Review article: Helicobacter pylori and gastro-oesophageal reflux disease—clinical implications and management

Authors

  • O’connor

    1. Department of Medicine, General Hospital, Tullamore, Co. Offaly, Ireland; and Faculty of Medicine, University College Dublin, Earlsfort Terrace, Dublin, 2, Ireland
    Search for more papers by this author

O’connor Consultant Physician/Gastroenterologist, Department of Medicine, General Hospital, Tullamore. Co. Offaly, Ireland.

Abstract

A significant proportion of patients with gastro-oesophageal reflux disease (GERD) have Helicobacter pylori infection, but it is unclear whether or not H. pylori should be treated in this clinical setting. The aim of this review was to critically assess the relationship between H. pylori and GERD and its potential implications for the management of GERD. Data for this review were gathered from the following sources up to April 1998—the biomedical database M EDLINE, a detailed review of medical journals, and a review of abstracts submitted to relevant international meetings.

On average, 40% of GERD patients carry H. pylori infection, with a reported infection prevalence ranging from 16% to 88%. To date, there has been no reported controlled trial of effective H. pylori therapy in GERD. GERD has been reported to develop de novo following the cure of H. pylori in peptic ulcer disease.

In the presence of H. pylori, proton pump inhibitor therapy appears to accelerate the development of atrophic corpus gastritis, a potentially precancerous condition. Conversely, proton pump inhibitor therapy seems to become less effective after cure of H. pylori. The mechanisms underlying these important contrasting phenomena are poorly understood.

The relationship between H. pylori and GERD is complex, and it is difficult to give definitive guidelines on the management of H. pylori infection in GERD. Controlled trials of H. pylori therapy in GERD are urgently needed, as well as further long-term data on both the natural history of gastric histopathological changes in the H. pylori-positive GERD patient treated with proton pump inhibitors, and the impact of H. pylori status on the clinical efficacy of antisecretory therapy. Pending these data, it is perhaps advisable to advocate cure of H. pylori in young patients with proton pump inhibitor-dependent GERD who, in the absence of anti-reflux surgery, are faced with the likelihood of long-term medical therapy.

INTRODUCTION

Gastro-oesophageal reflux disease (GERD) is a chronic relapsing condition associated with significant morbidity.1, 2 The prevalence of GERD appears to be on the increase, while that of peptic ulcer disease is waning. In a study of hospitalization rates in the USA between 1970 and 1995, El Serag & Sonnenberg3 found a fourfold drop in the rate for duodenal ulcer, compared with a sevenfold rise in that for erosive oesophagitis. Similar opposing time-trends were found for gastric cancer (twofold drop) and oesophageal cancer (sevenfold rise). The rising prevalence of GERD, coupled with the falling prevalence of peptic ulcer disease has also been noted after an analysis of endoscopy records over the past two decades.4, 5 A falling prevalence of H. pylori infection in developed countries has been cited as a possible reason for these epidemiological phenomena.

Previously, the presence or absence of H. pylori infection in patients with GERD has largely been ignored. A number of recent developments have changed that perception and generated considerable interest in the relationship between H. pylori and GERD.6, 7 A variable proportion of patients with GERD have H. pylori infection,8 but it is unclear whether the symptoms or natural history of GERD are influenced by curing H. pylori infection. There have been reports that GERD can develop de novo following the eradication of H. pylori in peptic ulcer disease.9 There is evidence that in the presence of H. pylori infection, proton pump inhibitor therapy—the mainstay of treatment in many GERD patients—may accelerate the development of atrophic corpus gastritis, a potentially precancerous condition.10 On that basis, it is argued that H. pylori should be eradicated in GERD patients. Ironically, the efficacy of proton pump inhibitors appears to diminish in the H. pylori-negative stomach.11 Finally, the predilection of H. pylori for gastric-type mucosa has sparked an interest in whether the infection is linked with the pathogenesis of Barrett’s oesophagus and, by inference, with oesophageal carcinoma. Given these considerations, the aim of this review was to critically assess the relationship between H. pylori and GERD, with a particular emphasis on potential implications for the management of GERD.

