3508 ESOPHAGEAL CYTOLOGY BY ENDOSCOPIC BRUSHING IN THE DIAGNOSIS OF GASTRO-ESOPHAGEAL REFLUX DISEASE (GERD) NOT COMPLICATED BY ESOPHAGITIS.

Abstract

Aim: To look for cytologic criteria by esophageal brushing useful for the diagnosis of GERD not complicated by esophagitis. Patients and methods: An esophago-gastroduodenoscopy with esophageal brushing for cytologic analysis was carried out in 30 patients who did not present any typical or atypical symptom of GERD (GERD-) and in 54 patients who presented a typical symptomatology of GERD (GERD+). The esophagus was normal in all asymptomatic patients; it was normal or presented an anatomical abnormality of junction but without esophageal lesions in the 54 patients with typical GERD. Brushing was performed during esophagoscopy 2-3 cm above the squamous columnar junction using a protected brush. A spreading out was performed by applicating brush on a blade. The blade was immersed in pure alcohol, enamelled, and then red using the Papanicolaou stain by two cytologists not informed of the clinical diagnosis. Results: 5 types of cytologic abnormalities were found with a perfect agreement between the two cytologists: 1) dysacidophilia, 2) polymorphic inflammatory infiltrate, 3) cytolysis 4) epithelial regeneration, and 5) altered gastric cells. In group GERD+, 48 patients out of 54 had at least one of the 3 first cytologic abnormalities. Considering that esophageal cytology by brushing was abnormal when there was at least one of the 3 first abnormalities, the sensitivity and specificity of cytology for the diagnosis of GERD were excellent, respectively 89 and 87%. Conclusion: In patients having a GERD there is a "cytologic esophagitis" in the absence of any macroscopic lesion of the esophagus. Esophageal cytology performed by brushing is a simple means, inexpensive, reproducible, sensible and specific for the diagnosis of GERD. Making esophageal cytology systematically when there is no esophagitis during esophagoscopy performed in a patient having an atypical or recent symptomatology of GERD could avoid the need for a further pH-metry.