Epidemiological studies of Helicobacter pylori and its relation to cancer and precancerous lesions in the upper gastrointestinal tract

Abstract: Helicobacter pylori infection is strongly associated with gastric cancer risk. The carcinogenic process is supposed to occur step by step, from a normal gastric epithelium via gastritis, atrophy, intestinal metaplasia to dysplasia. The overall aim of this thesis was to map out the pathway from Helicobacter pylori infection to cancer and identify risk factors for the steps. We also wanted to investigate whether H. pylori infection is associated with the risk of oesophageal cancer. We investigated in a hospital-based case-control study if H. pylori strains lacking the virulence factor CagA were associated with risk of gastric cancer. Because of an observed decrease in gastric cancer incidence, we also studied if the prevalence of atrophic gastritis, a precursor of gastric cancer, was decreasing. The age- and year- specific prevalence of atrophic gastritis was estimated in a cross-sectional study of the population of Västerbotten and Norrbotten 1990, 1994 or 1999. We further studied factors associated with the risk of atrophic gastritis. We performed another cross-sectional analysis in the control groups of two earlier population-based case-control studies to identify factors associated with being H. pylori sero-positive in adult life. The participants constituted an age- and sex-stratified random sample of the Swedish population. Finally, we used a population-based case-control study to investigate if H. pylori infection is associated with the risk of adenocarcinoma of the oesophagus or gastric cardia, or with squamous-cell carcinoma of the oesophagus, and if so, whether gastric atrophy might be in the causal pathway. The occurrence of IgG antibodies towards H. pylori was analysed by ELISA and immunoblot, the latter method also detected antibodies towards CagA. The presence of atrophic gastritis was determined through analysis of pepsinogen I in serum. Odds ratios were calculated by logistic regression. Comprehensive interview data made multivariate adjustments of confounding factors possible. We found that persons with antibodies to CagA ran the greatest risk of developing gastric cancer, but persons with CagA-negative infections still ran a four-fold increased risk in comparison with uninfected persons. CagA-seropositivity was also associated with increased odds of having atrophic gastritis, and both atrophic gastritis and infection with CagA-positive strains of H. pylori were associated with an increased risk of oesophageal squamous-cell carcinoma. H. pylori infection was, on the other hand, associated with a reduced risk of oesophageal adenocarcinoma, but there was no association between atrophic gastritis and risk of oesophageal adenocarcinoma. In the 55-64 year age group the prevalence of atrophic gastritis decreased in parallel with the falling incidence of gastric cancer. This probably indicates that the falling gastric cancer incidence is caused by factors acting before atrophic gastritis has developed. A declining prevalence of H pylori infection is the most plausible explanation. In the 35-44 year age group, however, the prevalence of atrophic gastritis increased, which might indicate that we will soon observe the lowest point of the gastric cancer incidence curve, followed by a new upward trend. A high body mass index was associated with increased odds of atrophic gastritis. Higher education and alcohol consumption were associated with a decreased risk. No clear tendency was seen for vegetable and fruit intake. The risk factors associated with a high H pylori seroprevalence in middle and older ages were mainly acting early in life, indicating that most infections were established during childhood. The risk factor pattern was further remarkably similar to that observed for stomach cancer.