METHODS

Data for this review were gathered from the medical literature available to April 1998. The biomedical database M EDLINE was searched using the keywords H. pylori and GERD. Studies were identified from a detailed review of medical journals, including Gastroenterology, Gut, European Journal of Gastroenterology and Hepatology, Alimentary Pharmacology and Therapeutics, American Journal of Gastroenterology, Scandinavian Journal of Gastroenterology, Lancet, British Medical Journal, and New England Journal of Medicine. Relevant abstracts were identified from those submitted to meetings of the American Gastroenterological Association, United European Gastroenterology Week, the annual Workshops of the European HelicobacterPylori Study Group, and the British Society of Gastroenterology.

H. PYLORI AND GERD

Table 1 summarizes the results of 26 studies that assessed the prevalence of H. pylori infection in adults with GERD.8, 12[13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30]–36 Of the 2112 GERD patients studied, 851 (40.3%) were H. pylori-positive, with an infection prevalence ranging from 16% to 88%. The prevalence of H. pylori in GERD increased with age, indicating an age cohort effect.8, 36 There was no difference between the sexes, and no significant difference in prevalence between the different grades of oesophagitis.8, 23, 28 In 13 case-control studies, 562 of 1426 GERD patients were H. pylori-positive (39.4%) compared with 1009 of 2010 control subjects (50.2%),13, 15- [16]17, 20, 22, 24, 26, 27, 29, 31, 34, 36 the difference in prevalence perhaps intimating that the pathogenesis of GERD might be related in some way to the absence of H. pylori. Hackelsberger et al.22 found that H. pylori-positive GERD patients had lower reflux grades than H. pylori-negative patients. while Sekiguchi et al.33 found H. pylori in 6 of 12 patients with mild reflux oesophagitis, but in none of 9 with severe reflux oesophagitis. Mihara et al.27 and De Koster et al.17 assessed the severity of histological gastritis and atrophy in relation to GERD, and found lower gastritis and atrophy scores in patients with GERD compared with controls, suggesting that the preservation of acid secretion might be the basis for the putative link between the absence of H. pylori infection and GERD.

Table 1.  .  Prevalence of H. pylori infection in GERD Thumbnail image of

In support of this hypothesis, Schenk et al.,32 in a prospective study of 137 GERD patients, found that 88 H. pylori-negative patients presented with more severe oesophagitis and a significantly higher prevalence of Barrett’s oesophagus. However, there was no significant difference between the H. pylori-positive and -negative groups in the minimum dose of omeprazole required to relieve symptoms and heal oesophagitis. Schnell et al.37 evaluated the prevalence of infection with H. pylori cag A+ strains in GERD and found that Cag A seronegativity was strongly associated with severe reflux disease, Barrett’s oesophagus, and dysplasia/adenocarcinoma. In addition, the need for long-term treatment with omeprazole was significantly associated with the presence of Cag A seronegativity. Awad et al.38 performed 24-h ambulatory combined oesophageal pH and manometry on 37 GERD patients, 29 of whom were H. pylori-positive (78%), and found no significant differences between the H. pylori-positive and -negative groups, although there was a trend towards lesser oesophageal acid exposure in H. pylori-positive patients.

In paediatric studies, Stewart et al.39 found H. pylori infection in only 3 of 80 children with GERD (4%) compared with 15 of 109 (16%) in a retrospective study by Rosioru et al.40 The latter study also found that the prevalence of oesophagitis was similar in the H. pylori-positive and -negative groups and the response to antisecretory therapy was independent of H. pylori status. In a prospective study of 173 children with GERD, Cargill et al.41 found H. pylori infection in 13 (7.5%) and—contrary to some adult studies—found more severe oesophagitis in H. pylori-positive patients. In a case-control study in the elderly, with an average age of 79 years, Liston et al.24 were unable to find an association between H. pylori and GERD.

