been undertaken to identify differential CS responses. Objectives: To identify asthma subphenotypes with differential responses to CS treatment using an unsupervised multiview learning approach. Methods: Multiple-kernel k-means clustering was applied to 100 clinical, physiological, inflammatory, and demographic variables from 346 adult participants with asthma in the Severe Asthma Research Program with paired (before and 2-3 weeks after triamcinolone administration) sputum data. Machine-learning techniques were used to select the top baseline variables that predicted cluster assignment for a new patient. Measurements and Main Results: Multiple-kernel clustering revealed four clusters of individuals with asthma and different CS responses. Clusters 1 and 2 consisted of young, modestly CS-responsive individuals with allergic asthma and relatively normal lung function, separated by contrasting sputum neutrophil and macrophage percentages after CS treatment. The subjects in cluster 3 had late-onset asthma and low lung function, high baseline eosinophilia, and the greatest CS responsiveness. Cluster 4 consisted primarily of young, obese females with severe airflow limitation, little eosinophilic inflammation, and the least CS responsiveness. The top 12 baseline variables were identified, and the clusters were validated using an independent Severe Asthma Research Program test set. Conclusions: Our machine learning-based approaches provide new insights into the mechanisms of CS responsiveness in asthma, with the potential to improve disease treatment.

Original languageEnglish (US)
Pages (from-to)1358-1367
Number of pages10
JournalAmerican journal of respiratory and critical care medicine
Volume199
Issue number11
DOIs
StatePublished - Jan 1 2019

Fingerprint

Cluster Analysis
Adrenal Cortex Hormones
Asthma
Phenotype
Sputum
Triamcinolone
Lung
Eosinophilia
Therapeutics
Research
Neutrophils
Macrophages
Demography
Learning
Inflammation

All Science Journal Classification (ASJC) codes

  • Pulmonary and Respiratory Medicine
  • Critical Care and Intensive Care Medicine

Cite this

Wu, Wei ; Bang, Seojin ; Bleecker, Eugene R. ; Castro, Mario ; Denlinger, Loren ; Erzurum, Serpil C. ; Fahy, John V. ; Fitzpatrick, Anne M. ; Gaston, Benjamin M. ; Hastie, Annette T. ; Israel, Elliot ; Jarjour, Nizar N. ; Levy, Bruce D. ; Mauger, David ; Meyers, Deborah A. ; Moore, Wendy C. ; Peters, Michael ; Phillips, Brenda R. ; Phipatanakul, Wanda ; Sorkness, Ronald L. ; Wenzel, Sally E. / Multiview cluster analysis identifies variable corticosteroid response phenotypes in severe asthma. In: American journal of respiratory and critical care medicine. 2019 ; Vol. 199, No. 11. pp. 1358-1367.
@article{9dff81e9cb0e4111b03ea0aa32be66e0,
title = "Multiview cluster analysis identifies variable corticosteroid response phenotypes in severe asthma",
abstract = "Rationale: Corticosteroids (CSs) are the most effective asthma therapy, but responses are heterogeneous and systemic CSs lead to long-term side effects. Therefore, an improved understanding of the contributing factors in CS responses could enhance precision management. Although several factors have been associated with CS responsiveness, no integrated/cluster approach has yet been undertaken to identify differential CS responses. Objectives: To identify asthma subphenotypes with differential responses to CS treatment using an unsupervised multiview learning approach. Methods: Multiple-kernel k-means clustering was applied to 100 clinical, physiological, inflammatory, and demographic variables from 346 adult participants with asthma in the Severe Asthma Research Program with paired (before and 2-3 weeks after triamcinolone administration) sputum data. Machine-learning techniques were used to select the top baseline variables that predicted cluster assignment for a new patient. Measurements and Main Results: Multiple-kernel clustering revealed four clusters of individuals with asthma and different CS responses. Clusters 1 and 2 consisted of young, modestly CS-responsive individuals with allergic asthma and relatively normal lung function, separated by contrasting sputum neutrophil and macrophage percentages after CS treatment. The subjects in cluster 3 had late-onset asthma and low lung function, high baseline eosinophilia, and the greatest CS responsiveness. Cluster 4 consisted primarily of young, obese females with severe airflow limitation, little eosinophilic inflammation, and the least CS responsiveness. The top 12 baseline variables were identified, and the clusters were validated using an independent Severe Asthma Research Program test set. Conclusions: Our machine learning-based approaches provide new insights into the mechanisms of CS responsiveness in asthma, with the potential to improve disease treatment.",
author = "Wei Wu and Seojin Bang and Bleecker, {Eugene R.} and Mario Castro and Loren Denlinger and Erzurum, {Serpil C.} and Fahy, {John V.} and Fitzpatrick, {Anne M.} and Gaston, {Benjamin M.} and Hastie, {Annette T.} and Elliot Israel and Jarjour, {Nizar N.} and Levy, {Bruce D.} and David Mauger and Meyers, {Deborah A.} and Moore, {Wendy C.} and Michael Peters and Phillips, {Brenda R.} and Wanda Phipatanakul and Sorkness, {Ronald L.} and Wenzel, {Sally E.}",
year = "2019",
month = "1",
day = "1",
doi = "10.1164/rccm.201808-1543OC",
language = "English (US)",
volume = "199",
pages = "1358--1367",
journal = "American Journal of Respiratory and Critical Care Medicine",
issn = "1073-449X",
publisher = "American Thoracic Society",
number = "11",

