Helicobacter pylori Biomarkers and Risk of Colorectal Oncogenesis—Response Free

We thank Kountouras and colleagues (1) for their interest in our work describing an association between antibodies to the Helicobacter pylori (H. pylori) virulence factor VacA and risk of colorectal cancer among 188 colorectal cancer cases and 370 matched controls in the prospective study, the Southern Community Cohort Study (2). In their letter, the authors note that our use of a serologic test to determine infection status through assessment of circulating antibody levels prevents us from discriminating between past and present infection. This is true, but as the authors themselves point out, past infection may be even more important for oncogenesis.

The letter writers also present their own data suggesting a direct and local effect of H. pylori on cancers arising within the colon. H. pylori rarely, if ever, colonize nongastric mucosa and an important limitation to the immunohistochemical study cited by the authors is that bacterial viability was not determined. Furthermore, the study described in the letter is based on a small sample size (50 patients with colorectal cancer, 25 patients with colorectal adenoma, and only 10 controls for comparison), which would not allow for adjustment for the significant confounding variable of age, as well as factors related to socioeconomic status and smoking history.

Finally, the authors present several potential mechanisms that may explain the H. pylori protein-specific association with colorectal cancer. We suggest that this association may also be related to the ability of pathogenic H. pylori strains to sculpt the composition of the gastrointestinal microbiome, which in turn may lower the threshold for colon cancer. It is also possible that individuals who are genetically susceptible to colon cancer may, in fact, select for more oncogenic H. pylori strains. However, these putative mechanisms require more detailed investigations, which is an important next step emanating from prospective case–control studies.

See the original Letter to the Editor, p. 365