What is the frequency of finding Helicobacter pylori in biopsies of patients with Barrett's esophagus? On which type of Barrett's metaplasia is it predominantly found? Could Helicobacter pylori be considered as a causative factor in Barrett's esophagus?
B.T. Cooper (Birmingham)
Helicobacter pylori has been linked to various disorders of the upper gastrointestinal tract particularly chronic gastritis, peptic ulceration and MALT associated lymphoma. The bacterium is especially adapted to survive on gastric epithelium and seem to be confined to it. There has been interest in the presence of H. pylori on heterotopic gastric epithelium and it has been found on islands of gastric epithelium within the duodenum and upper esophagus and in Meckel's diverticulum. The columnar epithelium of Barrett's esophagus (BE) is another tissue containing gastric type epithelium which might be colonized by
H. pylori and its presence would raise question as to whether it had a role in the causation or progression of Barrett's esophagus or in the development of complications such as esophageal ulcer or adenocarcinoma.
Since the first report of H. pylori on Barrett's epithelium , there have been 12 studies [2-13] reporting the prevalence of H. pylori in Barrett's esophagus (Table I); six of the studies have been prospective [7-11, 13].
The 12 studies involved 575 patients with Barrett's esophagus which was diagnosed by conventional endoscopic criteria. One hundred and fifty-eight patients (27.5%) had H. pylori identified in biopsies taken from the Barrett's esophagus segment. H. pylori was identified using histological techniques and/or culture in all but one study which used immuno-chemistry . However there was considerable variation in the prevalence of esophageal H. pylori between studies (from 0 to 62%) which may be explained in part by the bacterium's patchy distribution on Barrett's epithelium [6, 11]. In 5 studies, the data allowed relation of the H. pylori to a particular type of epithelium within the Barrett's esophagus segment [3, 9, 10, 12, 13]. In biopsies showing gastric fundal and/or junctional epithelium taken from 216 patients, 82 patients (38%) were H. pylori positive. In biopsies showing specialized columnar (intestinal metaplastic) epithelium taken from 155 patients, only 12 patients (7.7%) were positive. In 3 of the 5 studies, H. pylori was not identified in any biopsies taken from specialized columnar mucosa [3, 9, 13]. The bacterium was found to be adherent to gastric type mucous cells, in mucus overlying the Barrett's mucosa  and in association with MALT in Barrett's esophagus . Biopsies taken from the squamous epithelium just above the squamo-columnar junction in patients with Barrett's esophagus have been negative for H. pylori [5, 6, 9, 11]. Two studies used patients with reflux esophagitis as controls and none of these patients' distal esophageal biopsies were H. pylori positive [9, 11].
Five studies have addressed the issue of whether the antral mucosa was colonized with H. pylori in patients with Barrett's esophagus [5, 7, 9, 11, 13]. A total of 280 patients were studied and 99 patients (35%) had H. pylori present on antral mucosa. Two of the studies looked at the antral mucosa of control patients with reflux esophagitis and found a similar prevalence [9, 11]. These prevalences for H. pylori in antral mucosa of patients with Barrett's esophagus and reflux esophagitis are similar to those that would be expected from normal individuals of similar age within these populations. There is some evidence from analysis of the literature that patients with Barrett's esophagus who have H. pylori on their Barrett's mucosa are likely to have H. pylori positive antral gastritis but Barrett's esophagus patients with H. pylori positive antral gastritis do not necessarily have H. pylori in their Barrett's mucosa [5, 9]. This raises the possibility that the presence of H. pylori in the Barrett's esophagus segment might be related to gastroesophageal reflux. Thus patients with H. pylori in their stomachs and who have gastroesophageal reflux may get
H. pylori in their esophagus provided they have an appropriate mucosa in their esophagus on which the H. pylori can flourish i.e. Barrett's epithelium. It has been suggested that
H. pylori in the esophagus could be the result of contamination caused by careless biopsy technique  but other evidence does not support this .
There have been no specific studies to see whether the presence of H. pylori is associated with more rapid progression of the BE, although one small preliminary study suggested that colonization of the gastric fundus was associated with more severe reflux esophagitis and with Barrett's esophagus . More data is needed on this. There is no correlation between H. pylori in the esophagus and the length of the Barrett's segment , or the presence of hiatal hernia and/or associated reflux esophagitis . There is little data in the literature about H. pylori and complications but the few reports on esophageal adenocarcinoma [6, 9] and esophageal ulceration [4, 5] are conflicting. There is some evidence to suggest that H. pylori does little damage to Barrett's epithelium; the bacterium is often scantily and/or patchily distributed [6, 11] and some workers report very little inflammatory response to it . When inflammation is found, it is no more common in H. pylori positive Barrett's mucosa [10, 11, 13]. Recent preliminary data suggested an association between H. pylori and both dysplasia and p53 over-expression in Barrett's mucosa  but clearly much more work is needed. However to date, it seems unlikely that H. pylori is a major factor in either the development, progression or the complications of Barrett's esophagus.
In conclusion, H. pylori has been found in the metaplastic epithelium of about a quarter ot patients with Barrett's esophagus, but mainly on gastric fundic/junctional rather than specialized columnar epithelium in Barrett's esophagus. The bacterium is mostly found in patients who also have it in their gastric antral mucosa, suggesting that the mechanism of esophageal colonization is gastroesophageal reflux of the organism. There is little evidence to date that H. pylori has a role in the development or progression of Barrett's esophagus and its complications. It is not known whether eradication of H. pylori is of any benefit in patients with Barrett's esophagus although there is no evidence to suggest that it should be. Until there is good evidence to the contrary, H. pylori eradication cannot be recommended in patients with Barrett's esophagus.
1. Goodwin CS, Bliscow ED, Warren JR, Waters TE, Sanderson CR, Easton L. Evaluation of cultural techniques for isolating C. pyloridis from endoscopic biopsies of gastric mucosa. J Clin Pathol 1985;38:1127-1131.
13. Ricaute O, Fléjou JF, Vissuzaine C, Goldfain D, Rotenberg A, Cadiot G, Potet F. Helicobacter pylori infection in patients with Barrett's esophagus: a prospective immunohistochemical study. J Clin Pathol 1996;49:176-177.
14. Weston AP, Campbell DR, Horvat RT, Cherian R, Makdisi WF, Banerjee SK, Dixon A, McGregor D. MALT in Barrett's esophagus: prospective evaluation and association with gastric MALT, MALToma and H. pylori. Gastroenterology 1996:110:A613.
17. Young MA, Kim R, Clark MR, Van Bibber MM, Safatle-Ribeiro AV, Ribeiro U, Reynolds JC. The incidence of premalignant markers of esophageal cancer is increased in Barrett's patients with Helicobacter pylori. Gastroenterology 1996;110:A618.