Helicobactoer pylori affects the magnitude of acid inhibition by proton pump inhibitors.1-3 In an early study, omeprazole produced greater acid suppression in subjects with H pylori infection than in those without infection.1Direct evidence of the interaction between H pylori infection and the efficacy of omeprazole came from studies in which cure of the infection resulted in a decrease in the antisecretory effect exerted by omeprazole, both in healthy subjects2 and in patients with duodenal ulcer disease.3

In this issue (see page 468) Gillen et al confirm the effect of H pylori on the reduction in acidity produced by omeprazole. The authors have tried to elucidate the mechanism responsible for this effect. They report data which seem to refute the previously postulated hypothesis1-3 that ammonia is important in the interaction between H pylori and omeprazole.

Nevertheless, there are many good reasons favouring the ammonia buffering system hypothesis. Firstly, this theory takes into account the equilibrium between secreted protons and ammonia produced by H pylori. The ammonia buffering system has a minimal apparent effect in the absence of omeprazole therapy when the H⁺ concentrations in the stomach are high, but it becomes effective when H⁺ concentrations decrease in response to omeprazole treatment.1-4 Secondly, the loss of effectiveness of secretory inhibitors after eradication of the organism, along with the disappearance of ammonia from the stomach, further support this theory.2-5 Thirdly, the ammonia hypothesis could explain why this effect is less pronounced with H₂ receptor antagonists. As acid suppression achieved using H₂ receptor antagonists is lower than that with proton pump inhibitors, ammonia cannot influence intragastric pH.5Fourthly, ammonia produced by H pylori may increase H⁺ back-diffusion across damaged gastric mucosa, thus lowering H⁺ concentrations in the stomach.5 The effect of the ammonia buffering system would become particularly apparent with a further decrease in H⁺concentrations induced by omeprazole therapy. Finally, the mechanisms whereby H pylori and omeprazole interact may be dependent upon dose and duration of antisecretory therapy. In earlier studies, measurements were conducted after a one week course of omeprazole 20 mg daily.1-4 In Gillenet al’s study, a higher dose of omeprazole (40 mg o.d.) was administered for eight weeks. With this treatment, in contrast to low dose, short term therapy, H pylori infection is suppressed and ammonia production is therefore relatively low. Overall, the enhanced antisecretory effect of short term omeprazole therapy in H pyloriinfected subjects may be a result of the production of acid neutralising compounds by the bacterium.

Gillen et al propose an alternative mechanism for the greater pH raising effect of omeprazole therapy during H pylori infection. They postulate that the severity of H pylori induced corpus gastritis is exacerbated by omeprazole; this would lead to decreased acid secretion and increased pH. Unfortunately, no data to support this proposal are provided. Gastritis was not assessed. Furthermore, the gastritis hypothesis does not fully explain the increased intragastric pH observed during potent antisecretory therapy in subjects withH pylori infection. Firstly, the combination of high grade body gastritis and hyposecretion does not occur in patients with duodenal ulcer disease.3 These patients, as the authors themselves point out, are acid hypersecretors.7 Secondly, the increased efficacy of antisecretory therapy has been observed on the first day of administration of a reversible proton pump inhibitor (BY 841) to asymptomatic H pylori positive subjects.8 It is highly unlikely that redistribution ofH pylori to the gastric corpus and subsequent worsening of body gastritis occurs within a few hours of administration of acid inhibitory therapy. Finally, the decrease in response to omeprazole after eradication of H pylori is rapid and does not progress, whereas gastritis continues to improve during the first few months after eradication.9 Thus this mechanism is speculative.

Some drawbacks in Gillen et al’s study mitigate against the validity of the gastritis hypothesis. Firstly, the diagnosis of H pylori status was based on urea-¹⁴C breath tests alone. No single technique, however, can be considered ideal for the detection ofH pylori infection. The high pH and ammonia values observed by Gillen et al in breath test negative subjects (fig 1) are, in our view, suggestive that these subjects were, in fact, H pylori positive. Serological or histological analysis should have been conducted in these subjects. Secondly, gastric acid output studies were performed before and after omeprazole administration to separate groups ofH pylori positive and negative subjects, instead of studying subjects before and after cure of the infection. Thirdly, gastric acid secretion and pH were measured in single samples of gastric juice obtained during 30 minute periods. Consequently, hourly variations in gastric secretion and ammonia output, additional factors such as reflux of duodenal alkali into the stomach, and loss of gastric contents into the duodenum were not evaluated. In previous studies, this has been overcome by prolonged overnight measurements and by perfusing the stomach with volume markers.4 The question therefore arises whether the values obtained by Gillenet al are representative of acid secretion of the subjects studied. Fourthly, the role of other neutralising substances with a pKa above pH 7 such as secreted immunoglobulins and albumin leaking through a damaged mucosa was not evaluated.10 Finally, statistical analysis of the data is difficult to understand. Twenty H pylorinegative and 12 H pylori positive subjects were included in the study. Ammonia samples of a number of subjects are reported to be “not available”. We wonder whether the authors have included only the subjects with available ammonia values in their comparisons of group medians (acid and ammonia).

In conclusion, Gillen et al’s study confirms the previous observations that acid suppressive treatment in subjects with H pylori infection is more effective than in H pylori negative subjects.1-5 ,9 In addition, the authors show that this effect persists after long term treatment with proton pump inhibitors. However, the mechanism underlying the pH elevating effect of omeprazole therapy in H pylori positive subjects is not elucidated and awaits further study.