Helicobacter pylori and reflux disease
Correspondence to: Dr P. Sharma, Gastroenterology Section (111), Department of Veterans Affairs Medical Center, 4801 E. Linwood Blvd., Kansas City, MO 64128-2295, USA. E-mail: email@example.com
The falling prevalence of Helicobacter pylori infection and related diseases (peptic ulcer disease, gastric cancer) in developed countries has been paralleled by an increased recognition of gastro-oesophageal reflux and its complications. These epidemiological data do not support a role for H. pylori in the pathogenesis of reflux disease, but suggest a negative association with the increasing incidence of oesophageal diseases. This has led some investigators to propose a ‘protective’ role of H. pylori infection against the development of oesophageal diseases. In these patients, pre-existing lower oesophageal sphincter dysfunction, susceptibility to reflux, unmasking of latent reflux and the patterns and severity of gastritis are probably important factors contributing to the development of oesophageal diseases. The most likely mechanism by which H. pylori infection may protect against reflux is by decreasing the potency of the gastric refluxate in patients with corpus-predominant gastritis. The prevalence of H. pylori infection in patients with reflux disease is probably no greater than that in those without reflux, and there are conflicting data indicating that reflux symptoms or erosive oesophagitis develop after H. pylori eradication. It is also unclear whether H. pylori augments the antisecretory effects of proton pump inhibitors or accelerates the development of atrophic gastritis.
Helicobacter pylori infection leads to chronic active gastritis with the potential for the subsequent development of peptic ulcer disease, mucosa-associated lymphoid tissue lymphoma and distal gastric cancer in some patients. Host and environmental factors, as well as features of the infecting bacterial strain, determine the development of H. pylori -associated gastric diseases. 1 Patients with current/past peptic ulcer disease and those with early stages of mucosa-associated lymphoid tissue lymphoma are treated with H. pylori eradication therapy; however, the treatment of H. pylori in patients with reflux disease remains highly controversial. 2 Gastro-oesophageal reflux disease refers to the reflux of gastric contents into the oesophagus, leading to mucosal damage and/or symptoms of heartburn and regurgitation. These symptoms are extremely common; they are experienced by more than 40% of the adult US population on a monthly basis, but only 30–40% of those with reflux symptoms have erosive oesophagitis on endoscopy. 3, 4 A number of pathophysiological mechanisms have been reported in patients with reflux, including abnormal transient lower oesophageal sphincter relaxation, low lower oesophageal sphincter pressures, abnormal oesophageal motility, impaired oesophageal mucosal defences and the presence of hiatus hernia, etc., and attempts are now being made to add H. pylori to this growing list. Inadequate oesophageal clearance of refluxed material and the potency of the refluxate may be additional factors contributing to gastro-oesophageal reflux. H. pylori may influence the potency of the gastric refluxate by affecting the patterns of gastritis and gastric acid secretion. A number of other factors have been postulated, such as ammonia released by H. pylori (could potentially neutralize gastric acid), corpus gastritis associated with lower acid output, which then returns to normal when H. pylori is eradicated, and hypergastrinaemia leading to increased lower oesophageal sphincter pressure. 5, 6 This paper reviews the complex relationship between H. pylori and reflux.
Epidemiology of H. pylori and reflux disease
Although the prevalence of H. pylori is steadily decreasing in industrialized nations, probably due to improved hygiene and socio-economic factors, approximately one-half of the world's population is still colonized with H. pylori.7 On the other hand, reflux symptoms are extremely common in adults in Western countries, with approximately 8% of the US population experiencing heartburn on a daily basis.4 Thus, both reflux symptoms and H. pylori are common entities, and their prevalence is influenced by factors such as ethnicity, socio-economic status and advancing age. In Western countries, there has been a decrease in the incidence of H. pylori and related conditions [peptic ulcer disease and gastric (body and antrum) adenocarcinoma], but an increase in hospitalizations for reflux-related erosive oesophagitis and in the incidence of oesophageal adenocarcinoma.8 This has led H. pylori to be cited as a possible aetiological factor for this changing epidemiology.
