Antibiotic treatment may prevent deaths in the decades ahead

Infection with Helicobacter pylori is the main cause of human gastritis,1 the major cause of peptic ulcer,2 and an important risk factor for gastric cancer.3 In 1994 the US National Institutes of Health recommended antibiotic treatment of H pylori for patients with gastroduodenal ulcers.2 This approach has been swiftly adopted4 and will surely accelerate as newer and more effective regimens are introduced. A more difficult dilemma, however, concerns the management of the growing number of dyspeptic patients who are diagnosed with confirmed H pylori infection but no ulcer.

No convincing evidence exists to suggest that eradicating H pylori improves the symptoms of non-ulcer dyspepsia,5 but as testing for the infection becomes part of the routine investigation of dyspeptic patients the question arises whether treatment should be given as an insurance against future, more serious, illness. Indeed, ethical, and perhaps legal, difficulties may confront the doctor who diagnoses a patient as H pylori positive and then fails to do anything about it. Even if serious sequelae are uncommon, is it appropriate to tell these patients that they have an infection which conveys no known benefit to them, that increases their risk of peptic ulcer and cancer, that may be transmitted to their children, but does not require treatment?

Commonly used regimens comprise a one week course of an acid pump inhibitor combined with two antibiotics (amoxycillin, clarithromycin, or a nitroimidazole). The clarithromycin-nitroimidazole regimen achieves eradication in 85-95% of cases6 at a cost of around £20. Nevertheless, doctors are reluctant to treat H pylori in the absence of documented peptic ulcer for several reasons.

The major concern relates to the emergence of resistant organisms. These dangers have probably been exaggerated.7 Nitroimidazoles (metronidazole) and macrolides (erythromycin) have been widely used for many years. In 1994 over 50 million antibiotic prescriptions were issued in England, amounting to more than one course for every person.8 A second concern is whether infection with H pylori might conceivably benefit the people infected with it.9 A recent study suggested that the eradication of H pylori from patients with peptic ulcer might precipitate gastro-oesophageal reflux.10 More work on this is needed, but an alternative explanation is that reflux symptoms are unmasked when acid suppressive treatment is stopped.

The data required to determine the best management of patients positive for H pylori but without ulcers will eventually come from controlled clinical trials. Several trials are in progress or under discussion, though few are large or long enough to use cancer as an end point. Only clinical trials will provide the quantitative evidence of benefit to set against the cost of treatment, the problem of adverse drug reactions, and the microbiological consequences of widespread administration of antibiotics. Until such trials have been reported, one important factor to consider in evaluating the potential benefits of H pylori eradication is the risk of death from this infection.

In England and Wales over 7500 deaths occur each year from gastric cancer and over 4000 occur from peptic ulcer.11 Not all can be attributed to infection with H pylori. Nevertheless, three out of four gastric cancers may be attributable to the infection.12 The contribution of H pylori infection to deaths from peptic ulcer is harder to assess. Although most gastroduodenal ulcers are caused by H pylori infection, the other major cause is non-steroidal anti-inflammatory drugs. Results from a recent large, multicentre study in Britain suggest that about 40% of the 10 000 peptic ulcer bleeds that occur in people over 60 are related to drug treatment.13 In a high proportion, however, helicobacter infection probably also plays a part.14 A conservative estimate is that 65% of these bleeds are attributable to H pylori.

These figures suggest that over 8000 deaths each year are caused by the infection. If we assume that half the population at risk is infected this translates into a cumulative risk of dying from the infection by 85 years of about 1 in 35 for men and 1 in 60 for women, after allowing for competing causes of mortality (1 in 51 for men and 1 in 96 for women from gastric cancer and 1 in 154 for men and 1 in 173 for women from peptic ulcer).

Few common infections in developed countries have a mortality this high, and if the deaths associated with H pylori gastroduodenitis occurred immediately rather than after many decades this situation would be medically unacceptable. Of course, whether treatment of the H pylori infection would prevent these deaths is unknown–which is why trials are essential. The current problem, however, merits debate: H pylori may be a dangerous pathogen and the infection not trivial.

Some may argue that the serious consequences of infection, especially gastric cancer, are declining in incidence and therefore becoming less important. This is true, but only partly. Age standardised rates of gastric cancer are declining rapidly in many populations,15 but this effect is counterbalanced, especially in the developing world, by population growth and an increasing proportion of elderly people. The absolute number of patients with gastric cancer is likely to increase globally despite decreasing incidence rates (D Forman, unpublished observations). Furthermore, much of the decline in mortality from gastric cancer is probably attributable to the fall in the prevalence of H pylori infection in recent decades. The individual risk to the infected patient will remain unchanged.

In summary, it is timely to consider H pylori gastroduodenitis as a disease in its own right, with ulcer and cancer as its important complications. About 85% of infected individuals will not develop complications, but until a reliable means can be found to identify the 15% who become seriously ill all those with the disease must be considered to be at risk of a potentially fatal outcome.