}

Wu, W, Bang, S, Bleecker, ER, Castro, M, Denlinger, L, Erzurum, SC, Fahy, JV, Fitzpatrick, AM, Gaston, BM, Hastie, AT, Israel, E, Jarjour, NN, Levy, BD, Mauger, D, Meyers, DA, Moore, WC, Peters, M, Phillips, BR, Phipatanakul, W, Sorkness, RL & Wenzel, SE 2019, 'Multiview cluster analysis identifies variable corticosteroid response phenotypes in severe asthma', American journal of respiratory and critical care medicine, vol. 199, no. 11, pp. 1358-1367. https://doi.org/10.1164/rccm.201808-1543OC

Multiview cluster analysis identifies variable corticosteroid response phenotypes in severe asthma. / Wu, Wei; Bang, Seojin; Bleecker, Eugene R.; Castro, Mario; Denlinger, Loren; Erzurum, Serpil C.; Fahy, John V.; Fitzpatrick, Anne M.; Gaston, Benjamin M.; Hastie, Annette T.; Israel, Elliot; Jarjour, Nizar N.; Levy, Bruce D.; Mauger, David; Meyers, Deborah A.; Moore, Wendy C.; Peters, Michael; Phillips, Brenda R.; Phipatanakul, Wanda; Sorkness, Ronald L.; Wenzel, Sally E.

In: American journal of respiratory and critical care medicine, Vol. 199, No. 11, 01.01.2019, p. 1358-1367.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Multiview cluster analysis identifies variable corticosteroid response phenotypes in severe asthma

AU - Wu, Wei

AU - Bang, Seojin

AU - Bleecker, Eugene R.

AU - Castro, Mario

AU - Denlinger, Loren

AU - Erzurum, Serpil C.

AU - Fahy, John V.

AU - Fitzpatrick, Anne M.

AU - Gaston, Benjamin M.

AU - Hastie, Annette T.

AU - Israel, Elliot

AU - Jarjour, Nizar N.

AU - Levy, Bruce D.

AU - Mauger, David

AU - Meyers, Deborah A.

AU - Moore, Wendy C.

AU - Peters, Michael

AU - Phillips, Brenda R.

AU - Phipatanakul, Wanda

AU - Sorkness, Ronald L.

AU - Wenzel, Sally E.

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Rationale: Corticosteroids (CSs) are the most effective asthma therapy, but responses are heterogeneous and systemic CSs lead to long-term side effects. Therefore, an improved understanding of the contributing factors in CS responses could enhance precision management. Although several factors have been associated with CS responsiveness, no integrated/cluster approach has yet been undertaken to identify differential CS responses. Objectives: To identify asthma subphenotypes with differential responses to CS treatment using an unsupervised multiview learning approach. Methods: Multiple-kernel k-means clustering was applied to 100 clinical, physiological, inflammatory, and demographic variables from 346 adult participants with asthma in the Severe Asthma Research Program with paired (before and 2-3 weeks after triamcinolone administration) sputum data. Machine-learning techniques were used to select the top baseline variables that predicted cluster assignment for a new patient. Measurements and Main Results: Multiple-kernel clustering revealed four clusters of individuals with asthma and different CS responses. Clusters 1 and 2 consisted of young, modestly CS-responsive individuals with allergic asthma and relatively normal lung function, separated by contrasting sputum neutrophil and macrophage percentages after CS treatment. The subjects in cluster 3 had late-onset asthma and low lung function, high baseline eosinophilia, and the greatest CS responsiveness. Cluster 4 consisted primarily of young, obese females with severe airflow limitation, little eosinophilic inflammation, and the least CS responsiveness. The top 12 baseline variables were identified, and the clusters were validated using an independent Severe Asthma Research Program test set. Conclusions: Our machine learning-based approaches provide new insights into the mechanisms of CS responsiveness in asthma, with the potential to improve disease treatment.

AB - Rationale: Corticosteroids (CSs) are the most effective asthma therapy, but responses are heterogeneous and systemic CSs lead to long-term side effects. Therefore, an improved understanding of the contributing factors in CS responses could enhance precision management. Although several factors have been associated with CS responsiveness, no integrated/cluster approach has yet been undertaken to identify differential CS responses. Objectives: To identify asthma subphenotypes with differential responses to CS treatment using an unsupervised multiview learning approach. Methods: Multiple-kernel k-means clustering was applied to 100 clinical, physiological, inflammatory, and demographic variables from 346 adult participants with asthma in the Severe Asthma Research Program with paired (before and 2-3 weeks after triamcinolone administration) sputum data. Machine-learning techniques were used to select the top baseline variables that predicted cluster assignment for a new patient. Measurements and Main Results: Multiple-kernel clustering revealed four clusters of individuals with asthma and different CS responses. Clusters 1 and 2 consisted of young, modestly CS-responsive individuals with allergic asthma and relatively normal lung function, separated by contrasting sputum neutrophil and macrophage percentages after CS treatment. The subjects in cluster 3 had late-onset asthma and low lung function, high baseline eosinophilia, and the greatest CS responsiveness. Cluster 4 consisted primarily of young, obese females with severe airflow limitation, little eosinophilic inflammation, and the least CS responsiveness. The top 12 baseline variables were identified, and the clusters were validated using an independent Severe Asthma Research Program test set. Conclusions: Our machine learning-based approaches provide new insights into the mechanisms of CS responsiveness in asthma, with the potential to improve disease treatment.

UR - http://www.scopus.com/inward/record.url?scp=85066631255&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85066631255&partnerID=8YFLogxK

U2 - 10.1164/rccm.201808-1543OC

DO - 10.1164/rccm.201808-1543OC

M3 - Article

VL - 199

SP - 1358

EP - 1367

JO - American Journal of Respiratory and Critical Care Medicine

JF - American Journal of Respiratory and Critical Care Medicine

SN - 1073-449X

IS - 11

ER -