Clinical usefulness of virulence factors of Helicobacter pylori as predictors of the outcomes of infection. What is the evidence?

H elicobacter pylori is one of the commonest bacterial pathogens in humans, infecting between 80%–90% of the population in developing countries and between 30%–50% in developed countries (1). Although it was isolated in 1982, it is considered to be part of the natural human flora since time immemorial (2) and probably an historical equilibrium between the bacteria and the host has existed. H. pylori colonizes and proliferates in the mucus layer over the epithelium (3). The ability to survive and grow in gastric acid is apparently linked to its ability to maintain a tolerable pH by activation of internal urease production. In the majority of infected humans, there are no clinical consequences from H. pylori infection. Only 15%–20% of them will develop severe gastroduodenal pathology (3) and less than 1% will eventually develop a gastric tumour in their life. It is known that in many geographical areas no correlation exists between prevalence of infection and the risk of gastroduodenal diseases. Moreover, why H. pylori infection produces different outcomes and why a minority of infected develop a clinical disease are remaining questions. H. pylori strains appear to be highly diverse (4) and some strains seem to be more virulent than others (5). Therefore, it has been proposed (6) that the type of bacterial strains, in combination with host factors, may condition both the intensity and the pathway of the process. However, the strength of the scientific evidence provided in the literature is variable. The aim of this comprehensive review is to evaluate the most recent published epidemiological evidence on the role of virulence factors of H. pylori as markers of severity and specificity of the consequences of H. pylori infection. This means its clinical usefulness as predictors of the outcomes of infection. The markers of bacterial virulence factors so far described, that may influence H. pylori putative virulence and/or clinical outcomes, that were analysed are the following (Table I): the cytotoxin CagA, vacuolating cytotoxin (VacA), IceA, BabA2 and Neutrophil activating protein (NAP). We have identified all epidemiological studies published up to December 2002 available in MEDLINE (National Library of Medicine) that investigated the role of these virulence factors of H. pylori infection in the risk of gastric cancer or peptic ulcer. All casecontrol studies or prospective nested case-control studies were considered for the review. Studies based only on serological results of H. pylori antibodies not evaluating other virulence factors were excluded. Clinical studies based on a series of cases without controls, as well as some experimental studies using cell lines, were considered only to support or explain findings from etiological studies. We found 20 articles based on 14 case-control studies and 4 nested studies, which are summarized in Tables II and III, respectively. For each study, we evaluated the design and size, the study base, the features of controls, the covariates included and the results for each virulence factor analysed. Comparison of results with H. pylori antibodies alone was performed whenever this result was available.