Eosinophilic Esophagitis: Relationship of Subepithelial Eosinophilic I

OBJECTIVES: For technical reasons, the histologic characterization of eosinophilic esophagitis (EoE)-specific alterations is almost exclusively based on those found in the esophageal epithelium, whereas little is known about subepithelial abnormalities. In this study, we aimed to systematically assess the nature of subepithelial histologic alterations, and analyze their relationship with epithelial histologic findings, endoscopic features, and symptoms. METHODS: Adult patients with established EoE diagnosis were prospectively included during a yearly follow-up visit. Patients underwent assessment of clinical, endoscopic, and histologic disease activity using EoE-specific scores. RESULTS: We included 200 EoE patients (mean age 43.5 +/- 15.7 years, 74% males) with a median peak count of 36 intraepithelial eosinophils/hpf (IQR 14-84). The following histologic features were identified in the subepithelial layer: eosinophilic infiltration (median peak count of 20 eosinophils/hpf (IQR 10-51)), eosinophil degranulation (43%), fibrosis (82%), and lymphoid follicles (56%). Peak intraepithelial eosinophil counts were higher, identical, and lower when compared to the subepithelial layer in 62.5%, 7%, and 30.5% of patients, respectively. Anti-eosinophilic treatment at inclusion did not influence the relation between subepithelial and epithelial peak eosinophil counts. Subepithelial histologic activity correlated with epithelial histologic activity (rho 0.331, P < 0.001), endoscopic severity (rho 0.208, P = 0.003), and symptom severity (rho 0.179, P = 0.011). Forty percent (21/52) of patients with < 15 intraepithelial eosinophils/hpf had subepithelial peak counts of = 15/hpf. CONCLUSIONS: There is a significant but modest correlation between subepithelial histologic activity and epithelial histologic activity, endoscopic severity, and symptom severity. The long-term clinical impact of assessing subepithelial alterations in EoE needs to be further elucidated.