Barrett's esophagus: Macroscopic markers and the prediction of dysplasia and adenocarcinoma

Background and Aims: Surveillance endoscopy has been advocated for patients with Barrett's esophagus but the cost-effectiveness of this has been questioned. The aim of this study is to identify an optimum surveillance protocol by examining if macroscopic markers at diagnosis predict the development of dysplasia. Methods: The sample was 353 patients with Barrett's esophagus undergoing surveillance by a community-based group of gastroenterologists between 1981 and 2001. At diagnosis the presence of macroscopic and microscopic markers was noted. The presence and pattern of dysplasia and development of adenocarcinoma was documented during subsequent surveillance. Results: Three hundred and fifty-three patients (71% male) underwent regular surveillance over 19056 patient-months (median 42 months), having a median number of three surveillance endoscopies (range 1-40). Nine patients (seven male) developed adenocarcinoma (1/176 patient years) and four male patients developed high-grade dysplasia (1/397 patient years). Twelve of these 13 patients entered with one or more macroscopic markers: severe esophagitis, nodularity, Barrett's ulcer or stricture. Dysplasia risk was associated with macroscopic markers. Patients who entered with one marker were 6.7 times more likely to develop high-grade dysplasia/adenocarcinoma (HR = 6.7, 95% CI = 1.3, 35). Patients who entered with two or more markers were 14 times more likely to develop high-grade dysplasia/adenocarcinoma (HR = 14.1, 95% CI = 2.02, 102). Conclusions: The presence of severe esophagitis, Barrett's ulcer, nodularity or stricture at entry indicates a high-risk group for Barrett's esophagus. Cost-effectiveness of surveillance for these patients and those with dysplasia at entry would thus improve.