Stem Cell Divisions, Driver Mutations, and Carcinogenesis in Purebred Dogs

Most cancers are initiated by mutations that inactivate tumour-suppressor genes or activate oncogenes. Fitting a multistage model of carcinogenesis to the increase in cancer mortality with breed-specific size and lifespan in dogs has predicted that four somatic driver mutations are typically required to initiate cancer. This result is reconsidered here because it depends on the relationship between the number of at-risk cells and breed weight. Using a power function for this relationship results in higher quality models that support single driver mutations activating oncogenes. In addition, parameter estimates suggest that somatic mutation rates increase with weight, likely because of reduced investment in somatic maintenance. Regression of cancer mortality on body weight and lifespan shows that 56% of cancers in dogs are the result of mutations arising from somatic cell division, compared to 66% in humans. A further 7% of cancers may be due to inherited recessive mutations deactivating tumour suppressor genes, as indicated by the relationship between cancer mortality and breed germline homozygosity. Some of the remaining unexplained variation in cancer mortality may be explained by germline mutations underlying breed predispositions to specific cancers. Contrasting results with humans provide novel insights into the dynamics of carcinogenesis.