Validation of a Pathological Score for the Assessment of Bronchial Biopsies in Severe Uncontrolled Asthma: Beyond Blood Eosinophils

Background: Blood eosinophil count (BEC) is currently used as a surrogate marker of T2 inflammation in severe asthma but its relationship with tissue T2-related changes is elusive. Bronchial biopsy could add reliable information but lacks standardization. Objectives: To validate a systematic assessment of the bronchial biopsy for the evaluation of severe uncontrolled asthma (SUA) by standardizing a pathological score.Methods: A systematic assessment of submucosal inflammation, tissue eosinophilic count/field (TEC), goblet cells hyperplasia, epithelial changes, basement membrane thickening, prominent airway smooth muscle and submucosal mucous glands was initially agreed and validated in representative bronchial biopsies of 12 patients with SUA by 8 independent pathologists. In a second phase, 62 patients with SUA who were divided according to BEC & GE; 300 cells/mm3 or less underwent bronchoscopy with bronchial biopsies and the correlations between the pathological findings and the clinical characteristics were investigated.Results: The score yielded good agreement among pathologists regarding submucosal eosinophilia, TEC, goblet cells hyperplasia and mucosal glands (ICC = 0.85, 0.81, 0.85 and 0.87 respectively). There was a statistically significant correlation between BEC and TEC (r = 0.393, p = 0.005) that disappeared after correction by oral corticosteroids (OCS) use (r = 0.170, p = 0.307). However, there was statistically significant correlation between FeNO and TEC (r = 0.481, p = 0.006) that was maintained after correction to OCS use (r = 0.419, p = 0.021). 82.4% of low-BEC had submucosal eosinophilia, 50% of them moderate to severe.Conclusion: A standardized assessment of endobronchial biopsy is feasible and could be useful for a better phenotyping of SUA especially in those receiving OCS.& COPY; 2023 The Author(s). Published by Elsevier Espana, S.L.U. on behalf of SEPAR. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).