Is Helicobacter pylori status relevant in the management of GORD?

Authors


Professor S. Vigneri, Institute of Internal Medicine, University of Palermo, Via del Vespro n°141, 90127 Palermo, Italy. E-mail: vigneri@madeinsicily.com

Summary

There is growing interest in the relationship between H. pylori infection and gastro-oesophageal reflux disease (GORD). However, this relationship is complex, as yet not fully elucidated, and probably based on a multiplicity of factors. The prevalence of H. pylori infection in patients with GORD is similar, more often lower than in matched controls. There is a negative correlation between H. pylori infection and the severity of GORD. There are many hypothetical mechanisms by which H. pylori infection may protect from the development of GORD. Conversely, there are many possible mechanisms by which H. pylori infection could theoretically foster the GORD. Patients after H. pylori eradication may develop GORD, and this seems to suggest a protective role of H. pylori infection, but other possible explanations include weight gain after H. pylori eradication, changes in dietary habits and smoking, and pre-existing GORD.

H. pylori infected patients treated by various acid-inhibiting therapies such as proton pump inhibitors (PPIs), H2-receptors antagonists (H2-RA) or vagotomy, have an increase of their corpus gastritis severity, both in the activity of inflammation and in the density of organisms. Long-term therapy of GORD in H. pylori infected may lead to rapid progression of atrophic gastritis intestinal metaplasia and dysplasia, and increase the risk of developing gastric cancer. More recently it has been shown that H. pylori infection may interfere with the acid suppressive therapies used for treating GORD.

In our opinion the progression of gastritis depends on the threshold of acid output at which H. pylori can ‘flourish’. Recently interest is growing on gastric transitional zones and Helicobacter ecology. Any decrease of acid secretion changes the behaviour of H. pylori: the activity of gastritis improves in the antrum, but it deteriorates in the body. During proton pump inhibitor treatment, H. pylori redistribution occurs within the stomach, from an antral to a corpus or fundus prevalent pattern; corpus-fundus gastritis, exacerbated by PPI therapy, may result both in a diminished acid secretion and gastro-oesophageal reflux.

The interest in Barrett's oesophagus is growing due to the associated risk of adenocarcinoma. The literature seems to demonstrate that the prevalence of H. pylori infection of the stomach in Barrett's oesophagus patients is not different from that exhibited by controls, roughly one-third of the subjects. Intestinal metaplasia of the gastric cardia seems to be equally frequent in patients with and without GORD. Finally, it appears unlikely that a causal relationship exists between H. pylori infection and Barrett's-associated adenocarcinoma.

Introduction

Helicobacter pylori infection plays an important aetiological role in causing gastritis, peptic ulcer diseases, gastric adenocarcinoma and primary B-cell lymphoma of the stomach. These pathologies are frequent and for this reason several medical authorities believe that H. pylori infection, even without clinical manifestations, must be treated whenever it is identified. On the other hand, to date the epidemiological evidence demonstrates that only 20% of infected people will develop symptoms and lesions and that an aggressive ‘test and treat’ policy may promote the development of resistant bacteria and could cause morbidity and mortality due to the toxic effects of the antibiotic treatment. There is much evidence and growing interest in the idea that H. pylori is probably not very harmful to human beings and it may live in a symbiotic relationship with the human host. To date there is yet no direct evidence that H. pylori carries any benefit to human beings, but an exception is represented by the observation that H. pylori eradication might increase the incidence of Gastro-Oesophageal Reflux Disease (GORD) and it is the first demonstration of unfavourable risk/benefit ratio in the people who have an eradicating treatment. However, data on the association between H. pylori and GORD are conflicting. In this review, we aimed to identify the possible interrelationship among the H. pylori infection and reflux oesophagitis evaluating whether the symptoms or natural history of GORD are modified by eradication of H. pylori or whether GORD can develop ex novo following the cure of H. pylori infection. Furthermore we evaluate the proposed pathogenetic mechanisms involved in the link between H. pylori and GORD. Then we will focus on the evidence that antisecretory treatments in patients with H. pylori infection, may induce a redistribution of H. pylori to the gastric corpus with subsequent worsening of body gastritis and development of gastric atrophy and intestinal metaplasia a potentially precancerous lesion. In the light of this, the same authorities recommended a policy of H. pylori testing and treating of GORD patients before beginning long-term antisecretory therapy. But things are not so easy as we could think and the efficacy of proton pump inhibitors, seem to diminish after H. pylori eradication. Finally, a renewed area of interest is the relationship between H. pylori and Barrett's oesophagus. The present review will focus on these issues and critically evaluate the data gathered from medical literature.