Most studies have found no causal relationship between H. pylori infection and gastro-oesophageal reflux disease.9–12 Some reports have indicated a lower prevalence of H. pylori infection in patients with reflux symptoms, and have proposed that H. pylori reduces the risk of reflux oesophagitis. This negative association with the severity of oesophagitis has added weight to the theory of a protective role, but there are also reports to the contrary.9–12 Similarly, epidemiological studies have found either a small negative or no association between H. pylori infection and reflux; similar rates of H. pylori infection have been reported in patients with reflux (symptoms and/or erosive oesophagitis) and in controls.
In a study of 514 patients by Varanasi et al., the prevalence of H. pylori infection was 30.7% in patients with reflux oesophagitis and 42% in patients without oesophagitis.13 The role of H. pylori infection, gastric atrophy and gastric acid secretion was studied in 105 patients with and without erosive oesophagitis. The prevalence of H. pylori infection was significantly lower in patients with erosive oesophagitis compared to controls (34% vs. 76%), with acid secretion being significantly greater in patients with erosive oesophagitis, and correlated with the degree of gastric atrophy, i.e. higher atrophy score in patients without erosive oesophagitis.14 In a systematic review, O'Connor reported the overall prevalence of H. pylori to be 40% in gastro-oesophageal reflux disease patients (pooled data from 26 studies).15 However, in 13 case-controlled studies, only 39.4% of gastro-oesophageal reflux disease patients were H. pylori-positive compared with 50% of control subjects.
Studies from Asia, where the prevalence of H. pylori is still relatively high, have also reported conflicting results. To evaluate the relationship between H. pylori and reflux oesophagitis in a large number of Japanese subjects, Yamaji et al. measured serum H. pylori antibody and pepsinogen (surrogate marker of gastric atrophy) in 5732 patients with erosive oesophagitis, and demonstrated that both positivity for H. pylori antibody and low pepsinogen (suggestive of gastric atrophy) were negatively associated with the presence of reflux oesophagitis.16 Wu et al. found the prevalence of H. pylori infection to be lower in gastro-oesophageal reflux disease patients (31%) than in age- and sex-matched asymptomatic controls (61%) in Hong Kong.12 However, at Digestive Disease Week 2002, Watanabe et al. reported that only 2% of a randomly selected Japanese population (with an H. pylori prevalence of 44%) experienced daily reflux symptoms, and there was no association between the prevalence of H. pylori and gastro-oesophageal reflux disease.17
Overall, although studies have shown a low prevalence of H. pylori in patients with reflux, it is unclear whether there is a true inverse relationship between H. pylori and the prevalence of reflux. Some studies have examined the relationship of H. pylori prevalence (documented either by histology, serology or breath testing) with either symptom prevalence or the presence of erosive oesophagitis. A number of factors, including the lack of use of validated reflux symptom questionnaires, the definition of reflux, the method of H. pylori detection, inappropriate control groups and the lack of documentation of patterns of gastritis (i.e. antral-predominant vs. pan-gastritis), probably account for the discrepancies between the various published reports.
H. pylori infection and gastric acid secretion
H. pylori infection can have varying effects on gastric acid secretion. These are related to the distribution of gastritis in the stomach. H. pylori infection can increase, decrease or leave gastric acid secretion unchanged. Changes in acid secretion depend on the severity and location of inflammation. H. pylori infection can cause a gastritis that is located predominantly in the antrum, or can result in a more diffuse gastritis in which the corpus or body of the stomach is predominantly involved. Antrum-predominant gastritis has been shown to be associated with hypergastrinaemia, gastric hypersecretion and duodenal ulcer disease, and the eradication of the organism is associated with the correction of these abnormalities. 18 One month after the eradication of H. pylori in infected patients with duodenal ulcer disease, there are significant reductions in basal plasma gastrin concentration and in basal acid secretion. Other studies have examined gastric acid secretion and its response to gastrin-releasing peptide in healthy volunteers and patients with ulcer disease. 19 Patients with H. pylori infection and duodenal ulcer disease have several abnormalities of acid secretion compared with healthy volunteers. They have a substantial increase in basal acid output, and the response to gastrin-releasing peptide is exaggerated. They also have an increased maximal acid response to exogenous gastrin and an increased ratio of basal acid output to maximal gastrin-stimulated output. These effects result in the production of increased amounts of gastric acid in response to physiological stimuli and increased delivery of acid to the duodenum. All of these abnormalities resolve completely after the eradication of H. pylori , except for the increased maximal acid output in response to gastrin, which remains unchanged.
