Is the gastric mucosa normal in patients with GERD?
J.I. Wyatt, M.F. Dixon (Leeds)
Gastroesophageal reflux disease (GERD) is generally attributed to acid injury to the squamous esophagus, although the role of reflux of alkaline duodenal contents, particularly in the production of Barrett's esophagus, is gaining wider acceptance. In considering the state of the gastric mucosa in patients with GERD, it has to be borne in mind that in most populations at risk there will be a high prevalence of Helico-bacter pylori-associated chronic gastritis. Thus, one cannot expect the mucosa to be normal in the majority of patients with GERD, and a more pertinent question would be: "Is the pattern of gastritis found in patients with GERD different to that found in a control population?"
One could anticipate two variations to the pattern of gastritis in GERD patients. Firstly, the motility disturbance affecting gastroesophageal competence could also affect antropyloric function and give rise to duodenogastric reflux. Such individuals may exhibit reflux gastritis to a greater degree than control subjects without GERD. Secondly, given that GERD might be more frequently seen in association with increased gastric acid output, it could be anticipated that GERD patients will have the pattern of H. pylori-associated gastritis related to hyperacidity - namely an antral-predominant pattern - more frequently than control patients. The data available on both these aspects are strictly limited.
GERD and reflux gastritis
The concept of a distinctive form of gastritis related to reflux of bile and pancreatico-duodenal juice into the stomach is not a new one, but had been poorly documented prior to our 1986 publication . Although originally based on reflux in the postoperative stomach, we have since established that "spontaneous" reflux of bile and the presence of a reflux gastritis can occur in the intact stomach . While "pure" acid reflux could be entirely attributed to gastroesophageal malfunction, alkaline reflux into the esophagus must reflect passage through the stomach which
provides an opportunity for damage to the gastric mucosa. Although the existence of pyloric incompetence in patients with GERD , the presence of alkaline gastroesophageal reflux  and coexistence of alkaline reflux gastritis and esophagitis  have been known for several years, few studies have examined gastric biopsies and none has applied modern diagnostic criteria in the histological interpretation. Thus, Cowen  investigated the relationship between spontaneous enterogastric reflux gastritis and esophagitis and found that all 42 patients with intragastric bile reflux and GERD exhibited "chronic gastritis". However, of these 42 subjects, only 11 showed chronic atrophic gastritis and six had chronic active gastritis; the nature of the gastritis in the other 25 patients is not stated.
When we graded esophageal biopsies for the severity of esophagitis in 24 patients whose gastric biopsies revealed the histological changes of reflux gastritis, and compared them with those from 31 subjects whose gastric biopsies were normal, we found no significant difference in histological "esophagitis scores" [Wyatt and Dixon, unpublished observations]. However, the histological picture of reflux gastritis can result from factors other than exposure to bile. We have found identical appearances in patients on long-term aspirin and NSAID treatment and in some individuals who abuse alcohol . Therefore, our findings cannot be used to deny a link between esophagitis in the squamous mucosa and alkaline reflux gastritis.
Thus, while there are theoretical grounds for the coexistence of reflux gastritis and esophagitis, this has not been adequately validated. Given that alkaline reflux is considered to be a more important factor in the pathogenesis of Barrett's esophagus, an association is more likely to emerge if reflux gastritis was sought in these patients.
GERD and H. pylori-associated gastritis
It is now clearly established that H. pylori infection and the resultant chronic gastritis leads to sustained hypergastrinemia and increased acid output . It could be anticipated, therefore, that patients with symptomatic esophagitis will have a higher prevalence of H. pylori-associated gastritis than age-matched controls without esophagitis. However, what little evidence we have is contradictory. In children, Rosioru et al. found the prevalence of biopsy-proven esophagitis to be similar in H. pylori positive and negative groups .
Koop et al.  found a low prevalence of H. pylori-associated antral gastritis (24%) in patients with reflux esophagitis refractory to H2-blockers. A study on gastric acid secretion in GERD patients with and without concomitant duodenal ulcer  confirmed that those with coexistent DU had significantly higher gastric acid output than those with esophagitis alone, but did not demonstrate any greater severity of the esophagitis in the former group. While the great majority (if not all) of the DU patients in this study would have been H. pylori positive, those with esophagitis alone may or may not have been infected. A higher proportion of H. pylori negative patients in the latter group could explain why lower acid production was present, but this does not affect the authors' overall conclusion that acid output is not a major pathogenetic factor in GERD.
Some circumstantial evidence in favour of a relationship between H. pylori infection and GERD comes from histological studies. When esophageal biopsies from patients with antral predominant H. pylori gastritis (the pattern most closely related to increased acid production) were compared to biopsies from patients with a diffuse H. pylori positive pangastritis (a pattern in which normal or subnormal acid secretion occurs), the former group exhibited a higher mean esophagitis score (2.9 vs. 1.53; p < 0.01). However, it should be re-emphasized that histological esophagitis is poorly validated against objective measures of acid reflux such as 24-h pH monitoring, and no acid secretory data was available in this study.
A relationship between H. pylori-associated gastritis and GERD could only be properly investigated by studying the effects of eradication treatment on symptoms and acid reflux tests in H. pylori positive patients with GERD. Their outcome would be compared with non-eradicated controls. The results of such studies are awaited with interest.
7. El-Omar E, Penman I, Dorrian CA, Ardill JES, McColl KEL. Eradicating Helicobacter pylori infection lowers gastrin-mediated acid secretion by two-thirds in duodenal ulcer patients. Gut 1993;34:1060-1065.
10. Zhu H, Pace F, Sangaletti O, Bianchi Porro G. Gastric acid secretion and pattern of gastroesophageal reflux in patients with esophagitis and concomitant duodenal ulcer Scand J Gastroenterol 1993:28:387-392.