Surveillance for Barrett’s oesophagus: if you do it, do it properly

To the Editor: Reflux oesophagitis is an increasingly common medical condition, and up to 10% of all patients with reflux oesophagitis have associated columnar-lined oesophagus, or Barrett’s oesophagus (Box 1).1 This is associated with an increased risk of adenocarcinoma in Australian data of 1 in 176 patient-years.2 Current surveillance guidelines for Barrett’s oesophagus recommend second-yearly endoscopy with quadrantic biopsies at 2–3 cm,3,4 although recent reviews have suggested that the time interval can be extended.1 No prospective study has demonstrated that screening for Barrett’s oesophagus improves survival in the screened population. All the evidence is based on retrospective reviews. If, however, screening is to be of benefit, then adequate tissue sampling is required, otherwise the screening test provides false reassurance and is a poor use of endoscopy resources.

We retrospectively audited the endoscopies performed for Barrett’s oesophagus (with intestinal metaplasia) surveillance at our institution over 5 years (1996–2001). In this period, 253 endoscopies were performed as surveillance procedures for Barrett’s oesophagus. We reviewed the endoscopy and histopathology reports to determine whether an adequate number of biopsies had been taken.

We found that quadrantic biopsies (defined as four biopsies per 2 cm or less) at every 2–3 cm were performed in 27 of 72 (38%) short-segment, 27 of 150 (25%) long-segment (3–10 cm), and 1 of 13 (8%) extensive (> 10 cm) Barrett’s oesophagus. An acceptable number of biopsies were taken from 40% of patients. The number of biopsies taken per centimetre of Barrett’s oesophagus was inversely proportional to the length of Barrett’s oesophagus.

The median interval between surveillance procedures was 12 months. In most surveillance procedures (137/217; 63%) the endoscopy was performed following medical review in the outpatient department rather than because of planned call-back, although in only a few cases did it appear to be due to alarm symptoms such as dysphagia or weight loss. There was little consistency in recommendations among the medical staff. Only one cancer was identified through surveillance in this period, with another presenting in a patient previously on the call-back system who had been lost to follow-up. Three high-grade dysplasias were found.

Our retrospective audit revealed that the endoscopists were not following biopsy guidelines, and revealed significant variances in practice. Previously, we found a similar picture with post-polypectomy surveillance; with re-education and development of a prospective review of all cases, we have been able to greatly improve this aspect of practice.5

We were surprised at the result of our audit, and invite other endoscopy units to perform a similar audit of their own practice. It has encouraged us to develop a process similar to the one we have adopted to improve post-polypectomy surveillance. This should improve our practice and enable more efficient utilisation of the endoscopy facility.

1: Barrett’s oesophagus