There is a paucity of data on the clinical efficacy of H. pylori eradication therapy in GERD. For instance, a recent review of critical issues in the management of GERD could only quote anecdotal evidence that some GERD patients infected with H. pylori experienced symptom relief following eradication therapy.42 In a novel, controlled, double-blind study reported a decade ago, Borkent & Beker14 randomised 20 patients with reflux oesophagitis to treatment with colloidal bismuth and cimetidine or cimetidine alone. Cimetidine with colloidal bismuth gave significantly better results than cimetidine alone, hinting at the possibility that cure of H. pylori might be of benefit in GERD. In a small study reported in abstract form, Galimov et al.19 found a positive therapeutic effect from anti-Helicobacter therapy (not specified) in 34 of 43 GERD patients (78%). Controlled trials of H. pylori therapy in GERD are urgently needed in order to clarify what impact, if any, the cure of H. pylori infection might have on the symptoms and natural history of GERD.

GERD AND H. PYLORI CURE

During a study of possible H. pylori reinfection from patients’ spouses, Schutze et al.43 noted that 10 of 16 duodenal ulcer patients who had not been reinfected developed reflux oesophagitis during the follow up period of almost 4 years, raising, for the first time, the possibility that GERD might arise de novo following cure of H. pylori. Labenz et al.9 carefully followed up 450 duodenal ulcer patients treated for H. pylori without reflux oesophagitis at baseline endoscopy. H. pylori infection was cured in 244 patients and persisted in 216, and all patients underwent annual endoscopy or when symptomatic. From life table analysis, the estimated incidence of oesophagitis within 3 years was 25.8% after H. pylori eradication, compared with 12.9% in patients with persistent infection. Factors that predicted the development of reflux oesophagitis were: severity of corpus gastritis before cure, weight gain after cure, and male sex. Six months after H. pylori eradication therapy, Sacca et al.44 found mild reflux oesophagitis in 24 of 169 ulcer patients (14.2%) while Di Mario45 found oesophagitis in 10 of 202 ulcer patients (4.9%) 1 year after H. pylori cure.

It should be remembered however, that a variable proportion of peptic ulcer patients have coexistent GERD46- [47][48]49 and the phenomenon of new GERD after H. pylori cure could be explained, at least in part, by the unmasking of pre-existent disease. In a study where the endoscopic grade of GERD was systematically recorded before eradication therapy in 244 H. pylori-positive peptic ulcer patients, 49 (20%) had GERD, which tended to improve or remain stable in the short term after H. pylori therapy.50 In patients with a normal oesophagus before therapy, 7% showed evidence of reflux oesophagitis after H. pylori cure.

A number of possible mechanisms might account for the phenomenon of rebound reflux after H. pylori cure. Ammonia produced by H. pylori, including infection at the gastric cardia20, 51 or in a hiatal hernia,8 could buffer the gastric juice refluxing into the oesophagus. This neutralising effect would be lost after H. pylori cure. Healing of corpus gastritis with a consequent increase in acid secretory capacity after H. pylori cure is another possible pathogenetic mechanism, but the fact that acid output in duodenal ulcer patients decreases rather than increases in the first year after cure of infection is an argument against this.52 Labenz et al.9 found that weight gain was an independent risk factor for post-eradication reflux, but there is controversy over whether obesity per se plays a role in promoting GERD.53 Gastrin may modulate lower oesophageal sphincter pressure in healthy subjects,54 and it is conceivable that the fall in serum gastrin seen following H. pylori cure might in some way trigger gastro-oesophageal reflux. In essence, the extent and causes of rebound reflux after H. pylori cure are largely unknown and prospective studies are required which should include the measurement of oesophageal pH and manometry before and at defined time points after H. pylori therapy.

