The multistage theory of carcinogenesis and the age distribution of cancer in man.

The multistage theory of carcinogenesis is briefly reviewed, and the incidence function predicted by the theory is expressed as an "infinite sum" (convergent power series). This expression for the incidence function has the Armitage-Doll approximation as the first non-zero term and explicitly exhibits the dependence of the incidence function on the transition rates. It is then clear that if the transition rates are not small enough (approximately 10(-4/cell/yr), the Armitage-Doll approximation is inappropriate. Retention of one more term (in the series) leads to a better approximation in these cases. This is illustrated by a hypothetical example. Other assumptions implicit in the theory are discussed.