Helicobacter pylori and gastric cancer

Infection with Helicobacter pylori has been designated a cause of human cancer by the International Agency for Research on Cancer (part of the World Health Organization).1 The infection, which affects 30-50% of adults in Europe and over 80% in parts of the developing world, is the commonest cause of chronic non-atrophic gastritis, which, if severe and prolonged, can lead to loss of gastric glands (atrophic gastritis) and intestinal metaplasia: both markedly increase the risk of developing gastric cancer.1,2 H. pylori infection appears, by itself, capable of causing gastric mucosal atrophy,3 and there is some evidence that prolonged reduction in acid secretion, whether induced by disease or drugs, might accelerate this change.4,5 These observations pose difficult dilemmas for doctors treating patients with gastric disorders, especially those needing long-term treatment with potent antisecretory drugs.


EPIDEMIOLOGY OF GASTRIC CANCER
The risk of developing gastric cancer increases with age (especially over 50 years), with men twice as likely to be affected as women. The incidence is declining in the UK (to around 10-20 per 100,000 population) and most other developed countries, but remains high in Japan, China, and parts of South America and Eastern Europe. 6 The decline is largely due to a fall in cancers of the distal stomach and corpus, which, histologically, may be 'intestinal' or 'diffuse' in type. Environmental factors, (e.g. poverty, diets low in fresh fruit and vegetables or high in salt and preserved foods), especially early in life, appear to be important risk factors for the later development of gastric cancer at these sites. 6 Cancers of the gastric cardia and gastro-oesophageal junction appear to have a different aetiology and are becoming more common in Europe and North America. 6

THE ASSOCIATION WITH H. PYLORI
Several studies have shown that populations at high risk of gastric cancer also have a high prevalence of H. pylori infection, often from an early age. 2, 7 The association does not hold in all populations studied, nor does it explain the male predominance of gastric cancer. 8 More persuasive 'prospective' data come from three 'nested' case-control studies in which analysis was done on sera banked a number of years previously from patients who subsequently developed gastric cancer. [9][10][11] Each study showed that individuals with antibodies to H. pylori were more likely to develop gastric cancer: the relative risk after an interval of 15 or more years was 8.7 (95% confidence intervals [CI] 2.7-45) compared with seronegative subjects. 12 A recent meta-analysis of 19 cohort and case-control studies, involving a total of 2491 patients with gastric cancer and 3959 controls, found a summary odds ratio of 1.92 (95% CI 1.32-2.78) for gastric cancer in patients testing positive on serology for H. pylori infection, compared to seronegative subjects. 13 The relative risk was greatest in younger H. pyloriinfected patients (odds ratio 9.29 at age less than 30 years; 95% CI 3.43-34.04). Seropositivity increased the risk for both intestinal and diffuse types of cancer, but not for cancer of the cardia.
The epidemiological data are strengthened by evidence linking H. pylori infection with the precursor lesions of gastric cancer. 5 For example, in one prospective study, progression to atrophic gastritis and intestinal metaplasia developed in 16 of 58 (27%) H. pylori-infected patients during follow-up lasting 10-13 years, but in only 2 of 49 (4%) of non-infected patients (p<0.001). 3 About 60% of H. pylori isolates possess a cytotoxinassociated gene (CagA). 14 Such strains cause more severe gastritis, and more extensive mucosal atrophy and intestinal metaplasia, than CagA-negative strains, 14 and are more strongly associated with intestinal type gastric cancer. 15 The latter finding makes it less likely that infection with H. pylori (which, like gastric cancer, is associated with poverty, overcrowding and poor sanitation, especially in childhood) is simply a surrogate marker for other carcinogenic mechanisms.

