Pediatric eosinophilic esophagitis

Abstract

Background and aims: Eosinophilic Esophagitis (EoE) is an immune-mediated antigentriggered inflammatory disease, strictly located to the esophagus. The pathogenesis is still partly unknown, and there is a lack of biomarkers for the disease. A foreign body impacted in the esophagus could be a sign of EoE. Our aim was to investigate if children previously diagnosed with a foreign body in the esophagus had a missed diagnosis of EoE, to investigate inflammatory responses by an ex vivo biopsy provocation-based method, stimulating biopsies with milk, wheat, and egg extracts and finally, study a variety of biomarkers in both blood and saliva.

Methods: Study I was a population-based longitudinal study, were all children diagnosed with a foreign body in the esophagus in Stockholm, Sweden 2006-2016, were identified. In addition to a review of medical files, each family was contacted (n=325) and asked standardized questions. Children with symptoms indicating EoE were offered esophagogastroduodenoscopy (EGD). Study II was an experimental study on esophageal biopsies from 26 children, half of whom had active EoE and the other half served as controls. The biopsies were placed in test tubes and were stimulated with food extracts from cow’s milk, egg, and wheat. Supernatants were collected before and after stimulation and analyzed for 45 different inflammatory markers. Special staining of biopsies was also performed. Study III was a cohort study with a longitudinal part. Blood, saliva, and anamnestic data were prospectively collected from 52 children with EoE and 53 healthy controls. The analyses were performed with enzyme linked immunosorbent assays (ELISA).

Results: Study I. In the 325 pediatric cases involving foreign bodies, 207 (64%) underwent EGD during the event, while the remaining cases experienced spontaneous expulsion of the foreign body. Among children with a previous foreign body, either spontaneously released or endoscopically removed, 12 (3.7%) were diagnosed with EoE. Characteristic of EoE patients were dysphagia, food impactions, excessive drinking during meals, and food allergies. Study II. Markers showing significant differences between patients with active EoE and controls included Granzyme B (GzmB), IL-1ra, and CXCL8 (p < 0.05). Levels of GzmB were elevated, while levels of IL-1ra were reduced in patients with active EoE compared to controls and those with EoE in remission, both before and after food extract stimulation. CXCL8 levels increased in active EoE compared to controls only after stimulation. Additionally, the number of histologically detected GzmB-positive cells was notably higher in patients with active EoE compared to controls and those in EoE remission (p < 0.05). Study III. Analysis from 105 children divided into active EoE, remission, and healthy, revealed elevated levels of the biomarkers AEC, EDN, 15(S)-HETE, sIgG4, and sIgE milk in active EoE compared to healthy individuals. A combination of the biomarkers AEC, EDN, sIgE to egg white and wheat plus symptoms, showed an AUC of 0.92 in distinguishing between the three groups. Further, optimal cutoff values for these biomarkers were calculated that could discriminate between active EoE and healthy with a positive predictive value (PPV) of 100% and a negative predictive value (NPV) of 75%. Longitudinally, levels of 3 EDN, sIgG4 to Bos 4, Bos 5, Bos 8, gliadin, and birch, and sIgE to milk decreased in patients progressing from active EoE to remission (p<0.05).

Conclusions: Study I. Children with a foreign body in the esophagus are at risk of having EoE. Biopsies should be taken during foreign body removal and questions about swallowing problems and allergic diseases should be carefully explored in children who do not need EGD because of spontaneous release. Study II. The presence of GzmB in the esophageal mucosa of children with active EoE suggests its potential involvement in the pathogenesis of the disorder. Study III. Novel biomarkers associated with EoE were identified, and a panel was proposed that together with symptoms, could discriminate between active EoE, EoE in remission, and healthy individuals. Cutoff values to distinguish between active EoE and healthy was also presented. The findings may contribute to a less invasive diagnostic method and may be a potential surveillance tool for pediatric EoE patients.