CHRONIC GASTRITIS AND ANTISECRETORY THERAPY

One of the arguments in favour of eradication of H. pylori in GERD is the suggestion that long-term proton pump inhibitor therapy in H. pylori-positive patients may be associated with rapidly progressive atrophic gastritis of the gastric body, and with it an increased risk of gastric cancer.10 Large scale long-term follow up studies have firmly established atrophic gastritis as a gastric cancer precursor lesion.55H. pylori is the main cause of chronic gastritis world-wide, which can lead in time to atrophic gastritis.

Valle et al.56 retrospectively assessed the progress of H. pylori infection and chronic gastritis in 102 patients followed up for 32 years, and among the 85 patients who were initially H. pylori-positive, the prevalence of corpus atrophic gastritis rose from 13 (15%) to 32 (38%) at the end of the follow up. Among the 17 patients who were initially H. pylori-negative, the prevalence of atrophic gastritis rose from 0 to 2 (12%). In a study of the evolution of histological changes over a period of 10 years in 34 H. pylori-positive subjects, Niemala et al.57 found a significant decrease in the severity of antral gastritis accompanied by a significant increase in corpus gastritis, but without a significant change in the density of H. pylori infection. In a 12-year endoscopic follow up study in a population-based random sample of 81 H. pylori-positive Estonians, Villako et al.58 found the mean prevalence of atrophic gastritis was threefold more common in the corpus than in the antrum. Similarly, Kuipers et al.59 found the frequency of atrophic corpus gastritis in 58 H. pylori-infected patients increased, over a mean 11.5 years, from 24% to 45%, with an annual rate of increase of almost 2% as opposed to 0.3% in uninfected subjects.

Proton pump inhibitors are powerful antisecretory agents, for instance, omeprazole 40 mg can suppress over 80% of gastric acid secretion.60 Early studies that assessed histopathological changes in the gastric mucosa in patients on long-term omeprazole showed a progression of corpus gastritis in some patients.

In 74 patients with peptic ulcer disease treated chronically with omeprazole, Lamberts et al.61 found that atrophic gastritis increased from 1.8% to 20.8% after 5 years. Dent et al.62 found that of 16 GERD patients who had gastritis, 9 had progressed to atrophic gastritis after 1 year of maintenance treatment with omeprazole. In a prospective study of 91 patients with severe reflux oesophagitis treated with 20–40 mg omeprazole daily for a mean period of 48 months, Klinkenberg-Knol et al.63 found that the incidence of atrophic gastritis increased from less than 1% to 25% and that the progression of gastritis was more pronounced in patients with high serum gastrin levels. In 202 patients treated with omeprazole for at least 330 days, Solcia et al.64 found the incidence of chronic atrophic gastritis increased from 1% to 13%.

Subsequent studies suggest that proton pump inhibitor therapy-related changes in the topography and severity of chronic gastritis may be largely confined to H. pylori-positive subjects. Kuipers et al.10 evaluated 50 patients before and at the end of 8 weeks therapy with omeprazole 40 mg daily. Before therapy, 33 H. pylori-positive patients showed predominant colonization and inflammation of the antrum, whereas after therapy, infection and inflammation predominantly affected the gastric corpus. Seventeen H. pylori-negative patients showed no histological evidence of gastritis before or after therapy. Logan et al.65 showed similar proximal migration in the intragastric distribution of H. pylori during treatment with omeprazole. In a subsequent study in patients with reflux oesophagitis treated with omeprazole or fundoplication, Kuipers et al.66 found that among the patients treated with omeprazole, atrophic corpus gastritis developed in 18 of 59 H. pylori-positive patients (30.5%) compared with 2 of 46 H. pylori-negative patients (4.3%). In contrast, among the patients treated surgically with fundoplication, atrophic gastritis did not develop in any of the 31 H. pylori-positive patients or in the 41 −negative patients. These results gave a calculated annual rate of increase of atrophic gastritis of 6.1% in H. pylori-positive GERD patients treated with omeprazole compared with 0% in those treated by anti-reflux surgery. Kuipers’ study can be criticised as it was not a randomised prospective design, the patient cohorts were drawn from different countries and the mean age of the Dutch omeprazole-treated group was 62 years compared with 53 in the Swedish surgically treated group. Lundell et al.67 recently reported, in abstract form, the results of a controlled study with long-term follow up where 155 patients with GERD were randomised to treatment with omeprazole, and another 154 to anti-reflux surgery. The patient groups were well matched and had annual endoscopies and gastric biopsies over a median follow up period of 3 years. In contrast to Kuipers et al.,66 this study did not find evidence of accelerated development of atrophic gastritis in patients on long-term omeprazole.