In contrast with patients with antrum-predominant gastritis, patients with corpus-predominant gastritis have decreased acid secretion, which tends to return to normal when eradication therapy is administered. Gutierrez et al. assessed the effect of H. pylori eradication on basal acid output and maximal acid output in H. pylori-infected dyspeptic patients before and after successful eradication therapy.20 Gastric acid secretion was less than normal in infected patients with corpus gastritis and increased into the normal range after successful treatment of H. pylori infection, suggesting that gastric function can be restored to normal after the eradication of H. pylori infection in patients with chronic corpus gastritis. Similar data have been reported recently by Haruma et al., who studied the intra-gastric pH before and 1 year after the eradication of H. pylori in patients with moderate to severe atrophy of the gastric body.21 The median 24-h pH decreased after eradication therapy, and the percentage of time pH > 4 decreased from 65% to 28%, suggesting the restoration of acid secretion. Feldman et al. studied gastric acid secretion, oesophageal acid exposure and gastric emptying before and after H. pylori eradication in asymptomatic individuals.22 Basal gastric acidity increased in subjects whose infection was eradicated, but not in those with persistent infection. Basal and meal-stimulated gastric acid secretion did not change after H. pylori eradication. Pathological acid reflux developed in three of nine patients with the eradication of H. pylori and in one of five patients who did not have successful eradication.
In summary, changes in acid secretion depend on the pattern of gastritis. Most studies suggest that patients with antrum-predominant gastritis have increased gastric acid secretion that returns to normal after eradication. Patients with corpus-predominant gastritis may have reduced gastric acid secretion that returns to normal after eradication.
Effect of H. pylori eradication on 24-h oesophageal pH and motility
Tefera et al. studied the effect of H. pylori eradication on oesophageal acid exposure; pH studies were performed before and 12 weeks after eradication therapy in 25 subjects.23 There was no change in the total time pH < 4 in patients with eradication, and there was no significant change in reflux symptom scores. Manifold et al. studied 25 patients with H. pylori gastritis using ambulatory 24-h oesophageal and gastric pH-metry and gastric bilirubin monitoring before and 12 weeks after H. pylori eradication.24 No differences were found in oesophageal acid reflux, gastric alkaline exposure or gastric bilirubin exposure before and after the eradication of H. pylori. Wu et al. studied patients with erosive oesophagitis and controls without reflux disease. There was no difference in the severity of oesophagitis or in the oesophageal acid exposure between H. pylori-positive and H. pylori-negative groups.25 Subjects infected with H. pylori had significantly lower basal pressure in the lower oesophageal sphincter and decreased amplitude of distal peristalsis compared with individuals who were not infected with H. pylori. Ineffective oesophageal motility and failed oesophageal peristalsis were significantly more prevalent in H. pylori-positive patients. These data suggest that eradication therapy does not increase the distal acid exposure in the oesophagus and that chronic H. pylori infection may be associated with oesophageal dysmotility. In another study, Wu et al. investigated oesophageal acid exposure in 14 patients with gastro-oesophageal reflux disease and H. pylori infection, who were randomized to receive H. pylori eradication treatment, and in 11 patients randomized to receive omeprazole alone.26 There was no difference in the percentage of time the oesophageal pH < 4 before and 26 weeks after treatment between the groups. However, the percentage of time the oesophageal pH < 2 (P = 0.01) and pH < 3 (P = 0.02) was significantly increased in patients receiving H. pylori eradication treatment. A small number of patients developed worsening oesophagitis (n = 3). Overall, these studies show little effect on distal oesophageal acid exposure after H. pylori eradication.
