Marked Variability in the Endoscopic Diagnosis of Barrett'S Esophagus Using Japanese and Western Criteria

Chika Kusano,Tonya Kaltenbach,Roy M. Soetikno,Takuji Gotoda

DOI: https://doi.org/10.1016/j.gie.2009.03.1070

IF: 10.396

2009-01-01

Gastrointestinal Endoscopy

Abstract:Introduction: It is important to recognize and diagnose Barrett's esophagus (BE) with or without dysplasia, such that appropriate surveillance and interventions can be carried out. Recognition of the esophageal gastric junction (EGJ) is the basis for the diagnosis of BE. In Japan, the distal end of the lower esophageal palisade vessels (PV) is used to define the EGJ. However, in Western countries, the EGJ is defined as the proximal margin of the gastric folds (GF). We aimed to compare the endoscopic diagnosis of BE using the PV and the GF as a land mark of EGJ. Material and Method: We prospectively enrolled patients at a single US center who were referred for routine diagnostic esophagogastroduodenoendoscopy. Prior to the study initiation, a Japanese endoscopist formally demonstrated the endoscopic features of end of the palisade vessels and their detection technique. Six American endoscopists performed the endoscopic assessment, and recorded the location details of the PV, GF and BE. We assessed the recognition rates of the EGJ and the detection rates of BE, first by using the PV and later using the GF as a land mark of EGJ. Result: Over a six week period, we examined 82 patients. The mean patient age (SD) was 64.5 (14) years. The patients were primarily men (91%; n=73); with the majority (76%; n=62) being Caucasian and the minority (4%; n=3) Hispanic. Identification rate of EGJ was 87.8% (72/82) using PV and 97.5% (80/82) using GF. 28.0% (23/82) in cases were endocopically diagnosed as BE by PV (PV-BE cases) and 17.0% (14/82) by GF (GF-BE cases). The concordance rate for the detection of BE endoscopically was 68.3% (56/82). Reflux esophagitis was detected in 30.5% (25/82). We could not identify the EGJ due to esophagitis in 7 cases by using PV and in 2 cases using GF. BE (less than 1 cm) was identified in 73.9% (17/23 PV-BE cases) and 57.1% (8/14 GF-BE cases), respectively. Discussion: We showed different ratios in the endoscopic detection of BE based on the lack of a standard identification of EGJ. The difference in the prevalence rate of BE (especially short segment Barrett's esophagus (SSBE)) in Japan and Western countries may in part be attributed to differences in the endoscopic landmarks used to diagnose BE. As such, the actual risk for cancer in SSBE (especially less than 1 cm) based on Japanese criteria needs to be further investigated and clarified. A universal standard for identification of EGJ should be established.