Article Text


Giving carditis back to the heart
  1. R M GENTA
  1. Department of Pathology 113
  2. VA Medical Centre and Baylor College of Medicine
  3. 2002 Holcombe Blvd
  4. Houston, TX 77030, USA

Statistics from

See article on page 798

Anatomically vague (it was named after its proximity to the heart), histologically undistinguished (its mucosa is usually described as being “similar to the mucosa of the antrum”) and functionally considered a drab territory that connects two well characterised segments of the digestive system, the gastric cardia has long been ignored by gastroenterologists, pathologists, and physiologists alike.

Suddenly, in the past few years, this neglected Grenz zone has been catapulted to the centre of the gastroenterological stage. Another spin-off of Helicobacter pylori? In a sense, yes. But, whereas the wily bacterium is blamed for a long list of calamities occurring in the remainder of the stomach, duodenum, and other systems, in the cardia it is portrayed as a protector of mucosal integrity.1 An implausible defensor mucosae.

During the past few decades a dramatic rise in the incidence of adenocarcinoma of the cardia has been reported in the very populations in which the incidence of gastric cancer has been decreasing.2 The latter trend has been explained as a consequence of the declining prevalence of H pylori infection. However, the increased incidence of cancer of the cardia is paradoxical: why should this minuscule ill-defined portion of the proximal compartment of the stomach behave differently from its more distal areas? One possible explanation is that the lack of a consensus on what the cardia area is, the difficulty of establishing the precise site of origin of tumours of the gastro-oesophageal junction, and various inaccuracies in tumour coding in cancer registries may have unwittingly conspired to inflate trends in cardia cancer.3 In fact, most cancers classified as cardia tumours may arise in Barrett's epithelium in the distal oesophagus. The rise in acid reflux and Barrett's oesophagus in Western populations is probably related to dietary trends (more dietary fat), which have increased the prevalence of obesity,4 and by the increased gastric acid production made possible by the decreasing prevalence of H pylori induced atrophic gastritis.1 5 Recently, more investigators are inclined to accept the oesophageal origin of junctional tumours and have called for a consolidation of cancer arising in this area into a single clinicopathological entity.6 7 Nevertheless, discovering what happens in the cardia may be important, or even crucial, to understanding the events taking place at the junction.

In 1994 we tackled the issue of cardiac involvement byH pylori 8: in subjects withH pylori elsewhere in the stomach the mucosa of the cardia was almost always directly affected, and the inflammatory responses in the cardia and the antrum were similar in both quality and intensity. Our study did not suggest that the cardia was a compartment of the stomach affected by H pylori in a distinctive fashion. Therefore, we did not see the need for a special term (carditis?) any more than we believed that inflammation of the pylorus should be called pyloritis or the inflammation of the greater curvature grand curvitis. In 1996 Riddell introduced the term carditis,9 hitherto used exclusively in reference to inflammatory diseases of the heart, to connote “inflammation of the gastric cardia.” Since then, the perception of an epidemic of junctional cancer, the suggestions that H pylori eradication may result in increased acid reflux, and perhaps the relative virginity of the territory have provided the momentum for numerous investigations of the prevalence, nature, and importance of inflammation in the gastric cardia.

The results of these studies are remarkable for their disparity. Goldblum and colleagues10 found inflammation of the cardia in 23 patients with gastro-oesophageal reflux disease (GORD); only one of these patients was not infected with H pylori. Similarly, El-Serag et alfound no instances of inflammation of the cardia not associated withH pylori infection in 302 patients with reflux oesophagitis.11 In contrast, Chenet al found “some carditis” in 107 (92%) of 116 unselected subjects.12 Forty per cent of these subjects with moderate and 57% of those with severe carditis hadH pylori infection. The authors focused on these data to conclude that carditis is significantly associated withH pylori active gastritis, but not with symptoms or signs of gastro-oesophageal reflux. The same figures, however, show that 60% of individuals with moderate and 43% of those with severe inflammation in the cardia did not haveH pylori infection. In Voutilainen's Finnish series of 110 patients with junctional intestinal metaplasia, 101 (92%) had cardia gastritis, but only 50 (46%) hadH pylori infection; of the 1009 controls without junctional intestinal metaplasia 737 (73%) had carditis, but only 363 (36%) had H pyloriinfection.13 The study by Bowrey et al in this issue (see page 798) confirms that subjects with GORD have a high prevalence of cardia inflammation, even in the carefully documented absence of a H pylori infection.

Thus, we have several major discrepancies to explain. Why do some authors detect essentially no inflammation in the cardia ofH pylori-free subjects,10 11whereas others find that a large portion of uninfected subjects have isolated gastritis of the cardia (Bowrey et al)?12 13 A second issue raised by these data is the relation between cardia inflammation and reflux: some authors believe that inflammation in the cardia and evidence of GORD (Bowreyet al)13 go hand in hand, whereas other investigators fail to see evidence for such a correlation.10 11 Finally, if we do accept the existence of an entity to be provisionally called “non-H pylori isolated carditis,” we may have to appose the defeating qualifier “idiopathic” to indicate our ignorance about its aetiology. Having now created the novel entity “idiopathic non-H pylori isolated carditis” we must ask ourselves what its importance might be. The practical corollary of this theoretical question is, of course, whether it is sensible for endoscopists to go to the extra effort and retroflex the endoscope to obtain one or two biopsy specimens from the cardia. In the present state of knowledge, the histopathological answers they receive will be very unlikely to influence patient management.

Geographical differences are unlikely to explain the disparities found in the studies outlined earlier. Although the ethnic background of the subjects is rarely stated or taken into account in the analyses, most patients in Voutilainen's series were certainly Finns,13the majority of Bowrey's patients were likely to be Welsh, and Chen's subjects were Americans with various percentages of the ethnic groups that constitute the population of Boston.12 Yet, these diverse groups had comparable prevalence of “idiopathic non-H pylori isolated carditis”. Criteria used to define carditis are also unlikely to explain the substantial differences found in different populations. Most investigators have made a reference to their pathologists' use of the updated Sydney System.14 Even taking into account interobserver variation among pathologists, the discrepancies seem too wide to be explained by this interpretative variable alone. The inevitable conclusion is that, at this point, we do not have the necessary information to even guess the true prevalence of non-H pylori carditis in different populations, we do not know what it means, and have no idea what should be done about it. In this situation, the only sensible course of action is to encourage the performance of well designed studies, of which Bowrey's article is an example, in geographically and ethnically diverse populations of clinically well defined subjects. Data on the intensity and topographical distribution of gastritis ought to be available for each subject, so that correlations with inflammation in the cardia can be made. The combined information derived from such studies may eventually help to remove “idiopathic” from the provisional name given to the non-H pyloricarditis. It might even suggest that the term carditis should be given back to the heart.

See article on page 798


View Abstract

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.

Linked Articles