Towards a more robust comparative oncology: a Bayesian reanalysis of Peto's paradox and discussion of comparative cancer risk studies in vertebrates.

The multistage carcinogenesis model predicts that cancer risk should increase with body size and longevity owing to greater cell numbers and divisions, which provide more opportunities for mutations. However, the perceived lack of such associations across species, named 'Peto's paradox', suggests that larger or longer-lived animals may have evolved enhanced cancer suppression mechanisms. Empirical tests of this paradox have been limited by data availability, but large-scale zoo datasets now enable comparative analyses of cancer prevalence in vertebrates. Currently used statistical methods, however, often fail to adequately account for uncertainty in key model parameters. In this study, we use Bayesian methods to reanalyse these datasets and explore Peto's paradox, emphasizing the importance of quantifying uncertainty in comparative oncology. Our results show that body mass is positively associated with malignancy risk in mammals and amphibians, while it is negatively associated with cancer mortality in mammals. Longevity is positively associated with malignancy risk in non-avian sauropsids and amphibians. However, these relationships are accompanied by effect sizes with substantial uncertainty, primarily owing to small sample sizes. Through simulations, we demonstrate the limitations of current datasets and models. We also discuss the broader implications of Peto's paradox and suggest recommendations for improving future research on cancer risk across species.