Berstad et al.68 assessed H. pylori gastritis and epithelial cell proliferation in 90 GERD patients after long-term treatment with lansoprazole and, in agreement with Kuipers et al.,66 found a marked increase in corpus gastritis in H. pylori-positive patients. Epithelial cell proliferation also increased significantly in infected but not uninfected patients. In a similar study, Eissele et al.69 assessed the gastric mucosa in 42 patients treated daily with 30–90 mg lansoprazole for up to 5 years and found that chronic inflammation and atrophy of the corpus mucosa worsened exclusively in H. pylori-positive patients. They also found a 10-fold increase in the prevalence of argyrophil cell hyperplasia after long-term lansoprazole which was significantly related to H. pylori infection, chronic inflammation, and atrophy of the corpus mucosa. It should be noted that a worsening of H. pylori gastritis with antiacid therapy may not be confined to proton pump inhibitor therapy. Meining et al.70 have shown that treatment with ranitidine or aluminium-magnesium hydroxide aggravates the grade and activity of H. pylori gastritis in duodenal ulcer patients.

The mechanisms underlying progression of corpus gastritis associated with H. pylori infection and antisecretory therapy remain unclear. The pathological changes seem to occur despite the absence of a significant increase in the density of H. pylori infection in the fundic part of the stomach.10, 71 Perhaps a reduction of acid levels allows increased ingress of bacterial toxins, such as ammonia, with resultant mucosal damage.72 A deeper location of H. pylori in the gastric glands might also enhance mucosal damage.73 Gastrin exerts trophic effects on the gastric mucosa74 and both H. pylori infection and proton pump inhibitor therapy cause chronic hypergastrinaemia in humans.75 Interestingly, El-Nujumi et al.76 have shown that eradicating H. pylori reduces the rise in gastrin seen during subsequent long-term omeprazole therapy. Antigenic mimicry exists between H. pylori and gastric mucosa, and H. pylori-mediated autoimmunity is probably of fundamental importance in the pathogenesis of corpus atrophic gastritis.77 It is tempting to speculate that hypoacidity might in some way influence the grade of antigenic mimicry of H. pylori and thus the behaviour of chronic gastritis. There is also a link between HLA-DQA genotype and the incidence of atrophic gastritis in H. pylori infection,78 and differences in HLA genotype could be an important host factor influencing the mucosal response under hypoacidic conditions.

Whether the above data indicate a need for universal eradication of H. pylori in GERD patients who are taking, or about to start, proton pump inhibitor therapy is a moot point. Firstly, the long-term natural history of any mucosal changes related to proton pump inhibitor therapy in the presence of H. pylori infection needs to be assessed and any relevant clinical risk documented. More importantly, the value of eradicating H. pylori in proton pump inhibitor-dependent GERD should be measured in a controlled study.