Epidemiology

Many studies have evaluated the relation between GORD and H. pylori by determining the prevalence of H. pylori infection in these patients. Recent studies have shown that the prevalence of H. pylori infection in patients with reflux oesophagitis is similar, or more often lower than in matched controls.1,2, 3,4,5 In a recent review the author 6 summarized the results of 26 studies and observed that between the GORD patients studied, H. pylori diagnosed by various methods ranged from 16% to 88% and overall 851 (40.3%) among 2112 GORD patients studied, were H. pylori-positive. However, there were no differences in prevalence between the different grades of oesophagitis and sexes. The prevalence increased with age, probably owing to an age-cohort effect. In the studies where control populations were analysed there was no significant difference in the prevalence of H. pylori between the patients with reflux disease and control group and when all of these studies are analysed collectively, the overall low prevalence of H. pylori perhaps would suggest not only that the bacterium is very unlikely to have a significant role in the pathogenesis of GORD, but that the pathogenesis of GORD might be related in same way to the absence of H. pylori. Further support in favour of the hypothetical protective role of H. pylori in GORD is the observation of a negative correlation, in GORD patients, between H. pylori infection and the severity of GORD. Schenk et al. 7 observed that H. pylori-negative patients had at baseline endoscopy a more severe oesophagitis and more frequent Barrett's oesophagus than the H. pylori-positive patients. Different results were found by other authors 8 since no correlation was detected between H. pylori status and the severity of oesophagitis. Some studies evaluated also the presence of H. pylori in the squamous oesophageal mucosa of patients with reflux disease; all but one 9 failed to detect the H. pylori in that side. 1–14 There is a strong causal relation between gastro-oesophageal reflux and oesophageal adenocarcinoma and now we know that the more frequent, more severe and longer-lasting the symptoms of reflux are the greater the risk, is regardless of the presence of Barrett's oesophagus. 15 The prevalence of H. pylori infection is decreasing in white people in developed countries and this is paralleled by an increased incidence of GORD and, primary adenocarcinoma of the oesophagus.16,17 These opposing time trends of gastroduodenal pathologies (peptic ulcer disease and gastric cancer) vs. oesophageal disease are consistent with the hypothesis that the declining infection rates of H. pylori in developed countries, could result in a concomitant rise in the occurrence of gastro-oesophageal reflux disease and oesophageal adenocarcinoma. 18 Recently some studies suggested that patients infected with CagA+ strains were less prone to suffer with severe GORD, its complications, including Barrett's oesophagus, oesophageal distal adenocarcinoma and cancer of the cardia.19,20 Asians and Blacks are significantly less prone to develop oesophageal adenocarcinoma than white people and since they are frequently infected with CagA+ strains, the latter might have a protective role. 21 In conclusion, the overall epidemiological evidence would indicate that H. pylori is more likely to be beneficial rather than neutral or even dangerous as far as, the development of GORD and its complications is concerned.

