da Silva, J.

Since most cancers are initiated by driver mutations arising in somatic cells, the risk of cancer should be explained by the probability of driver mutations as a function of the product of the number of stem cells and the stem cell division rate for a tissue. However, such models of driver-mutation initiated carcinogenesis fail to adequately predict cancer risk. It has been suggested that the missing component is greater adaptive cancer suppression in large tissues with high rates of stem cell division. This may manifest as either a greater number of driver mutations required to initiate cancer or a lower stem cell mutation rate. These hypotheses are tested here using nonlinear regression to fit models that incorporate variation in the number of driver mutations and the stem cell mutation rate to data on the tissue-specific cancer risk, number of stem cells, and stem cell division rate. The greatest concordance between predicted and observed cancer risks is obtained by single driver mutations across tissues and stem cell mutation rates that decline with increasing lifetime numbers of stem cell divisions. This provides evidence of adaptive cancer suppression among tissues.