H. PYLORI AND EFFICACY OF ANTISECRETORY THERAPY

Any discussion about the relationship between H. pylori and GERD should include reference to the possibility that the presence of H. pylori and its eradication might change the therapeutic efficacy of antisecretory agents, particularly proton pump inhibitors. The idea that H. pylori status might influence the efficacy of proton pump inhibitors was initially hinted at by reports that during treatment with omeprazole, the intragastric pH was higher in patients with duodenal ulcer than in healthy controls.79, 80 In an important study that tested the hypothesis that H. pylori infection was associated with increased intragastric pH during omeprazole therapy, Verdu et al.11 performed ambulatory 24 h gastric pH recordings in 18 H. pylori-positive and 14 H. pylori-negative healthy subjects who received 1-week courses of placebo or omeprazole, 20 mg daily, with a 4–6 week wash-out period between treatments. During the placebo phase, median 24 h pH values were not different between the H. pylori-positive and -negative subjects. However, during omeprazole treatment, median 24 h pH values were significantly higher in H. pylori-positive subjects, both in the gastric corpus and antrum. The same workers went on to study whether H. pylori also augmented the pH-increasing effect of omeprazole in patients with duodenal ulcer and assessed the impact of cure of infection. H. pylori eradication resulted in a marked decrease of the pH-increasing effect of omeprazole in ulcer patients, an effect which was most apparent at night-time.81 In contrast, baseline intragastric pH remained unchanged after cure of infection. In a one year follow-up study, Labenz et al.82 showed that the effect of curing H. pylori on the pH response to omeprazole persisted over time. Ranitidine also loses efficacy on nocturnal intragastric acidity after cure of H. pylori, but the effect seems less than that on proton pump inhibitor efficacy.83

In an attempt to assess the influence of H. pylori infection on the clinical response to antisecretory therapy, Carlsson et al.84 evaluated data from three controlled randomised double blind studies involving 1350 patients with GERD, and found that the 36% of patients infected with H. pylori had a significantly lower risk of relapse during maintenance treatment with omeprazole compared with the H. pylori-negative patients. Healing of oesophagitis and relief of heartburn was similar in H. pylori-positive and -negative patients during treatment. The clinical implication to be taken from these data is that the patient effective dose of proton pump inhibitors in GERD may have to be increased in the absence of H. pylori infection.

There are a number of reasons why proton pump inhibitor therapy might be less effective in the H. pylori-negative stomach. H. pylori infection causes corpus gastritis, which may impair secretory function and, as already discussed, proton pump inhibitor therapy in the presence of H. pylori may worsen mucosal inflammation in the gastric corpus.10 Corpus gastritis is also known to be associated with loss of M3 muscarinic receptors that mediate acid secretion.85H. pylori generates ammonia and other amines, leading to a higher concentration of acid neutralising substances in the stomach. H. pylori infection may also increase alkaline duodenogastric reflux.86H. pylori infection may induce nitric oxide synthase and nitric oxide with the capacity to inhibit acid secretion.87In vitro, H. pylori releases acid inhibiting substances but their impact on in vivo acid secretion remains unclear.88H. pylori gastritis is associated with decreased antral somatostatin and increased gastrin release.75, 89 The rise in gastrin which is seen during treatment with omeprazole is higher in H. pylori-positive subjects than in H. pylori-negative controls.90 Labenz & Malfertheiner91 have hypothesized that gastrin-mediated parietal cell stimulation could decrease intracanalicular pH, which in turn might lead to both increased transformation of omeprazole to its active moiety and also to an increase in the number of active proton pumps susceptible to inhibition.

H. PYLORI AND BARRETT’S OESOPHAGUS

H. pylori has a distinct predilection for gastric-type mucosa.92 In Barrett’s oesophagus, the normal squamous mucosa is replaced by the specialized columnar epithelium and this metaplastic change is thought to occur in response to prolonged gastro-oesophageal reflux.93 Barrett’s oesophagus is primarily important because of its malignant potential, conferring a 30 to 120-fold increased risk of oesophageal carcinoma compared to the general population.94 Hence, the relationship between H. pylori and Barrett’s oesophagus has been assessed on the premise that gastric-type epithelium in the oesophagus might be susceptible to colonization and that infection might in some way enhance the malignant potential of Barrett’s oesophagus.