HELICOBACTER PYLORI and GORD: PATHOPHYSIOLOGICAL INTERRELATIONS

There are many hypothetical mechanisms by which H. pylori infection may protect from the development of gastro-oesophageal reflux disease. Different factors contribute to aetiopathogenesis of gastro-oesophageal reflux disease, but the aggression of refluxed material and the impaired oesophageal clearance are the most important mechanisms leading to GORD development. H. pylori infection is associated with an abnormal acid secretion but this effect depends on different conditions. At first H. pylori lowers intragastric acidity by the release of substances that inhibit gastric acid secretion, 22 moreover the bacterium produces by its ureasic activity, a large amount of ammonia. 23 The latter acts as a powerful neutralizing substance only at elevated gastric pH while demonstrating no effect on the physiological pH in the normally secreting stomach. In the oesophagus, where physiologically pH is high, the gastric juice refluxing could be neutralized by ammonia generated by H. pylori activity. On the other hand, gastric acid secretion depends on corpus gastritis because is demonstrated that patients with multifocal atrophic gastritis have lowered gastric acid secretion or complete achlorhydria. Theoretically, H. pylori infection would protect against GORD and this hypothesis is supported by the observation that eradication of H. pylori, with partial reversal of the hypochloridria, 24 represents a risk factor for the development of reflux. 25 Further, gastrin may increase the Lower Oesophageal Sphincter (LOS) pressure 26 and H. pylori infection in both patients with duodenal ulcer and non-ulcer dypepsia is associated with raised basal, meal stimulated, Gastrin Release Peptide (GRP) stimulated gastrin concentration,27,28, 29,30,31 while eradication of H. pylori results in a decrease in gastrin concentrations.32,33 GORD is a motility disorder and the main abnormality is incompetence of the LOS due to a steady low pressure or its inappropriate Transient Lower Oesophageal Sphincter Relaxations (TLOSR). It is possible that substances released by H. pylori or produced by the gastritis could reduce the frequency of (TLOSR). It is also possible suppose that H. pylori gastritis raises the threshold for triggering TLOSR and reflux episodes by altering gastric distension. The role of H. pylori for gastric emptying remain controversial34,35,36 even if there is some indirect evidence that H. pylori has a role in delaying gastric emptying and the degree of this effect correlates with the severity of antral inflammation. 37 However, to date evidence would suggest H. pylori does not predispose to GORD through alterations in gastric emptying. Some evidence links H. pylori corpus gastritis CagA gene and protection against GORD and its complications 19 and Graham and Yamaoka 38 recently described H. pylori infection as a biological antisecretory agent, because according to them, the presence of CagA-positive H. pylori infection not only protects against GORD, but also protects against its complications, because corpus gastritis in patients with H. pylori infection and CagA-positivity induces a lower gastric acid secretion than does H. pylori infection without CagA-positivity. Conversely, there are many possible mechanisms by which H. pylori infection could theoretically foster the gastro-oesophageal reflux disease. H. pylori infection with antrum predominant gastritis induces an increase in gastrin release and gastric acid secretion so therefore in same patients with GORD higher acidity and volume of gastric juice might worsen reflux disease. H. pylori in the cardia may influence by direct and indirect effects the sphincter pressure and competence. In patients not suffering from GORD, H. pylori density and severity of gastritis are similar in the cardia and antrum, 39 while they are lower in patients with GORD. 40 We know that H. pylori gastritis induces the release of prostaglandins, cytokines and nitric oxide which may have an adverse effect on LOS, by promoting an inflammatory response, mucosal damage and by sensitizing afferent nerves driving a vicious cycle.41,42,43 In conclusion to date the role of H. pylori infection on the pathogenesis of GORD seems to be very unlikely. All of these proposed pathophysiological interrelations are at present strongly speculative and from the present data it is impossible to came to a definitive conclusion.

HELICOBACTER PYLORI ERADICATION and GORD

Several studies in healthy subjects and duodenal ulcer patients noted that many patients after H. pylori eradication developed reflux oesophagitis during the follow-up.44,45,46 Recently, studies from Europe and Japan demonstrated an increased prevalence of erosive oesophagitis in peptic ulcer patients eradicated for H. pylori infection without reflux oesophagitis at baseline endoscopy.47,48 In the German study, 47 patients with duodenal ulcer were studied. The patients with persistent infection had 12.9% of incidence of oesophagitis within 3 years, compared with 25.8% of the eradicated patients. In 70 infected Japanese patients with duodenal ulcer, atrophic gastritis and gastric ulcer, after H. pylori eradication an erosive oesophagitis developed in 14.3% of them. 48 These results seem to suggest a protective role of H. pylori infection against the new development of GORD after the eradication of the bacterium, but other possible explanations are hypothesized, including weight gain after H. pylori eradication, changes in dietary habits, smoking and pre-existing GORD in patients with H. pylori positive peptic ulcer. There is some evidence that H. pylori negative patients have more severe grade of oesophagitis than H. pylori positive ones, 7 but more recently Tefera et al. 49 found no significant change in the severity of oesophageal acid exposure or reflux symptoms 3 months after H. pylori eradication in patients with pre-existing reflux oesophagitis. Carbone et al. 50 observed no variations of reflux pattern before and after H. pylori eradication in patients with dyspepsia and negative endoscopy. These overall results do not support definitively the hypothesis that H. pylori eradication might induce or aggravate gastro-oesophageal acid reflux.

