This prospective study followed 170 patients enrolled over the past 13 years. All enrolled patients were diagnosed with BE on the basis of two endoscopies 6 months apart showing apparent columnar epithelium on endoscopy and intestinal metaplasia on histopathology. Of 170 patients, 167 were male and 3 were female; the mean age was 62 years (range 30-85 yrs). These patients have been followed for a mean time of 57 months (4.8 yrs; range 6-156 months). The mean length of BE was 5.2 cm (range less than 1 to 16 cm); 42/170 patients (25%) have short segment BE, defined as less than 3 cm in length. Race was not noted, nor was frequency of endoscopic follow-up.
Seven of 170 patients were found to have adenocarcinoma of the esophagus at the initial or the 6-month endoscopic examination. Therefore, the prevalence of esophageal adenocarcinoma in this study is 4%.
Adenocarcinoma of the esophagus developed in four patients over the course of the study and was diagnosed a mean of 27 months after study entry (range 8-53 months). One of these patients had short segment BE; all had high-grade dysplasia. The average age of affected patients was 60 years; 3/4 were white males (sex and race of the fourth patient were not provided). Two patients had an esophagectomy and are doing well several years later. Two patients were not considered to be surgical candidates and have died. The prevalence of adenocarcinoma in this study exceeded the incidence.
The incidence of adenocarcinoma of the esophagus in this study is 1/208 patients years, which is lower than that in previous prospective studies, for which the pooled incidence is 1/107 patient years. The current study is larger with longer follow-up time than these previously published prospective studies.
There is potential for error in this study because few patients with cancer were enrolled, as illustrated by the fact that one less patient diagnosed with adenocarcinoma would have yielded an incidence of 1/278 patients years, whereas one more patient with adenocarcinoma would have yielded an incidence of 1/167 patient years. Further, large- scale multicenter studies are needed to better define the incidence of esophageal adenocarcinoma in patients with BE in the United States.
The study confirms previous information that high grade dysplasia and adenocarcinoma may coexist. The finding of adenocarcinoma in short-segment Barrett's suggests these patients should be entered into careful separate screening programs.
ACG Highlights Home Page