Article Text

Consideration of histopathological subtypes and biopsy techniques in Barrett’s oesophagus surveillance programmes
  1. Havering Hospitals NHS Trust,
  2. Gubbins Lane, Romford,
  3. Essex RM3 OBE, UK
  2. A C WICKS,
  1. University Division of Gastroenterology,
  2. Leicester General Hospital,
  3. Gwendolen Road,
  4. Leicester LE5 4PW, UK

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Editor,—In their recent study Macdonald et al (Gut 1997;41:303–7) concluded that as only 1 in 143 patients was identified as having carcinoma of the oesophagus as a result of surveillance, the policy should be reconsidered. This conclusion raises several questions: Which patients should undergo surveillance and does the histopathological subtype matter? In this study patients eligible for the endoscopy surveillance programme had to be under the age of 70, without other serious coexisting disease. The endoscopic Barrett’s segment was >3 cm in all cases and patients with all histological subtypes (gastric 78% and specialised intestinal metaplasia 18%) were included. The definition of Barrett’s oesophagus has a long and controversial history.1 Specialised intestinal metaplasia is increasingly recognised as being the most common histological subtype,2 ,3 and patients with this epithelial type are at high risk of developing caner.4 ,5On the basis of this, many investigators now use the term Barrett’s oesophagus to indicate the presence of specialised intestinal metaplasia.6 Furthermore, a recently proposed re-classification of Barrett’s oesophagus suggests that patients with specialised intestinal metaplasia extending beyond the gastro-oesophageal junction are the only group in which endoscopic surveillance is currently warranted.1 How many biopsy specimens should be taken? In the Leicester study specimens were typically taken from all four quadrants at the mid-point of the involved mucosa. However, given the heterogeneous nature of the Barrett’s epithelium,6-9 it may be necessary to take samples in every quadrant every 2 cm along the length of the Barrett’s epithelium in order to detect areas of high grade dysplasia and carcinoma in situ. This is the regimen currently recommended by the Working Party Report to the World Congress of Gastroenterology.10

Perhaps if Barrett’s surveillance programmes concentrated on patients with specialised intestinal metaplasia, and involved the collection of multiple biopsy samples, this would improve the yield and cost effectiveness of such programmes? Hopefully in the future, our collective experiences of Barrett’s surveillance and a clearer understanding of the pathophysiology of Barrett’s oesophagus, might enable a consensus to be reached.



Editor,—We are grateful to Drs Fitzgerald and Burnham for their comments on our paper. It is important to remember that our report is a retrospective study encompassing our clinical practice over the past 10 years. As discussed in the paper, the definition of what constitutes Barrett’s oesophagus has continued to alter over this period, the inclusion of short segment Barrett’s being an additional complicating factor. We agree that many workers in the area now consider that the presence of intestinal metaplasia is an absolute requirement for a diagnosis of Barrett’s oesophagus to be made and that intestinal metaplasia constitutes a risk factor for the development of carcinoma.

In our paper, we quoted the criteria suggested by the World Health Organisation for screening for any disease.1-1 As there is little evidence that these criteria have been met for screening of Barrett’s oesophagus, particularly with regard to altering the outcome of patients,1-2 caution must be shown in interpreting the guidelines produced at the World Congress of Gastroenterology1-3 as a “gold standard”. We agree that performing four-quadrant biopsies every 2 cm is likely to pick up additional areas of mucosal abnormality. It must be remembered, however, that in our study the average length of the Barrett’s mucosa was 7 cm and would therefore require an additional eight biopsy samples per patient. This has major implications for providing a screening service as it will increase the time taken to perform the procedure, thereby reducing the number of procedures that can be performed on an endoscopy list. It is also possible that this would result in a reduction in the patient’s willingness to undergo further screening. In addition, the histological costs would be increased by about £60 per patient per endoscopy.

There is general agreement that large multicentre studies are required for a conclusive assessment of the value of screening patients with Barrett’s mucosa to be reached. Focusing on subgroups, such as those with co-existing strictures, may prove to be the only cost effective approach.


  1. 1-1.
  2. 1-2.
  3. 1-3.