HELICOBACTER PYLORI and GORD TREATMENT: WHAT ARE THE INTERRELATIONS?

In the last few years it has been shown that H. pylori infected patients treated by various acid-inhibiting therapies such as Proton Pump Inhibitors (proton pump inhibitors), H2-Receptors Antagonists (H2-RA) or vagotomy, have an increase of their corpus gastritis severity, both in the activity of inflammation and in the density of organisms. Rapid progression of atrophic gastritis may lead to intestinal metaplasia and dysplasia, increasing the risk of developing gastric cancer, since gastric body atrophy is a gastric cancer precursor lesion. GORD long-term therapies in H. pylori infected, especially with proton pump inhibitors may accelerate the development of atrophic gastritis, intestinal metaplasia and perhaps of gastric cancer. Therefore, the need for eradicating H. pylori infection in patients receiving long-term acid inhibition has been strongly recommended in light of this. More recently it has been suggested that H. pylori infection may interfere with the acid suppressive therapies used for treating GORD. Therefore, H. pylori infection seems to produce a greater decrease in gastric acidity during treatment with either H2-RA or PPIs. Conversely curing the infection significantly reduces the pH increasing effect of proton pump inhibitors while the loss of efficacy with ranitidine seems to be less pronounced. Therefore, in H. pylori infected patients with proton pump inhibitor-dependent GORD, have we to eradicate the bug or not? We have a situation in which one has to choose between two possible undesirable consequences!

