Peto’s paradox and the promise of comparative oncology