Endoscopic findings in patients with Schatzki rings: evidence for an association with eosinophilic esophagitis

Aim: To investigate endoscopic findings in patients with Schatzki rings (SRs) with a focus on evidence for eosinophilic esophagitis (EoE).

Methods: We consecutively approached all adult patients scheduled for elective outpatient upper endoscopy for a variety of indications at the German Diagnostic Clinic, Wiesbaden, Germany between July 2007 and July 2010. All patients with endoscopically diagnosed SRs, defined as thin, symmetrical, mucosal structures located at the esophagogastric junction, were prospectively registered. Additional endoscopic findings, clinical information and histopathological findings with a focus on esophageal eosinophilia (≥ 20 eosinophils/high power field) were recorded. The criteria for active EoE were defined as: (1) eosinophilic tissue infiltration ≥ 20 eosinophils/hpf; (2) symptoms of esophageal dysfunction; and (3) exclusion of other causes of esophageal eosinophilia. Gastroesophageal reflux disease was excluded by proton pump inhibitor treatment prior to endoscopy. The presence of ≥ 20 eosinophils/hpf in esophageal biopsies in patients that did not fulfil the criteria of EoE was defined as esophageal hypereosinophilia.

Results: A SR was diagnosed in 171 (3.3%; 128 males, 43 females, mean age 66 ± 12.9 years) of the 5163 patients that underwent upper gastrointestinal-endoscopy. Twenty of the 116 patients (17%) from whom esophageal biopsies were obtained showed histological hypereosinophilia (≥ 20 eosinophils/hpf). Nine of these patients (8 males, 1 female, mean age 49 ± 10 years) did not fulfill all diagnostic criteria of EoE, whereas in 11 (9%) patients with ≥ 20 eosinophils/hpf, a definite diagnosis of EoE was made. Three of the 11 patients (27%) with definite EoE had no suspicious endoscopic features of EoE. In contrast, in the 25 patients in whom EoE was suspected by endoscopic features, EoE was only confirmed in 7 (28%) patients. Patients with EoE were younger (mean age 41.5 ± 6.5 vs 50.5 ± 11.5 years, P = 0.012), were more likely to have a history of allergies (73% vs 29%, P = 0.007) and complained more often of dysphagia (91% vs 34%, P = 0.004) and food impaction (36% vs 6%, P = 0.007) than patients without EoE. Endoscopically, additional webs were found significantly more often in patients with EoE than in patients without EoE (36% vs 11%, P = 0.04). Furthermore, the SR had a tendency to be narrower in patients with EoE than in those without EoE (36% vs 18%, P = 0.22). The percentage of males (73% vs 72%, P = 1.0) and frequency of heartburn (27% vs 27%, P = 1.0) were not significantly different in both groups. The 9 patients with esophageal hypereosinophilia that did not fulfil the diagnostic criteria of EoE were younger (mean age 49 ± 10 years vs 58 ± 6 years, P = 0.0008) and were more likely to have a history of allergies (78% vs 24%, P = 0.003) than patients with < 20 eosinophils/hpf. Predictors of EoE were younger age, presence of dysphagia or food impaction and a history of allergies.

Conclusion: A significant proportion of patients with SRs also have EoE, which may not always be suspected according to other endoscopic features.