Anti-secretory drugs and atrophic gastritis

Therapy with PPIs, in particular long-term treatment induces a redistribution of H. pylori and gastritis with an increase of the density of bacteria and a rapidly progressive atrophic gastritis in the corpus. The latter represents a precancerous lesion and this is the most important argument in favour of eradication of H. pylori in patients with GORD. Our group was the first to point out that omeprazole short-term therapy modifies the gastric distribution of H. pylori and we suggested that after omeprazole therapy, which causes a significant reduction in gastric acidity, H. pylori spreads and flourishes in the body, if it has not already done so. 51 Later on, we demonstrated that the change in the pH of the gastric environment produced by omeprazole long-term treatment induced not only a further significant increase in the body density of H. pylori,52,53 but also a worsening in the activity of corpus gastritis. 54 However, we did not observe any significant increase in gastric intestinal metaplasia or gastric dysplasia after 18 months of omeprazole 20 mg daily or omeprazole 20 mg every other day therapy. Subsequent observations confirmed that the changes in distribution and activity of chronic gastritis. which are proton pump inhibitors therapy related, depend on the H. pylori infection55,56, 57,58, 59,60,61 and it seems that there is no difference between different proton pump inhibitors treatments utilized. 56–60 Kuipers et al. 57 found evidence of an accelerated development of atrophic corpus gastritis (30.5%) in patients on long-term omeprazole therapy (range 3–8 years) for GORD in contrast to patients who underwent Anti-Reflux Surgery (ARS) (0%), suggesting, like other authors, that the patients in need of long-term acid suppressive therapy should receive bacterial eradication therapy if they are H. pylori-positive. 56–63 Furthermore, Berstad et al. 58 observed an increase of epithelial cell proliferation in the corpus during long-term treatment with lansoprazole, only in the patients with H. pylori infection and such proliferation increased with increasing degree of inflammation. These observations too support the notion that H. pylori eradication should be seriously considered before long-term treatment with proton pump inhibitors is initiated. However, some studies failed to demonstrate any evidence of an accelerated development of atrophic gastritis in patients infected with H. pylori, on long-term profound acid inhibition. Lundell et al. 64 compared the clinical efficacy of omeprazole and ARS in the long term management of GORD. A total of 155 patients were randomized to undergo ARS and 155 patients were randomized to continue omeprazole therapy. During a follow-up period of 3 years, in H. pylori infected patients with chronic GORD, it was found omeprazole therapy is not associated with an increased risk of development of corpus glandular atrophy and intestinal metaplasia. The annual increment of atrophy was 8.4% in the H. pylori-negative subjects who had moderate to severe gastritis before treatment whereas it was 8.3% in H. pylori infected subjects. Also Stolte et al. 60 observed that lansoprazole aggravates the gastritis in the corpus, but the PPIs does not induce any increase in the incidence of atrophy/intestinal metaplasia after 12 months of treatment. Another study by Lamberts et al. 65 concluded that prolonged achloridria induced chronic atrophic gastritis over the duration of the study (5 years) but did not transform type I metaplasia into type III metaplasia, in turn leading to dysplasia. However, ranitidine and aluminium-magnesium hydroxide also worsen the activity of H. pylori gastritis in duodenal ulcer patients 66 but, at present, we do not know why patients with H. pylori infection and treated with acid-suppressive therapy may have an increased risk concerning the development of the body atrophic gastritis. Several mechanisms have been proposed such as an increased antigenic toxic effect depending on a deeper location of H. pylori in the gastric glands, 67 antigenic mimicry between H. pylori and gastric mucosa influenced by hypoacidity 68 and differences in HLA genotype. 69 In our opinion the progression of body gastritis depends on the threshold of acid output at which H. pylori can ‘flourish’. The most important ecological determinant in H. pylori infection is sensitivity to local acid output, in particular the pH in microenvironment of the stomach. 70 The baseline individual acid output allows H. pylori to flourish in the antrum in normal acid-secreting individuals or in both the antrum and the body in persons having local acid output below the inhibitory threshold. The interactions between H. pylori and gastric acidity are poorly understood and much remains to be learned about the mechanisms by which environmental pH and H. pylori may interrelate, but the survival of H. pylori in the gastric environment may depend on both the development of specialized intrinsic defences and the bacterium's ability to induce physiological alterations in the host environment. 71 Recently interest is growing on gastric transitional zones and Helicobacter ecology. 72 The transitional zones are dynamic rather than static areas and the local acid levels modifies the behaviour of H. pylori and the pattern of distribution of gastritis in the stomach. The gastritis is confined to the antrum by high local acid levels and high acid output may be the limiting factor in the development of body gastritis. Any reduction of acid secretion, more pronounced with the proton pump inhibitors, changes the behaviour of H. pylori and the activity of gastritis improves in the antrum, when it deteriorates in the body. Meyer-Rosberg et al. 73 have shown that in the pH range 3.5–8 the bacterium flourishes and maintains the Proton Motive Force (PMF), electrochemical gradient for proton across the plasma bacterial membrane, which ensure a continuous supply of energy through adenosine triphosphate synthesis. The distribution of the organism and its activity is consistent with the effect of pH and urea on the PMF of H. pylori. The pH elevation in the antrum resulting from inhibition of acid secretion and a further elevation of pH by bacterial urease activity, lessens the number of bacteria. Contemporarily the increase of fundic pH induced by proton pump inhibitors allows us to obtain values that permit protein synthesis and growth and so H. pylori becomes prevalent in the fundus and the activity of gastritis increases. This pattern is reversible if acid secretion is restored ( Fig. 1). However an FDA advisory panel concluded that the ‘available evidence, does not indicate that chronic proton pump inhibitors therapy of GORD increases the risk of gastric atrophy, intestinal metaplasia or gastric cancer’. 74

Figure 1.

At the pH range 3.5–8 the bacterium flourishes and maintains the Proton Motive Force (PMF), electrochemical gradient for proton across the plasma bacterial membrane, which ensure a continuous supply of energy through adenosine triphosphate synthesis 73. The pH elevation in the antrum resulting from inhibition of acid secretion and a further elevation of pH by bacterial urease activity, lessens the number of bacteria. The increase of fundic pH induced by proton pump inhibitors allows us to obtain values that permit protein synthesis and growth and so H. pylori becomes prevalent in the fundus and the activity of gastritis increases.

