Choose your death: adaptive cell senescence predicts a late-life decrease of cancer prevalence
Résumé
Carcinogenesis shares many proximal factors with ageing (including genomic instability and telomere attrition), and is fueled by the life-long, random accumulation of (epi)genetic mutations. Cancer incidence should therefore increase with age, and cancer risk be highest for longest-lived species. However, cancer epidemiology is actually quite distinct from the epidemiology of ageing-related diseases. Cancer incidence curves were shown to stabilize and decrease for older ages, both in humans and in homogenic rats; while there is no statistical association between cancer risk and a species’ longevity. The solution to these paradoxes could lie in the senescent-cells theory of ageing: while senescent cells take no longer part to tissue renewal and functioning, they do not enter carcinogenesis either. Cell senescence could therefore be an adaptation to cancer. We implemented a multi-scale evolutionary demography model in which carcinogenesis, cell senescence, and apoptosis are competing to determine cell fates; and longevity decreases with the accumulation of senescent and cancerous cells. This model predicts that the accumulation of senescent cells can be adaptive for a range of parameters compatible with biological data. The epidemiological pattern predicted by this model depicts a decrease of cancer incidence at older ages. Interestingly, this model is compatible with a large range of sizes and longevities and could explain why longest-lived, largest animals do not have the highest cancer risks.