Table 2 summarizes the results of 19 studies8, 16, 23, 28, 34, 35, 95- [96][98][100][104][105]107 that assessed the prevalence of H. pylori infection in adults with Barrett’s oesophagus. Of 818 patients studied, 240 (29.3%) had evidence of H. pylori colonising Barrett’s mucosa, with an infection prevalence ranging from 0% to 88%. In nine studies,8, 16, 28, 35, 95, 97, 99, 102, 103 gastric mucosa was also assessed, and 96 of 317 patients (30.1%) were H. pylori-positive with an infection prevalence ranging from 13% to 63%. Five studies16, 28, 34, 95, 102 assessed gastric biopsies from appropriate control groups and of 323 patients studied, 91(28.8%) were H. pylori-positive. H. pylori in Barrett’s oesophagus is almost always associated with gastric H. pylori infection, suggesting that the stomach is the likely primary site of infection, with secondary colonization of columnar mucosa in the oesophagus. Colonization of Barrett’s epithelium is generally low grade101 and no significant difference has been found in the severity of inflammatory changes between H. pylori-positive and -negative Barrett’s oesophagus.95, 101 In addition, there is no apparent relationship between H. pylori status and the presence of Barrett’s ulcer, oesophageal strictures, or the total length of Barrett’s oesophagus.102, 106

Table 2.  .  Prevalence of H. pylori infection in Barrett’s oesophagus Thumbnail image of

Weston et al.107 prospectively assessed the presence of mucosa associated lymphoid tissue (MALT) in 139 patients with Barrett’s oesophagus and found that the prevalence of gastric and oesophageal H. pylori in patients with Barrett’s MALT was significantly higher compared with patients with Barrett’s oesophagus without MALT. The MALT lesion in Barrett’s mucosa was focal, and no patients had MALT lymphoma. In a study of the prevalence of premalignant markers in Barrett’s oesophagus, Young et al.108 assessed the presence of dysplasia, overexpression of the tumour suppressor gene P53, and type 3 intestinal metaplasia in multiple biopsies from 37 Barrett’s patients, 6 of whom had H. pylori infection. Dysplasia and overexpression of P53 were significantly commoner in the patients with H. pylori infected Barrett’s mucosa and all these patients had type 3 intestinal metaplasia. Thus, while H. pylori infection would appear to play little or no role in the development of Barrett’s oesophagus, the possibility that H. pylori might be important in the pathogenesis of malignant transformation of Barrett’s epithelium cannot be discounted and further studies are warranted.

CONCLUSIONS

The relationship between H. pylori and GERD is complex and confusing. A significant proportion of GERD patients have coincident H. pylori infection but there are little or no data on what impact a cure of the infection has on the symptoms or natural history of GERD. Post-eradication rebound GERD in ulcer patients raises the possibility that GERD might worsen after cure of H. pylori. In the presence of H. pylori, proton pump inhibitor therapy appears to accelerate the development of corpus gastritis and this undesirable effect is an argument for curing H. pylori in GERD patients. On the other hand, proton pump inhibitor therapy seems to become less effective after cure of infection. The mechanisms underlying these important contrasting phenomena are poorly understood. Finally, H. pylori can colonize Barrett’s oesophagus with possible effects on the malignant potential of this condition.

At present, it is difficult to give definitive guidelines on the management of H. pylori infection in GERD. Controlled trials of eradication therapy in H. pylori-positive GERD are urgently needed, as well as further long-term data on both the natural history of gastric histopathological changes in the H. pylori-positive GERD patient treated with proton pump inhibitors, and the impact of H. pylori status on the clinical efficacy of antisecretory therapy. Pending these data, it is perhaps advisable to advocate a cure of H. pylori in some GERD patients, for instance, young patients with proton pump inhibitor-dependent GERD who, in the absence of anti-reflux surgery, are faced with the likelihood of long-term medical therapy.

ACKNOWLEDGEMENT

Sincere thanks are due to Mrs Sandra Keating, Medical Librarian, General Hospital, Tullamore, for her help in collating the references, and to Mrs Iris Martin for typing the manuscript.

Ancillary