Anti-secretory drugs and Helicobacter pylori: interrelations on drugs' efficacy

H. pylori interferes with efficacy of acid inhibition by proton pump inhibitors. Verdùet al. 75 demonstrated by 24-h intragastric profile that omeprazole produced greater acid suppression in H. pylori infected subjects than in non-infected subjects. H. pylori increases the acid-suppressive effects of proton pump inhibitors and direct evidence of that came from studies in which eradication of H. pylori infection resulted in a decrease in the antisecretory effect produced by omeprazole and lansoprazole, in patients with duodenal ulcer disease and in healthy subjects, respectively.76,77H. pylori exaggerates also the acid-suppressive effect of H2-RA and the cure of H. pylori infection seems to decrease the antisecretory effect of ranitidine, in prevalence at night.78,79 However H. pylori eradication and consequent corpus gastritis resolution eliminates this effect, which is more evident during proton pump inhibitors treatment than during treatment with H2-RA ranitidine. According to what we have just said it may be argued that H. pylori cure may not result in the best interest of patients with GORD requiring long-term acid suppression for the control of symptoms and oesophageal lesions. Recently Gillen et al. 80 confirmed that the presence of H. pylori infection leads to a more profound suppression of acid secretion during omeprazole treatment in healthy volunteers. Besides they refuse the hypothesis that this effect could be explained by neutralization of intragastric acid by bacterial ammonia production. What is the mechanism responsible for the effect of H. pylori infection on the reduction in acidity produced by omeprazole? Gillen et al. 80 postulate that during proton pump inhibitors treatment H. pylori redistribution occurs within the stomach, from an antral to a corpus or fundus prevalent pattern and corpus-fundus gastritis, exacerbated by omeprazole therapy, may result both in a diminished acid secretion and gastro-oesophageal reflux. Some considerations against the ‘gastritis hypothesis’ have been done. 81 First, the patients with duodenal ulcer are hypersecretors and they have not corpus gastritis, but they respond to the treatment with proton pump inhibitors. Secondly, the increased efficacy of proton pump inhibitors has been demonstrated on the first day of administration of a reversible proton pump inhibitor (By841) to H. pylori-positive subjects 82 and it is unlikely that the body gastritis can occur within few hours of proton pump inhibitors administration. In third place, after eradication of H. pylori the efficacy of omeprazole treatment comes down in a few days and does not progress, whereas gastritis continues to improve during the first months after eradication. 83 The second hypothesis to explain the greater pH raising effect of proton pump inhibitor therapy during H. pylori infection, postulates that ammonia produced by H. pylori could buffer gastric acid, with or without corpus gastritis. In fact the ammonia buffering system is effective only when H+ concentrations decrease in respond to omeprazole therapy. Moreover after eradication of H. pylori and consequent disappearance of the ammonia, the acid secretion inhibitors reduce their effectiveness. The acid reduction obtained with H2-RA is lower than that achieved with proton pump inhibitors and for this reason the ammonia buffering system has a lower effect. Finally the interaction between H. pylori and omeprazole depends on the dose and duration of the acid-inhibiting therapy. However the mechanism producing the pH elevating effect of omeprazole therapy awaits further explanations and to date there is little evidence that this effect has any clinical relevance for the maintenance treatment of GORD with the acid-inhibitors drugs. In fact during long-term treatment with omeprazole, both H. pylori-positive and -negative GORD patients require the same dose of drug. 7–84 Moreover, the effect of omeprazole and of ranitidine on acid reflux as measured by oesophageal pHmetry and by GORD symptoms, is not significantly influenced by the H. pylori85 and the rates of recurrence of symptoms and oesophagitis are very similar in both H. pylori-positive and -negative patients. 84 On the contrary Holtmann et al. 5 showed that in patients with grade II or grade III oesophagitis, healing of lesions and relief of symptoms during treatment with pantoprazole are significantly better in H. pylori infected patients than in patients without infection. At least, rebound hypersecretion might also have a determinant role. Gillen et al. 86 demonstrated that at day 15 after omeprazole, rebound hypersecretion occurs at greater rate in H. pylori-negative patients and they suggested that the underlying mechanism is that of hypergastrinaemia induced hyperplasia of ECL and parietal cells observed in patients treated with long-term omeprazole therapy. 87 It occurs early in treatment and irrespective of H. pylori status, despite higher gastrin levels in H. pylori-positive patients during omeprazole therapy. The absence of rebound increased basal acid output in H. pylori infected patients could be referred to corpus gastritis. The rebound phenomenon at withdrawal of omeprazole might be important in H. pylori-negative patients raising many questions concerning the withdrawal of acid-suppressive treatment in GORD. In conclusion, there is not sufficient evidence to say, at present, that proton pump inhibitors maintenance therapy has to be titrated on H. pylori status and therefore it could be correct the recommendation that ‘testing for H. pylori infection is not indicated in patients on long-term treatment or in those considered for treatment with a proton pump inhibitor for GORD’, as stated by the recent guidelines of the American College of Gastroenterology, 88 but until further evidence and definitive long-term data on these crucial arguments will not be available, we think it would be advisable to choose each time our behaviours in relation to other concomitant conditions.

