A generalized theory of somatic evolution

Abstract

The modern Multi-Stage Model of Carcinogenesis (MMC) was developed in the 1950s through the 70s and postulated carcinogenesis as a process of rounds of Darwinian selection favoring progressively more malignant cell phenotypes. Through this period, almost nothing was known about driver mutations in cancers. Also, stem cells and cellular tissue organization were poorly characterized. The general multi-stage process was later confirmed by experimental studies, and cancer risk and incidence has been explained as primarily a function of mutation occurrence. However, the MMC has never been formally tested for its ability to account for current knowledge about cancer evolution. In particular, different numbers of cancer drivers required for different cancers and vast discrepancies in the organization of stem cell compartments for different tissues appear inconsistent with the very similar age distribution of the vast majority of cancers. In this regard, the initial theoretical idea underlying MMC is often over-interpreted with little connection to modern evidence, and a general theory of somatic evolution still does not exist. In this study, we applied Monte Carlo modeling and demonstrated the effect of various parameters, such as mutation rate, mutation effects and cell division, on the MMC performance. Our modeling demonstrates that the MMC requires considerable modification in order to describe cancer incidence. We elucidate the required conditions for how somatic cell selection should operate within the MMC in order to explain modern data on stem cell clonality and cancer, and propose a generalized theory of somatic evolution based on these results.