S119 Mapping mouse models of severe asthma onto human disease. (15th November 2017)

• Record Type: Journal Article
• Authors: Kazi, K; Adcock, IM; Pavlidis, S
• Is Part Of: Thorax. Volume 72(2017)Supplement 3
• Page Start: A72
• Page End: A72
• Publication Date: 2017-11-15
• Subjects: Chest -- Diseases -- Periodicals; Thorax; Chest -- Diseases; Periodicals; Periodicals; 617.54
• DOI: 10.1136/thoraxjnl-2017-210983.125
• Languages: English
• ISSNs: 0040-6376

Abstract

Abstract : Introduction: Severe asthma represents a significant unmet need in terms of therapeutics. Drug development in asthma has been slow and expensive and one of the reasons for this is that positive findings for drugs tested in preclinical models have not readily translated into successful therapies in man. Aims and Objectives: We sought to improve the predictive power of existing models of asthma by using novel bioinformatics techniques to align these models with subsets of human asthma. Methods: We applied differential gene expression analysis to transcriptomic data from whole lung samples of 6 murine models of asthma and oxidative stress to produce gene signatures that represented each model. These signatures were then used to calculate enrichment scores (ESs) for transcriptomic data from bronchial biopsies taken from 81 asthmatic and 26 healthy subjects from the U -BIOPRED cohort using gene set variation analysis. These ESs were taken as an index of similarity between each model and each patient and were used to drive further analyses using topological data analysis and goodness of fit modelling. Results: We found that no single mouse model was aligned well with all asthmatics. We identified three clusters of patients who were represented to varying degrees by different mouse models and who displayed clinical features that aligned well with phenotypes of asthmatics identified previously by clustering analyses based upon clinical features and biological markers. Patients in cluster χ were defined by neutrophilic sputum, later onset of disease, higher incidence of sinusitis, more frequent exacerbations and more airflow limitation. Patients in cluster Y1 showed significantly lower sputum neutrophils, a trend towards higher sputum eosinophils, a significantly later onset of airways disease and a trend towards higher BMI. Patients in cluster Y2 showed a significantly higher percentage of neutrophils in the blood, a trend towards increased sputum lymphocytes and were more likely to identify aspirin as a trigger (p=0.06). Conclusion: Our evidence supports the assertion that it is possible, on a transcriptional level, to align mouse models of asthma to subsets of human asthma and that doing so may have significant implications for the expedience of drug development in asthma.