HELICOBACTER PYLORI INFECTION and BARRETT'S OESOPHAGUS

The interest in Barrett's oesophagus is still growing since the early description of this entity in 1950 89 for two main reasons: Firstly, Barrett's oesophagus is associated with GORD, and also with an increased risk of adenocarcinoma, 90 thus representing a link between a common benign condition and a very malignant disease; Secondly, the incidence of adenocarcinoma of the oesophagus and cardia is increasing at the fastest rate among gastrointestinal human cancers. 91 Since H. pylori exhibits a special affinity for gastric-type epithelium, and since Barrett's metaplasia contains columnar-lined epithelium, it is to be expected that H. pylori will also be able to attach the Barrett's epithelium, at least of the gastric type, independently from any involvement of H. pylori infection in the pathogenesis of oesophageal mucosal inflammation. The results so far present in the literature seem to demonstrate that the prevalence of H. pylori infection of the stomach in Barrett's oesophagus patients is not different from that exhibited by controls, roughly one-third of the subjects. The colonization of metaplastic epithelium by the bacterium has been tested only in a minority of studies, but appears to be marginally lower. 6 It seems therefore that the stomach represents the primary site of infection, with secondary colonization of columnar mucosa in the oesophagus. Furthermore, most H. pylori-positive patients show a very low bacterial load in their metaplastic epithelium, and no significant difference has been found in the severity of inflammatory changes between H. pylori-positive and H. pylori-negative Barrett's oesophagus patients. 92 Finally, recent work has confirmed that within the oesophagus, H. pylori adheres only to gastric type metaplasia, which is not considered premalignant for adenocarcinoma. 93 In conclusion it is most probable that H. pylori has no aetiological role on the development of Barrett's oesophagus, nor in the oesophagitis associated with this metaplastic change and so the colonization of Barrett's epithelium probably reflects only a shift from gastric localization. Another intriguing point is the prevalence of H. pylori infection and the intestinal metaplasia of the gastric cardia; it is in fact not known at present whether inflammation of the cardia indicates GORD and/or is a manifestation of pangastritis caused by H. pylori. 94 Recently two studies have shed some light on this issue. 40–95 In the first study, biopsies were obtained from the antrum, corpus and cardia from 135 H. pylori infected patients with gastritis, ulcer disease, or reflux oesophagitis. 132/135 (97.7%) of them showed active carditis, resembling antral gastritis in most patients, but with less marked bacterial density and inflammatory process. 40 The authors conclude that H. pylori gastritis commonly involves the cardia and that intestinal metaplasia in the cardia is a common finding in H. pylori gastritis. It is unclear the reason why the cardia histological density of the bacteria and inflammatory responses are lower than in the antrum. In the second work, 95 22 GORD patients and 11 controls were compared in relationship to endoscopic and bioptic evaluation of inflammation, H. pylori infection and intestinal metaplasia in distal oesophagus, cardia, fundus and antrum. It turned out that neither the prevalence of H. pylori infection (controls 48%; GORD 41%) nor cardia inflammation (controls 41%; GORD 40%) differed between the two groups. All 11 controls and 22 of 23 patients with GORD (96%) who had cardia inflammation had H. pylori infection too. Cardia intestinal metaplasia was more common among controls (22%) than among GORD patients (3%, P < 0.01); all patients with cardia intestinal metaplasia had cardia inflammation, seven had H. pylori infection, and six had metaplasia elsewhere in the stomach. The authors conclude that the prevalence of cardia inflammation is similar in patients with and without GORD, and is associated with H. pylori infection. Also, in this study, cardia intestinal metaplasia is associated with H. pylori-related cardia inflammation (P = 0.01) and intestinal metaplasia elsewhere in the stomach, indicating that it is distinct from Barrett's oesophagus. The final point is the association, if any, between H. pylori infection and Barrett's associated adenocarcinoma. Again, two recent works have contributed to the improvement of our knowledge on this previously uninvestigated issue.96,97 Quddus et al. 96 report on 19 cases of adenocarcinoma arising in Barrett's oesophagus, which were examined for the presence of H. pylori after staining with three different techniques: all sections of Barrett's oesophagus, with or without dysplasia, adenocarcinoma and stomach (when available) were uniformly negative for the presence of H. pylori. The authors conclude that neither gastric nor oesophageal infection with H. pylori is a requisite for the development of adenocarcinoma in Barrett's oesophagus. 96 The second study aimed at comparing the prevalence of H. pylori and increasing grades of dysplasia. Biopsies from 19 malignant and 94 benign cases of Barrett's oesophagus were analysed histologically for H. pylori; 34% of non-dysplastic Barrett's epithelium was colonized with H. pylori compared with only 17% of dysplastic/malignant cases (P = 0.04). No relationship was found between H. pylori status and (a) length of Barrett's oesophagus; (b) the presence of strictures or ulcers; (c) previous anti-reflux surgery. The authors therefore confirmed that H. pylori, colonization of Barrett's oesophagus is not particularly common, and that a negative correlation exists with increasing severity of dysplasia. 97 To summarize, from both studies it appears that it is unlikely that a causal relationship exists between H. pylori infection and Barrett's-associated adenocarcinoma.