H. pylori eradication and the development of reflux disease
In a study of patients with duodenal ulcer disease who received eradication therapy, Labenz et al. estimated that the incidence of oesophagitis was 25.8% at 3 years in patients with successful eradication of H. pylori, compared with 12.9% at 3 years in patients with ongoing infection.27 Factors that predicted the development of reflux oesophagitis were the severity of corpus gastritis, weight gain and male gender. This study has been criticized because the two groups of patients came from different parts of Germany, and it focuses on endoscopic oesophagitis rather than symptoms. In contrast, Vakil et al. studied patients with endoscopically documented duodenal ulcer disease and H. pylori infection (demonstrated by rapid urease test and histology).28 There were 111 patients who had heartburn before eradication. Heartburn resolved in 68% of patients with successful eradication. Of the 101 patients who did not have heartburn before eradication therapy, 17 developed heartburn (12 of 78 patients infected with H. pylori and five of 23 patients with successful eradication) (no significant difference). Erosive oesophagitis developed in one of the 211 patients undergoing endoscopy at 6 months. These data suggest that heartburn may improve after successful eradication of H. pylori in patients in whom it is present before eradication therapy is administered. New symptoms of heartburn develop in a small number of patients, but erosive oesophagitis is uncommon. McColl et al. found that patients with dyspepsia and ulcer disease frequently had heartburn (31%).29 The benefit of eradication therapy in resolving the dyspeptic symptoms was similar in patients with reflux-type dyspepsia (20%) and ulcer-type dyspepsia (23%). Fallone et al. studied 87 patients with duodenal ulcer disease and H. pylori infection: 63 patients had successful eradication of H. pylori and 24 did not.30 Symptoms of gastro-oesophageal reflux disease were found in 18 of 63 (29%) patients with successful eradication and two of 24 (8%) patients with persistent H. pylori infection, but the presence of pre-treatment heartburn was not described and patients were not stratified on the basis of pre-existing symptoms. Oesophagitis was found in 13 of 63 (21%) patients with successful eradication vs. one of 24 (4%) patients with failed eradication; however, the results failed to reach statistical significance, and life-table analysis did not reveal a significant increase in the risk of development of symptoms or oesophagitis. In an interesting study from Japan, 286 Japanese subjects with gastritis, gastric ulcer or duodenal ulcer received eradication treatment or placebo.31 Endoscopic oesophagitis was seen in 36 patients in the eradication group and in only one patient in the control group. Life-table analysis showed that the estimated 3-year prevalence of gastro-oesophageal reflux disease was 18% in patients treated for H. pylori and 0.3% in patients receiving no eradication treatment. Most reflux disease cases were mild (Los Angeles grade A in 24 patients, grade B in 11 and grade C in one). The incidence of reflux oesophagitis was higher in patients with corpus gastritis than in patients without corpus gastritis. The presence of hiatus hernia was significantly associated with the development of oesophagitis. These data suggest that the cure of corpus gastritis in patients with other predisposing factors for gastro-oesophageal reflux disease may account for the development of symptoms or erosive oesophagitis.
H. pylori and the treatment of reflux disease
Intra-gastric pH values are significantly higher during treatment with proton pump inhibitors in patients who are infected with H. pylori than in those who are not.32H. pylori contributes by neutralizing acid in the stomach through the activity of urease. Some investigators have therefore suggested that the healing of erosive oesophagitis may be better with proton pump inhibitor therapy in patients who are infected with H. pylori, but this has been shown in only one study. Holtmann et al. showed that patients with H. pylori infection who were treated with a proton pump inhibitor (pantoprazole) had better symptom relief and better healing of oesophagitis at 4 and 8 weeks than patients who were not infected with H. pylori.33 The effects were much less pronounced at 8 weeks than at 4 weeks. Carlsson et al. studied 1350 patients with gastro-oesophageal reflux disease treated with omeprazole.34 They found that the healing rates and symptomatic relief were similar in patients infected with H. pylori and in those who were not. Preliminary data from Vakil et al. suggest that H. pylori infection does not affect the healing rates in erosive oesophagitis treated with esomeprazole or lansoprazole.35, 36 Schenk et al. assessed the dose of omeprazole required for maintenance of healing in erosive oesophagitis in patients with and without H. pylori infection.37 The maintenance dose was similar in patients with and without H. pylori infection.
