Minimal change esophagitis: prospective comparison of endoscopic and histological markers between patients with non-erosive reflux disease and normal controls using magnifying endoscopy

INTRODUCTION: More than half the patients with gastroesophageal reflux disease (GERD) show no endoscopic abnormality or minimal change esophagitis (non-erosive reflux disease, NERD). We investigated the value of endoscopic and histological markers for the prediction of NERD before and after treatment with 20 mg esomeprazole. METHODS: Between July and October 2002, consecutive patients presenting for upper endoscopy were stratified into GERD and non-reflux patients (control group) with the help of a questionnaire. The endoscopist was blind to the presence of reflux symptoms. Using magnifying endoscopes minimal change esophagitis was defined by the presence of vascular injection or vascular spots above the Z-line, villous mucosal surface and islands of squamous cell epithelium below the Z-line. Targeted and random biopsies were taken below and above the Z-line. Patients with endoscopically visible classical signs of esophagitis (Los Angeles A-D) or histologically proven Barrett's esophagus were not further investigated in the study (drop out). The esophageal specimens were histologically evaluated for erosions, infiltration with leukocytes, hyperplasia of basal cells and length of papillae. Patients with NERD were treated with 20 mg esomeprazole/day for 4 weeks and reevaluated by endoscopy as described before. RESULTS: 39 patients with heartburn and 39 patients without reflux symptoms (controls) were finally included in the analysis (per protocol). Patients with NERD significantly (p = 0.005) more often showed endoscopic signs of minimal change esophagitis (27/39) than the control group (8/39). An increased length of papillae (14/39 versus 2/39; p = 0.005) and basal cell hyperplasia (17/39 versus 4/39; p = 0.009) were significantly more common in the heartburn group. After treatment with esomeprazole, no significant endoscopic or histological differences between the NERD and control group could be observed. CONCLUSIONS: Minimal change esophagitis can be seen with high resolution magnifying endoscopy. By combining endoscopic and histological markers NERD can be predicted with a sensitivity of 62% and a specificity of 74%. Treatment with esomeprazole for 4 weeks reverses the slight alterations to normal