OBJECTIVE: This study aimed to identify sources of underestimation of the association between Helicobacter pylori infection and non-cardia gastric cancer, in a high-risk European population. METHODS: Non-cardia gastric cancer patients (n = 420) recruited in two major hospitals in North of Portugal and population controls (n = 1,389) were evaluated. Whole-cell IgG antibodies against H. pylori were quantified by ELISA and Western Blot testing was conducted in a subsample (272 cases and 186 controls) allowing for the detection of current infection marker and CagA. Sex- and age-adjusted odds ratios (OR) and 95% confidence intervals (95% CI) were computed. RESULTS: When assessing infection by ELISA, the OR for its association with gastric cancer decreases and reverts as IgG titers increased, from 1.96 (95% CI: 1.09-3.54) for borderline positive results (16.0-21.9 RU/ml) to 0.52 (95% CI: 0.36-0.74) for the highest IgG levels (> or = 102.0 RU/ml). The same pattern was observed across strata of age and blood collection timing with stronger associations among younger subjects and those providing blood samples earlier after diagnosis. The presence of CagA (Western Blot) was associated with an increased risk of gastric cancer (OR = 11.32; 95% CI: 5.64-22.73). CONCLUSION: The use of methods with low sensitivity to detect past infection leads to a substantial underestimation of gastric cancer risk in high-prevalence settings.
OBJECTIVE: This study aimed to identify sources of underestimation of the association between Helicobacter pyloriinfection and non-cardia gastric cancer, in a high-risk European population. METHODS:Non-cardia gastric cancerpatients (n = 420) recruited in two major hospitals in North of Portugal and population controls (n = 1,389) were evaluated. Whole-cell IgG antibodies against H. pylori were quantified by ELISA and Western Blot testing was conducted in a subsample (272 cases and 186 controls) allowing for the detection of current infection marker and CagA. Sex- and age-adjusted odds ratios (OR) and 95% confidence intervals (95% CI) were computed. RESULTS: When assessing infection by ELISA, the OR for its association with gastric cancer decreases and reverts as IgG titers increased, from 1.96 (95% CI: 1.09-3.54) for borderline positive results (16.0-21.9 RU/ml) to 0.52 (95% CI: 0.36-0.74) for the highest IgG levels (> or = 102.0 RU/ml). The same pattern was observed across strata of age and blood collection timing with stronger associations among younger subjects and those providing blood samples earlier after diagnosis. The presence of CagA (Western Blot) was associated with an increased risk of gastric cancer (OR = 11.32; 95% CI: 5.64-22.73). CONCLUSION: The use of methods with low sensitivity to detect past infection leads to a substantial underestimation of gastric cancer risk in high-prevalence settings.