Abstract Background/Aims Eosinophilic granulomatosis with polyangiitis (EGPA) is a rare anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis characterised by asthma, eosinophilic inflammation, and multisystem involvement. Despite asthma being present in over 95% of cases, respiratory manifestations remain poorly defined limiting disease phenotyping. This study aimed to evaluate the clinical significance of biomarkers of type-2 (T2) inflammation (i.e., blood eosinophil count BEC, fractional exhaled nitric oxide FeNO and immunoglobulin E IgE) on disease activity; investigate the impact of ANCA status on various aspects of EGPA, including additional biomarkers (i.e., C-reactive protein CRP and immunoglobulin G IgG), lung function and high-resolution computed tomography (HRCT) imaging; and identify distinct respiratory phenotypes that form the basis of a model for EGPA disease subclassification. Methods Retrospective observational single-centre analysis of 52 EGPA patients reviewed at Hammersmith Hospital between January 2011 and January 2025 was performed. Variables measured contemporaneously with patient symptom questionnaires at least 18 months post-remission induction included T2 biomarkers, ANCA status at diagnosis, spirometry, and HRCT data. Following multiple imputations for missing data, factor analysis was conducted to identify latent variables and factors to use in a two-step cluster analysis. From this multivariate analysis, respiratory phenotypes were derived. Differences between clusters, particularly in relation to lung function and imaging, were identified using appropriate statistical tests. Statistical analysis was performed in IBM SPSS Statistics V29.0. Results Clustering successfully identified three distinct phenotypes. Cluster 1 (22/52) was T2-inflammatory predominant, ANCA-positive with severe disease onset; cluster 2 (10/52) was upper-airway predominant, 40% ANCA-positive with non-severe onset; cluster 3 (20/52) exhibited T2 inflammation with severe disease onset but was ANCA-negative. Clusters differed significantly in spirometry and HRCT imaging. Cluster 1 demonstrated findings consistent with inflammatory airway disease, cluster 2 with small vessel disease such as pulmonary vasculitis and cluster 3 features of small airway disease with microvascular alveolar surface damage. Limited significant differences in T2 biomarkers stratified by ANCA status were observed. However, IgE demonstrated trends in line with previous findings with higher levels in ANCA-positive patients. Conclusion This study suggests that heterogeneity of EGPA phenotypes cannot be explained by ANCA status alone. A combination of factors including airway involvement, additional biomarkers, and structural imaging changes contributes to distinct disease profiles. Importantly, HRCT imaging revealed structural differences across clusters, as a marker of damage, underscoring the importance of this imaging modality in the comprehensive assessment of an individual with EGPA and suggesting radiomics as a promising tool to further facilitate disease phenotyping. Identification of these phenotypes provides a framework for further research of the respiratory manifestations of EGPA, paves the way to clearer understanding of disease pathophysiology, and emphasises the need for improved biomarker identification and selection to better characterise under-recognised subgroups, overall highlighting the potential for future phenotype-directed therapeutic strategies in EGPA. Disclosure A. Faisal: None. H. Wilson-Morkeh: None. S. Siddiqui: Honoraria; S.S. has received speaker fees from AstraZeneca, GSK, Chiesi, Areteia Therapeutics, Sanofi and Medscape. Grants/research support; S.S. has received grants from AstraZeneca. Other; S.S. has provided advisory services to AstraZeneca, GSK, Chiesi, Areteia Therapeutics, Sanofi and Medscape.
Faisal et al. (Wed,) studied this question.