Table 1.

Evaluation of the association between colorectal cancer and H. pylori, S. bovis, JCV, and HPV using the Bradford Hill criteria

H. pyloriS. bovisJCVHPV
StrengthMeta-analysis OR = 1.4ORs for detection of S. bovis in stool range from 1.0 to 10.7ORs for JCV DNA in cases vs controls range from 1.0 to 6.2ORs for HPV DNA in cases vs controls range from 2.7 to 9.1
TemporalityNot evaluatedNot evaluatedNot evaluatedNot evaluated
ConsistencyMost studies OR ≥ 1S. bovis endocarditis studies consistently show elevated risks for colorectal neoplasia. However, stool studies of colorectal cancer cases and controls are not consistent.Inconsistent study resultsEarly studies do not detect HPV DNA in colorectal neoplastic tissue. Nine recent case-control studies consistently report an increased risk.
SpecificityH. pylori is a known cause of gastric cancer.S. bovis is a cause of septicemia and endocarditis.JCV is known to cause progressive multifocal leukoencephalopathy.HPV is a known cause of several anogenital cancers.
Biological plausibilityAssociation between gastrin, gastric H. pylori infection, and colorectal cancer is plausible. Direct infection of the colon is unlikely.S. bovis is known to inhabit the colon, and molecular studies indicate that S. bovis proteins have carcinogenic properties.JCV large T-antigen has carcinogenic properties, but its ability to infect the colon is under debate.HPV infects epithelial cells, and its oncogenic properties are well described, but the ability of HPV to enter the colon and rectum is debatable.
CoherenceGeographic distribution of colorectal cancer differs significantly from gastric cancer.Many colorectal cancers exhibit overexpression of cyclooxygenase-2. S. bovis proteins up-regulate cyclooxygenase-2 in vitro.JCV is a neurotrophic virus and is very common in the population.Colorectal cancer occurs in epithelial cells, but the risk factors for HPV infection are not known for colorectal cancer.
Biological gradientNo evidence for gastrin.Not evaluatedOne study finds viral copy number higher in cases than in controls.Not evaluated
Not evaluated for H. pylori.
AnalogyH. pylori causes gastric cancer.H. pylori induces inflammation in the stomach, resulting in cellular proliferation and increased gastric cancer risk. S. bovis could have a similar mechanism in the colon.SV40, another polyomavirus, is hypothesized to cause certain human cancers, including brain cancer. However, this has not been proven.HPV causes adenocarcinoma in the cervix.
ExperimentNot evaluated, but it is possible to treat H. pylori infection.Not evaluated, but it is possible to treat S. bovis infection.Not evaluated, and currently, it is not possible to prevent JCV through vaccination.Not evaluated, but it is possible through HPV vaccination studies.
  • NOTE: 0, not evaluated; +, weak evidence; ++, moderate evidence; +++, strong evidence.