Endoscopy with random biopsies is the standard technique for the diagnosis of Barrett's esophagus (BE). We studied the usefulness of endoscopic biopsies in the diagnosis and surveillance of patients with BE.
We reviewed all patients with BE controlled at the Hospital Cl铆nic from February 2002 to March 2010.
During the study period, 493 endoscopies were performed in 206 patients. We excluded 86 patients who had undergone one endoscopy only, hence the final sample consisted of 117 patients (mean age 61 [12] years, 73%men) with 407 endoscopies and a mean follow-up of 45 (38) months. Fifty-seven patients had a short-BE (49%), 28 a long-BE (24%) and 32 (27%) could not be inferred from the report. In 25 cases (21%), intestinal metaplasia (IM) was not detected at the endoscopic index. During follow-up, biopsies were negative for IM at some point in 45 patients (38.4%) and there was a change in the histologic grade in 100 cases (24.6%). The presence of a short-BE (43%in short-BE vs 7%in long-BE, P = .001) and few biopsies (3 [1.5] vs 6.2 [4.5]; P = .005) were associated with a higher frequency of negative biopsies for IM.
Random endoscopic biopsies are not a good tool for diagnosis and surveillance of patients with BE.