Statistics from Altmetric.com
Traditional teaching holds that the normal stomach has three types of mucosae. The oxyntic mucosa of the gastric body and fundus has glands with parietal cells that secrete acid, and chief cells that secrete digestive enzymes. The mucosa of the antrum is comprised of mucus-secreting cells and endocrine cells that produce gastrin, which regulates acid production by the oxyntic mucosa. Finally, the most proximal portion of the stomach (the gastric cardia), a region with ill-defined borders, allegedly is lined by ‘cardiac mucosa’ comprised almost exclusively of mucus-secreting cells. Cardiac mucosa has been assumed to function as a buffer zone, preventing the damage that might result if the acid-sensitive, squamous mucosa of the oesophagus joined directly with the acid-secreting oxyntic mucosa of the gastric body. This dogma went unchallenged until 1997, when Chandrasoma proposed that cardiac mucosa is not a normal structure, but rather is acquired when GORD causes columnar metaplasia of squamous epithelium in the distal oesophagus.1 Furthermore, he contended that cardiac mucosa evolves into the intestinal metaplasia of Barrett's oesophagus in the setting of persistent GORD. Recent studies by investigators in Glasgow provide support for Chandrasoma's proposals.
In an earlier investigation, the Glasgow group studied the gastro-oesophageal junction (GOJ) region of 51 healthy volunteers.2 No subject had abnormal acid reflux by conventional pH monitoring in which a pH electrode was positioned 5 cm above the lower oesophageal sphincter (LOS). Using pH electrodes positioned within the LOS, however, significant differences in intrasphincteric acid exposure levels were noted between subjects with small and large waist circumferences. Compared with the 24 subjects with small waists, the 27 with large waists had more proximal extension of gastric acid within the LOS, resulting in protracted acid exposure of the distal-most, squamous-lined oesophagus. Accompanying this intrasphincteric acid incursion, subjects with large waists had a significantly greater length of cardiac mucosa than subjects with small waists (2.5 mm vs 1.8 mm in biopsy specimens taken across the squamocolumnar junction), and the extent of cardiac mucosa correlated positively with abdominal fat and with age. In addition, all cardiac mucosa identified was inflamed, irrespective of the subject's waist circumference. The investigators concluded that central obesity is associated with extension of gastric acid into the LOS (ie, intrasphincteric reflux), and with lengthening of the cardiac mucosa that is likely due to metaplasia induced by intrasphincteric reflux.
In a report published in this issue of Gut, Derakhshan et al3 provide further support for the hypothesis that cardiac mucosa develops through reflux-induced metaplasia. Using gastric and oesophageal biopsy specimens from the same healthy volunteers described above, as well as from 15 patients with long-segment Barrett's oesophagus, the researchers scored inflammatory changes and performed immunohistochemical staining for markers of intestinal and gastric differentiation (CDX2, Villin, TFF-3, Li-Cadherin, MUC-1, MUC-2 and MUC-5AC). The intensity of inflammation in cardiac mucosa was similar to that in Barrett's metaplasia, and significantly greater than that in squamous, oxyntic and antral mucosae. Furthermore, the cardiac mucosa of the healthy volunteers exhibited an immunohistochemical staining pattern almost identical to that of non-intestinal type Barrett's metaplasia. Finally, the distal oesophageal squamous mucosa of some healthy volunteers expressed the columnar differentiation markers TFF-3 and Li-Cadherin, perhaps heralding the onset of columnar metaplasia, and the expression of these markers correlated with central obesity. All of these findings support, but do not prove the hypothesis that cardiac mucosa at the GOJ is metaplastic, even in apparently normal individuals.
A problem that confounds the interpretation of studies on metaplasia at the GOJ is the lack of a widely accepted, standardised terminology for key elements of the structures and processes involved. There are no clear anatomical limits for the gastric cardia; there can be substantial disagreement among pathologists on precisely what constitutes cardiac mucosa, and some even object to the very use of the term ‘cardiac mucosa’. Further muddying the waters is international disagreement on the requirement for intestinal metaplasia with goblet cells to establish a diagnosis of Barrett's oesophagus (required in the USA, not in the UK). Even when there is unequivocal oesophageal intestinal metaplasia, individual biopsies of the Barrett oesophagus can reveal a mosaic of intestinal, cardiac and oxyntic-type mucosae.4 Consequently, it is not entirely clear what mucosal type predominates in the non-intestinal-type Barrett's metaplasia specimens in Derakhshan's study, but it is most likely predominantly cardiac in type. The observation that cardiac mucosa in Barrett's oesophagus exhibits the same immunohistochemical differentiation markers as cardiac mucosa at the GOJ certainly is consistent with the hypothesis that both are metaplastic in origin. However, this observation is also consistent with the hypothesis that cardiac mucosa simply exhibits the same immunohistochemical signature irrespective of how and where it develops.
The term ‘metaplasia’ itself is not standardised. Metaplasia often has been defined as the process in which one type of adult cell replaces another. However, metaplasia might be better defined as a change from one type of tissue (with multiple differentiated cell types) into another, and such a change is likely to result from alterations in the expression of key developmental transcription factors in undifferentiated progenitor cells that have the capacity to produce and maintain multiple cell types.5 ,6 In Barrett's oesophagus, GORD is considered the condition that reprograms progenitor cells to express columnar rather than squamous developmental transcription factors, but the identity of the Barrett's progenitor cells remains unclear. Studies have suggested that the progenitors might be native to the oesophagus (cells in the squamous epithelial basal layer or in the ducts of submucosal glands), or they might migrate into the oesophagus from the GOJ or proximal stomach when the squamous epithelium is damaged by GORD.6 ,7 Derakhshan's study suggests that these progenitor cells might first give rise to cardiac mucosal metaplasia, which later evolves into intestinal metaplasia.
What can we glean from available studies on cardiac mucosa? It is present in a very narrow strip (usually <3 mm) at the GOJ in most apparently normal individuals, but this is not proof that it is a normal structure. Autopsy studies of soldiers killed in the Korean War revealed atherosclerotic lesions in most, but atherosclerosis is not considered a normal condition. Furthermore, there is considerable evidence to suggest that cardiac mucosa is metaplastic in some, if not all cases, and that cardiac mucosa is the precursor of intestinal metaplasia with goblet cells. For example, in patients who have oesophagectomy with oesophagogastrostomy, which often is complicated by reflux oesophagitis and columnar metaplasia in the remnant oesophagus, cardiac mucosa appears to develop before intestinal metaplasia.8 Like Derakhshan, other investigators have shown that cardiac-type mucosa in the oesophagus expresses molecular markers of intestinal differentiation,9 and cardiac mucosa can exhibit DNA content abnormalities similar to those of intestinal metaplasia with goblet cells.10 The Glasgow group has shown that cardiac mucosa almost always shows signs of chronic inflammation, a common cause of metaplasia, and that the extent of cardiac mucosa correlates positively with age, abdominal obesity, and the extent of acid exposure in the distal-most oesophagus.2 The report by Derakhshan adds to this accumulating body of indirect evidence supporting the notion that cardiac mucosa often is metaplastic, and the precursor of Barrett's intestinal metaplasia.3
Competing interests None.
Provenance and peer review Commissioned; internally peer reviewed.
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.