Aspects of gastroesophageal reflux and risk for esophageal cancer : an epidemiological approach

Abstract

Aims: The aims of this study were to confirm, in a prospective study, the link between gastroesophageal reflux. and risk of adenocarcinoma of the esophagus and gastroesophageal junction observed in retrospective studies, to investigate into the effects of anti-reflux surgery, and to shed light on some possible mechanisms behind the alleged carcinogenicity.

Methods: We estimated relative risk for esophageal adenocarcinoma in a cohort of 35,274 male patients hospitalized for gastroesophageal reflux disease (GERD) and another cohort of 6,406 male patients who underwent antireflux surgery identified in the Swedish Inpatient Register, using the general Swedish population as reference. Further, in a nationwide population-based case-control study in Sweden, antibodies against H. pylori and cytotoxin associated gene-A-positive (CagA+) antigens, and pepsinogen I concentration were measured using sera from 97 esophageal adenocarcinoma, 85 esophageal squamous-cell carcinoma patients and 499 randomly selected controls. We also followed 21,265 patients hospitalized for pernicious anemia in Sweden between 1965 and 1999 for an average of 7.1 years and estimated relative risks of esophageal cancer by histology. Finally, polymorphisms in XPD codon 751 (Lys→Gln) were genotyped and compared between cases and controls using material in the above-mentioned case-control study.

Results: More than 6-fold and 14-fold excess risks for esophageal adenocarcinoma were observed among unoperated and operated gastroesophageal reflux disease patients, respectively. The risk among operated patients remained elevated regardless of time after surgery. We found a significantly reduced risk for esophageal adenocarcinoma associated with H. pylori infection. The inverse association remained in the stratum without stomach atrophy. In contrast, subjects with CagA positive serology had an about 2-fold excess risk for esophageal squamous-cell carcinoma. Compared with the general population, a significant excess risk for esophageal squamous-cell carcinoma was observed in pernicious anemia patients, while no significant risk elevation or reduction was observed for esophageal adenocarcinoma. XPD codon 751 Lys/Gln and Gln/Gln genotypes, compared with Lys/Lys genotype, were both associated with a more than doubled risk for esophageal adenocarcinoma. The combined effects of these genotypes and symptomatic gastroesophageal reflux departed from additivity, but only with borderline significance.

Conclusions: Gastroesophageal reflux is strongly associated with tile risk of esophageal adenocarcinoma. However, the high risk of developing esophageal adenocarcinoma remains after antireflux surgery as practiced in Sweden. The effects of earlier intervention against GERD need to be evaluated. Infection with H. pylori is inversely associated with risk of esophageal adenocarcinoma, but the hypothesized pathway via atrophy - reduced acid load in the esophagus appears unlikely. Gastric atrophy following infection with CagA+ strains of H. pylori or type A atrophic gastritis_may increase the risk for esophageal squamous-cell carcinoma. XPD 75 1 Gln allele is a potential genetic marker for susceptibility to esophageal adenocarcinoma, and may interact multiplicatively with gastroesophageal reflux on the carcinogenesis of this tumor.