Effect of H. pylori eradication on relapse in patients with established gastro- oesophageal reflux disease
Moayyeddiet al. recently demonstrated, in a prospective controlled trial, that H. pylori eradication does not influence the relapse rates in gastro-oesophageal reflux disease patients.38H. pylori-infected patients with a chronic (> 1 year) history of heartburn and normal endoscopy or grade A oesophagitis were recruited. Patients were randomized to receive 20 mg omeprazole, 250 mg clarithromycin and 500 mg tinidazole twice a day for 1 week, or 20 mg omeprazole twice a day with identical placebos. A second concurrently recruited control group of H. pylori-negative patients were given open-label 20 mg omeprazole twice a day for 1 week. Relapse of gastro-oesophageal reflux disease (defined as moderate or severe symptoms) occurred in 83% of each group during the 12-month study period. Life tables revealed no statistical difference between the groups in the time to first relapse. Schwizer et al. studied 70 patients with gastro-oesophageal reflux disease. All patients received lansoprazole, 30 mg twice daily, for 10 days, followed by 30 mg once daily for 8 weeks.39 Patients infected with H. pylori received clarithromycin and amoxicillin or placebo for 10 days. Controls were patients not infected with H. pylori. Patients were followed up at 2-week intervals for 6 months to assess the development of gastro-oesophageal reflux disease symptoms. At the end of the study, endoscopy was repeated, and oesophageal and gastric pH studies were performed. H. pylori-infected patients relapsed earlier than patients in whom H. pylori had been eradicated and controls who had never had H. pylori infection. There was no difference in 24-h pH values at the start and end of the study, suggesting that H. pylori eradication did not affect distal oesophageal acid exposure. These data suggest that the eradication of H. pylori has no adverse effect on the relapse rate in gastro-oesophageal reflux disease, and may be beneficial.
H. pylori and gastric atrophy in reflux disease
The relationship between H. pylori infection and reflux disease is further complicated by studies which suggest an increased risk of gastric atrophy development in patients who are H. pylori-positive and are being treated with long-term proton pump inhibitor therapy. In a small subset of patients with H. pylori infection, chronic gastritis can lead to the development of gastric atrophy and intestinal metaplasia, potential precursor lesions for gastric adenocarcinoma. In addition, in H. pylori-positive patients receiving proton pump inhibitor therapy, the severity of corpus gastritis may increase; however, whether long-term proton pump inhibitor therapy accelerates the development of gastritis with atrophy (surrogate marker for gastric cancer) is very unclear. In an initial study by Kuipers et al., none of the H. pylori-infected gastro-oesophageal reflux disease patients treated with anti-reflux surgery developed gastric atrophy, compared to 31% of patients treated with proton pump inhibitor therapy (omeprazole) for an average of 5 years.40 However, in this study, the patients in the proton pump inhibitor group and the surgery group were from different countries, and the patients in the surgery group were significantly younger than those in the omeprazole group. Similar results were reported by Eissele et al.41 In addition, in an open-label, long-term trial of patients with reflux oesophagitis treated for a mean of 6 years with omeprazole, there was an increase in both the severity of corpus gastritis and glandular atrophy in H. pylori-infected patients.42 Amongst the H. pylori-infected patients, gastric atrophy was detected in 12% at baseline and 39% on follow-up. Therefore, one of the reasons why H. pylori should be eradicated is concern about the development of gastric atrophy in patients on long-term proton pump inhibitor therapy in the presence of H. pylori infection. However, in a trial of gastro-oesophageal reflux disease patients randomized to omeprazole vs. anti-reflux surgery, Lundell et al. did not find any evidence of accelerated development of atrophic gastritis in patients on long-term proton pump inhibitor therapy;43H. pylori-infected patients in both treatment groups (34% in the anti-reflux surgery group and 26% in the omeprazole group) developed gastric atrophy.
