Barrett's esophagus: a discrepancy between macroscopic and histological diagnosis

The diagnosis of Barrett's esophagus at present requires endoscopic and histological confirmation of specialized intestinal metaplasia. This study prospectively analyzed the endoscopic and histological prevalence of Barrett's esophagus and the risk factors for the presence of Barrett's esophagus among patients being treated in an endoscopy unit.

A total of 474 unselected patients (58% men; mean age 52 y) were included in the study. Two biopsy specimens each were taken from below and above the squamocolumnar junction and from the antrum and gastric body. Four-quadrant biopsies were taken every 1-2 cm to confirm a macroscopic suspicion of Barrett's esophagus.

Barrett's esophagus was suspected at endoscopy in 109 patients (23%). Of the 109 patients with endoscopically suspected Barrett's esophagus, only 46 (42%) had the finding confirmed histologically. The sensitivity and specificity for the endoscopic diagnosis of Barrett's esophagus were 62% and 84%, respectively. A multivariate logistic regression analysis identified age (P = 0.0001; odds ratio per life-year 1.087; 95% CI, 1.046-1.139), male sex (P = 0.0020; OR 6.346; 95% CI, 2.094-22.314), and the number of biopsies (P = 0.0025; OR 1.661; 95% CI, 1.247-2.392) as factors associated with evidence of intestinal metaplasia on biopsy.

The striking discrepancy between the endoscopic findings and the histological diagnosis may be due to the focal distribution of intestinal metaplasia. This emphasizes the importance of an adequate biopsy protocol. In addition, better methods of detecting focal islands of intestinal metaplasia that are not visible at conventional endoscopy are needed.