Factors related to the occurrence of gastric adenocarcinoma subtypes

Abstract: Gastric cancer is the second most common cancer worldwide, but the etiology remains poorly understood. A spontaneous global decline has reduced the number of victims in most Western countries while Japan and China have screening programs in place due to the magnitude, and the serious prognosis, of this malignant disease. Secular and geographic variations indicate environmental risk factors, but also reveals discrepancies between subtypes of gastric cancer both according to tumor histology (intestinal/diffuse) and site (cardia/non-cardia) emphasizing the need to perform subtype-specific etiologic research. The aims of this thesis were to elucidate patterns of occurrence, and, to identify and to study risk factors of gastric adenocarcinoma subtypes. The population-based case-control study, conducted from February 1989 through January 1995, also set the structure for two descriptive studies within the same source population (in total 1.3 million) in two geographic areas in Sweden with differing gastric cancer incidence. We identified all new cases of gastric adenocarcinoma in the study area, with a completeness of case ascertainment exceeding that of the Swedish Cancer Register (papers I and II). All Swedes aged 40-79, living in the two areas were eligible cases if diagnosed with gastric adenocarcinoma and subject to the random selection of frequency (2: 1) matched controls. Cases (567) and controls (1165) were interviewed about various exposures, and sera was collected from (542) individuals (papers III-V). We found that the decline in gastric cancer can be entirely attributed to non-cardia tumors, supporting the hypothesis that cardia and non-cardia cancers are distinct entities with diverging etiologies. With a strict classification of site, we could not confirm the feared rise in cardia cancer. The rate of the decline was identical for the intestinal and diffuse types, indicating environmental determinants for both histologic types. Despite an excellent overall completeness, the accuracy in national registration of cardia tumors was surprisingly low and reported increase in cardia incidence during the past decades could theoretically be explained by misclassification. A low intake of antioxidants was associated with an increased risk of all gastric cancer subtypes and a high parallel intake of ascorbic acid, ß-carotene and [alpha]-tocopherol was significantly associated with reduced risk, suggesting synergistic mechanisms. Our findings also imply that other bioactive compounds in fruit and vegetables could be important in reducing the risk of gastric cancer. Exposure to metal work and phenoxy acids was associated with increased risk of gastric adenocarcinoma, but causality is unclear. With a more correct classification of exposure, the risk associated with H. pylori appears stronger than earlier, confined to non-cardia tumors, and substantial for both intestinal and diffuse type tumors. Interactions between antioxidants, H. pylori infection and cigarette smoking may considerably modify the risk of all types of gastric adenocarcinoma, indicating that H. pylori induces oxidative stress that could be ameliorated by dietary antioxidants.