It is still unclear whether gastric atrophy occurs in reflux patients due to the synergy between H. pylori infection and proton pump inhibitor therapy; in addition, the clinical significance of gastric atrophy in these patients on long-term therapy is unknown. Although atrophic gastritis is a surrogate marker for the development of gastric cancer, and the diagnosis of atrophy is difficult for pathologists, currently it is still controversial whether or not patients on long-term proton pump inhibitor therapy should be tested and treated for H. pylori infection; however, a test and treat strategy in this situation has been proposed by the Maastricht Consensus Conference as well as by other investigators.44, 45
H. pylori and reflux complications: barrett's oesophagus and oesophageal adenocarcinoma
Barrett's oesophagus is a condition in which the squamous mucosa in the distal oesophagus is replaced by metaplastic columnar tissue. It is the pre-malignant lesion for adenocarcinoma of the oesophagus and oesophago-gastric junction.46 The possible protective effect of H. pylori infection (especially cagA+ strains) is probably related to the fact that complications of gastro-oesophageal reflux disease, such as Barrett's oesophagus and oesophageal adenocarcinoma, are significantly more frequent in Caucasians than in African-American or Asian populations, which is inverse to the prevalence of H. pylori infection.47 This is entirely based on epidemiological associations, without evidence of causality. Weston et al. recently reported H. pylori infection in 73 of 208 (35.1%) patients without dysplasia, compared with 17 of 47 (36.2%) with low-grade/indefinite dysplasia, two of 14 (14.3%) with high-grade dysplasia and three of 20 (15%) with adenocarcinoma.48
Does the cagA status matter?
Virulent strains of H. pylori, including those with a cytotoxin-associated gene (cag) pathogenicity island, i.e. cagA+, have been reported to cause significant gastric inflammation and cytokine production, and patients harbouring these cagA+ strains have a greater risk of developing gastric atrophy and intestinal metaplasia, duodenal ulcers and gastric adenocarcinomas.49–52 Patients harbouring cagA+ strains of H. pylori are more likely to develop peptic ulcer disease and distal gastric cancer than are individuals infected with cagA– strains. Also, the presence of virulence factors (VacA S1 genotype and/or cagA) has been reported to be significantly lower in gastro-oesophageal reflux disease patients compared to those with peptic ulcer disease in a group of H. pylori-infected patients tested in The Netherlands.53 A number of other studies also suggest that infection with H. pylori, including cagA+ strains, may protect against the development of erosive oesophagitis, Barrett's oesophagus and its associated dysplasia and adenocarcinoma. A recent study by Warburton-Timms et al. reported that the risk of severe oesophagitis was significantly decreased for patients infected with cagA+ strains of H. pylori (odds ratio, 0.57; 95% confidence interval, 0.41–0.80); however, there was no difference in the prevalence of H. pylori between those with and without reflux oesophagitis.54 In a study of 210 patients with gastro-oesophageal reflux disease and Barrett's oesophagus, Vicari et al. found that the prevalence of H. pylori infection in patients with Barrett's oesophagus was 34%, compared with 46% in an age-matched control group (not significantly different).55 However, when compared with controls (42%), the prevalence of cagA+H. pylori was significantly decreased in patients with Barrett's oesophagus (13.3%) and Barrett's oesophagus with dysplasia/cancer (0%). Similarly, Chow et al. reported that the presence of the cagA+ strain was associated with an increased risk for distal gastric adenocarcinoma, but a reduced risk for oesophageal and gastric cardia adenocarcinoma.56 A recent meta-analysis presented at Digestive Disease Week 2002 reported a negative association between the prevalence of both H. pylori and cagA+H. pylori and reflux disease, Barrett's oesophagus and oesophageal adenocarcinoma.57
These findings indicate that cagA+H. pylori may potentially protect against complications of gastro-oesophageal reflux disease, such as Barrett's oesophagus and associated dysplasia/adenocarcinoma.
The exact association between H. pylori and reflux disease continues to be an enigma. The incidence of oesophageal diseases, including oesophageal adenocarcinoma, has increased in recent years, whereas the incidence of H. pylori-related ulcer disease and distal gastric adenocarcinoma has declined. This suggests an inverse relationship between H. pylori infection and oesophageal disease. Adenocarcinoma of the distal oesophagus and the gastro-oesophageal junction has been linked to gastro-oesophageal reflux disease, whereas distal gastric cancer has been related to H. pylori. Current data demonstrate that H. pylori eradication does not, in itself, cause gastro-oesophageal reflux disease. There is a negative association between the prevalence of H. pylori infection, especially cagA+ strains, and Barrett's oesophagus and cancer; however, causality has not been established. The distribution and severity of gastritis and gastric atrophy may be important factors in a subset of patients with a predisposition to gastro-oesophageal reflux disease, rather than just the presence or absence of H. pylori. At this time, it seems that H. pylori probably plays a minor role in susceptible individuals in the development of oesophageal diseases.58 However, this is an evolving area with ongoing research and the future may very well alter our current thinking.