POSSIBLE CARCINOGENIC MECHANISMS
Prolonged inflammation of the stomach from any cause can eventually lead to gastric mucosal atrophy, which results in decreased secretion of acid. This allows colonisation of the stomach by bacteria able to convert dietary and salivary nitrate into nitrite, which can then form potentially carcinogenic N-nitroso compounds. 6 Additionally, H. pylori infection markedly decreases the normal secretion of ascorbic acid into the gastric lumen. 16 Ascorbic acid is thought to protect against gastric carcinogenesis, for instance, by inhibiting the production of N-nitroso compounds and scavenging reactive oxygen metabolites, which can damage cell membranes and alter DNA. 16

HELICOBACTER PYLORI AND GASTRIC CANCER
Vol 36 No 8 August 1998 DTB Several other effects of H. pylori infection might increase the likelihood of developing gastric cancer. The infection causes epithelial cell proliferation, which may encourage cell mutations, 17 and appears to promote the release of reactive oxygen metabolites from inflammatory cells. 18 Many people are infected with H. pylori but very few develop gastric cancer. Gastric cancer more commonly develops in patients in whom the gastritis is widespread or mainly affects the corpus rather than the antrum, 5,19 and in those in whom it progresses more rapidly to atrophy, as when there is sustained achlorhydria. 5 Many host and environmental factors appear capable of modifying this progression.

INTERVENTION STUDIES
Proof that infection with H. pylori causes gastric cancer can ultimately only come from intervention studies. Such studies would need to compare outcome in cancer-free H. pyloriinfected subjects randomised to eradication therapy or no therapy, and would have to observe tens of thousands of subjects for at least 10 or 15 years. Studies of patients who already have gastric atrophy or intestinal metaplasia would permit smaller and shorter trials, and some are under way. However, while eradication of H. pylori reverses chronic non-atrophic gastritis, it is unclear whether it can reverse these more advanced lesions, 5,14 so the findings in such patients might not predict the treatment effect in the general population of H. pylori-infected patients.
Limited data have already suggested possible benefit from treatment. In a non-randomised Japanese study of 132 H. pyloriinfected patients who underwent endoscopic resection for early gastric cancer, those in whom the infection was eradicated suffered no new gastric cancers over 4 years of follow-up, while further early gastric cancers developed in 6 of 67 (9%) patients who received no eradication therapy (p=0.01). 20

DILEMMAS FOR CLINICAL PRACTICE
Until definitive data have shown that eradication therapy reduces the risk of developing gastric cancer, a policy of routine screening and treatment of whole populations would be premature. In the meantime, practitioners have to decide how to treat individual patients found to be infected.
There is general agreement that patients with peptic ulcer who are infected with H. pylori should receive anti-Helicobacter treatment. In the West, however, most patients who seek advice for dyspepsia have gastro-oesophageal reflux or nonspecific dyspepsia, rather than peptic ulcer or cancer. Although many of these will have H. pylori infection, so far there is little evidence that treating their infection will help their symptoms. 21 Many patients who have dyspepsia are given acid-suppressing drugs, and long-term use of them, particularly proton pump inhibitors, could potentially affect the complex balance between H. pylori infection, acid secretion and cancer risk. 4,5

