An epidemiologic perspective on the stem cell hypothesis in human carcinogenesis.

BACKGROUND

Tomasetti and Vogelstein have hypothesized that the patterns of cancer incidence in various cells and tissues are highly correlated with the estimated lifetime number of stem cell divisions. The authors reviewed the risks in tissues of 17 types of cancer from the United States and 69 additional countries. Positive correlations were observed consistently between the tissue - specific cancer incidence and the estimated lifetime number of stem cell divisions. The authors concluded that approximately two-thirds of global cancer incidence may be attributed to random DNA replication errors.

METHODS

An epidemiologic perspective is presented that may serve as a counterpoint in interpreting organ-specific cancer risks. The unifying nature of the Tomasetti/Vogelstein hypothesis must be viewed in the context of diverse and contrasting global trends and patterns of types and "causes" of cancers that are closely linked with economic development and cultural lifestyle practices. The presentation is organized by reviewing the global burden of cancer; concepts of causal inferences and counterfactual assumptions; multifactorial causes of hepatocellular carcinoma and a hierarchy of causes that varies internationally; tobacco carcinogenesis and the multiplex associations with 19 cancer sites and tissues; profile in contrasts in transit through the small and large intestine.

OBSERVATIONS AND CONCLUSIONS

It is readily recognized that DNA replication errors and number of stem cell divisions may vary in individuals and populations due to external environmental genotoxic chemicals and biologic agents, and internal hormonal and metabolic factors. There is a striking contrast in the risk of adenocarcinoma in the small intestine with that in the large intestine. Tomasetti and Vogelstein indicated that the cumulative number of divisions of stem cells over a lifetime in normal epithelial mucosal cells from colorectal cancer patients was 4 time greater than in the epithelial tissue from patients with adenocarcinoma of the small intestine. Their conclusion would suggest a "seed" and "soil" interaction rather than exclusively the independence of either component. Namely, that the contrasting physiological, biochemical, microbial and immunological features in the lumen and on the mucosal surface of the large intestine, in contrast to that in the small intestine, would foster molecular genetic and epigenetic events that are advantageous to neoplasia in the large intestine.