Eosinophilic oesophagitis endotype classification by molecular, clinical, and histopathological analyses: a cross-sectional study

Background: Eosinophilic oesophagitis is understood in terms of quantifiable histological, endoscopic, and molecular features. Data are scant for inter-relations of these features and their potential to identify distinct disease endotypes. We aimed to identify clinical–pathological correlations between endoscopic and histological disease variables by transcription profiling of the oesophagus of patients with eosinophilic oesophagitis of varying severity and disease activity states. Methods: We did a cross-sectional study across ten hospital sites in the USA associated with the Consortium of Eosinophilic Gastrointestinal Disease Researchers. We analysed oesophageal biopsy specimens taken from paediatric and adult patients with eosinophilic oesophagitis (discovery cohort), using the eosinophilic oesophagitis diagnostic panel (EDP), a set of 96 informative transcripts. Histological and endoscopic features were assessed by quantification of oesophageal eosinophils and use of the eosinophilic oesophagitis histology scoring system (HSS) and the eosinophilic oesophagitis endoscopic reference score (EREFS). Associations among the various histological, endoscopic, and molecular features were analysed by Spearman correlation. Results were replicated in a biologically independent, single-centre, validation cohort of patients with active eosinophilic oesophagitis. Findings: The discovery cohort contained 185 samples and the validation cohort comprised 100 specimens. In the discovery cohort, EDP showed intersite consistency, significant correlation with oesophageal eosinophils (p<0·0001), and similar findings between paediatric and adult patients. Of eight HSS domains, basal zone hyperplasia correlated with the EDP (median Spearman ρ 0·47 [IQR 0·36–0·60]). Of five EREFS features, distal furrows correlated with the EDP (median Spearman ρ 0·42 [0·32–0·50]). By analysing active eosinophilic oesophagitis in the discovery cohort, the EDP identified three clusters associated with distinct endotypes (termed EoEe1–3) despite similar eosinophil levels. EoEe1 was associated with a normal-appearing oesophagus (risk ratio [RR] 3·27, 95% CI 1·04–10·27; p=0·0443), an inverse association with a history of oesophageal dilation (0·27, 0·09–0·82; p=0·0105) and showed relatively mild histological, endoscopic, and molecular changes. EoEe2 showed an inflammatory and steroid-refractory phenotype (RR 2·77, 95% CI 1·11–6·95; p=0·0376) and had the highest expression of inflammatory cytokines and steroid-responding genes. EoEe3 was associated with a narrow-calibre oesophagus (RR 7·98, 95% CI 1·84–34·64; p=0·0013) and adult onset (2·22, 1·19–4·12; p=0·0155), and showed the highest degree of endoscopic and histological severity and the lowest expression of epithelial differentiation genes. These endotypes were replicated in the validation cohort by clustering and with an eosinophilic oesophagitis endotype-prediction algorithm. Interpretation: Our new disease classification stratifies patients with eosinophilic oesophagitis into subgroups with potential clinical and therapeutic significance and provides a framework for a precision medicine approach to eosinophilic oesophagitis. Funding: National Institutes of Health.