H. pylori infection, acid secretion and cancer risk
H. pylori infection and inflammation affecting predominantly the antrum (the most common pattern) is associated with hypergastrinaemia, increased acid output and duodenal ulcer, but low risk of gastric cancer. 19 In a minority of patients, however, infection causes gastritis chiefly in the body of the stomach: acid output is reduced (and sometimes abolished) but recovers to a variable degree after eradication of the infection. 22 This pattern of inflammation is often accompanied by varying degrees of mucosal atrophy, and is characteristic of patients who develop gastric cancer. 19,22 Why some patients develop a predominantly corpus gastritis is not known. However, it appears more likely in people with a naturally low acid output, 5,19,22 and there is evidence that genetic factors (e.g. HLA antigen status), diet, age at infection, and properties of the infecting strain (e.g. cytotoxin production, stimulation of acid-inhibitory factors, or of antibodies that cross-react with the proton pump) are also important. 5,19,22 The effect of antisecretory drugs The use of antisecretory drugs in H. pylori-infected patients reverses the predominantly antral pattern of gastritis and increases the severity of corpus gastritis. 23-25 These changes, which are seen both with proton pump inhibitors and H 2 -receptor antagonists, are reversible after short-term treatment, but persist in patients receiving long-term maintenance therapy with omeprazole for severe gastro-oesophageal reflux. 4 In one 5-year study of long-term management of reflux, the annual increase in atrophic gastritis (6.1%) was higher in 59 H. pylori-infected patients on omeprazole than in 46 uninfected patients on omeprazole (0.8%), or in 31 infected patients who had fundoplication (0%) (p<0.001). 4 This non-randomised study, though very important, is problematic: the omeprazole and fundoplication groups were from different populations and those treated with omeprazole were on average 9 years older (gastric atrophy results from long-lasting gastritis and its development may increase exponentially with age). The findings have, moreover, recently been challenged by data from a shorter randomised trial, so far published only as an abstract. 26 After an average follow-up of 3 years, this study found no increase in gastric atrophy or intestinal metaplasia among 155 H. pyloriinfected patients on long-term omeprazole compared with 155 infected patients treated by anti-reflux surgery.

IMPLICATIONS FOR TREATMENT
H 2 -receptor antagonists have been widely prescribed for over 20 years and proton pump inhibitors for more than 5 years. During this time, the incidence of distal and body cancer has fallen. Nevertheless, the evidence that acid suppression in an infected person changes the topography and nature of H. pyloriassociated gastritis, so that it comes to resemble that most commonly associated with these cancers, has led many experts to conclude that it might be prudent to eradicate H. pylori in patients with a proven need for long-term acid-suppressant therapy. 5,27 Although this seems sensible, it would inevitably expose more patients to unwanted effects, such as antibiotic-induced diarrhoea, and might promote the emergence of resistant strains of H. pylori. There is also some evidence that patients may be less sensitive to the acid-suppressing effects of proton pump inhibitors after eradication of H. pylori infection, and that eradication might make reflux more troublesome in some patients. 28 It has also been suggested that infection with H. pylori might confer some protection against gastro-oesophageal reflux and its complications, including Barrett's oesophagus and associated carcinoma. 29,30 Adenocarcinoma of the cardia and oesophagus have become much more prevalent in areas where the prevalence of H. pylori infection has declined, and are associated with the absence of H. pylori infection with CagA-positive strains. 30 Much more needs to be known about the relationship between H. pylori infection and gastro-oesophageal reflux.
These considerations make it all the more important to consider carefully whether long-term treatment with acid suppression is the best option for younger patients with already long-standing reflux, who are clearly likely to need drug therapy for years. Lifestyle measures should be re-emphasised, and surgery is sometimes a suitable alternative for such patients.

CONCLUSION
Infection with Helicobacter pylori is the commonest cause of chronic gastritis, which, if sufficiently extensive and prolonged, can lead to mucosal atrophy, intestinal metaplasia and gastric cancer. Epidemiological evidence is accumulating for a causal association between H. pylori infection and cancer of the gastric corpus and antrum. The incidence of these cancers is declining in areas where H. pylori infection is becoming less prevalent. By contrast, cancers of the cardia and lower oesophagus are becoming more common.
In patients infected with H. pylori, long-term therapy with potent antisecretory drugs alters the distribution of the infection within the stomach. The pattern of gastritis then resembles that more commonly associated with the development of mucosal atrophy, and found in the small minority of all H. pylori-infected patients who eventually develop cancer of the gastric corpus and antrum. For these reasons, many of our consultants consider it prudent to eradicate H. pylori infection in patients with severe gastro-oesophageal reflux who have proven to need long-term acid suppression. Whether this could impair the control of reflux symptoms, or promote the development of complications of reflux, are important and, as yet, unanswered questions.