Disrupted mucosal vascular barrier in eosinophilic esophagitis

Abstract

Esophageal barrier has been investigated until now on the epithelial side only. Gut vascular barrier dysfunction has been recently implicated in a number of immune-mediated gastrointestinal disorders. We here characterized the esophageal vascular barrier (EVB) in eosinophilic esophagitis (EoE). Probe-based confocal laser endomicroscopy (pCLE) was performed in two EoE and two reflux esophagitis (RE) patients. The vascular barrier marker plasmalemma vesicle-1 (PV-1) was investigated as a measure of barrier disruption, by both immunohistochemistry and qPCR, in esophageal biopsies of 16 patients with EoE, 15 with RE, and 15 healthy controls (HC). In EoE, but not RE, pCLE revealed leakage of the EVB, which was restored in one patient after dupilumab treatment. PV-1 was significantly increased in EoE in comparison to RE and HC, both in terms of protein and transcript levels, supporting vascular leakage. EVB is disrupted in active EoE. Further studies are needed to understand the diagnostic and pathogenic implications of this finding.

Keywords: Confocal laser endomicroscopy; Dupilumab; Esophageal mucosa; Plasmalemma vesicle-1; Reflux esophagitis.

Fig. 1

Composite panel of images showing a representative case of reflux esophagitis (RE; A, B), one case of eosinophilic esophagitis (EoE) before (C, D) and after treatment with dupilumab (E, F), and another case with active EoE (G, H). Macroscopically, the endoscopy in this case showed reflux esophagitis Los Angeles grade A (A); at probe-based confocal laser endomicroscopy (pCLE) after fluorescein administration (B) no evident leakage is present, and the normal squamous epithelium is recognizable. In EoE, endoscopy showed some rings, furrows, and exudates in the middle esophagus (C, G); at pCLE, after fluorescein administration, leakage from small vessels is evident (red arrows), while the normal squamous mucosa is not recognizable (D, H). After a 3-month treatment with dupilumab, the macroscopic appearance of the esophagus is improved, with no visible rings, furrows, or exudates (E); at pCLE, after fluorescein administration, no leaky vessels were noticed (F). Scale bar = 20 μm.

Fig. 2

(A) Representative images (40x) of esophageal biopsy specimens stained with plasmalemma vesicle-1 (PV-1) from 16 patients with eosinophilic esophagitis (EoE), 15 patients with reflux esophagitis (RE), and 15 healthy controls (HC). In EoE (upper image), PV-1 staining was well evident in numerous vessels, poorly recognizable –but still evident– in RE, and completely absent in HC. (B) PV-1 staining score was significantly higher in EoE compared to both RE and HC, with no significant difference between RE and HC (C) PV-1 mRNA was significantly more expressed in EoE compared to both RE and HC, with no significant difference between RE and HC. Legend: *p < 0.01; **p < 0.001; ***p < 0.0001; ****p < 0.00001. GAPDH glyceraldehyde 3-phosphate dehydrogenase.