Conclusion

In conclusion the relationship between H. pylori infection and GORD is complex, as yet not fully elucidated, and probably based on a multiplicity of factors. Although H. pylori may in theory represent a cause for GORD, from the available data it is likely that H. pylori could even represent a protective factor for GORD, in particular if and when corpus gastritis is present. 98 Investigating and treating H. pylori infection is not recommended in patients with GORD that do not need proton pump inhibitors long-term treatment. The need for H. pylori eradication before embarking on long-term proton pump inhibitor treatment for reflux oesophagitis is at present not completely demonstrated, but same studies reported that in presence of H. pylori infection, gastric acid inhibition could induce an atrophic gastritis with theoretic risk of developing gastric cancer. However, because several arguments pros and cons are reported, H. pylori eradication in GORD patients who need long-term therapy cannot be considered as definitively established. There is evidence suggesting that H. pylori increases the efficacy of proton pump inhibitors, while the cure of H. pylori infection seems to decrease the antisecretory effect of proton pump inhibitors and H2-RA ranitidine. According to some authors,99,100 if some types of H. pylori strains are clearly virulent, while others may have neutral or beneficial role, it is advisable to make a critical re-evaluation regarding to world-wide elimination of this bacterium. Finally, it seems that H. pylori infection of Barrett's epithelium (gastric-type metaplasia) is possible as an epiphenomenon of gastric colonization. Therefore, H. pylori is of no pathogenetic relevance for Barrett's oesophagus development, progression and cancerization. Furthermore, cardia gastritis or ‘carditis’ seems to be equally frequent in patients with and without GORD, and cardia intestinal metaplasia seems not to be an objective marker of Barrett's epithelium.

Ancillary