A biological model of carcinogenesis has been developed that can be expressed mathematically, and therefore can be studied using computer-based analyses. It is based on several assumptions: Carcinogenesis occurs in two stages: initiation and transformation (to malignant tumors); the carcinogenic events can occur only in stem cells or their functional equivalents; these events can occur only during the active part of the cell cycle; each of the events occurs in a probabilistic fashion. Cell dynamics are thus an extremely important part of carcinogenesis. Any agent can have an impact on the carcinogenic process by either directly altering the genome (genotoxic) or increasing the proliferative rate of the tissues: increasing the number of cell divisions through which a spontaneous alteration in the genome can occur; or an agent can affect both of these. Effects on the genome and on cell proliferation can have different dose-responses. Extrapolation to low doses requires consideration of the dose-response for each effect. Differences in mechanisms affecting cell proliferation and genetic changes need to be considered in determining tresholds. The model was originally validated utilizing tumor incidence data from multiple experiments with the carcinogen, N-[4-(5-nitro-2-furyl)2-thiazolyl]-formamide (FANFT), in rats. FANFT is a strong mutagen, is metabolically activated to a reactive electrophile, binds to DNA, and also increases cell proliferation. Modeling analyses demonstrate that the tumor dose-response curve for FANFT can be explained based on a combination of the individual dose-response of its genotoxic and cell proliferation effects. Modeling of a nongenotoxic compound, sodium saccharin (NaS), has also been evaluated. It is not metabolized to a reactive electrophile (it is actually nucleophilic), does not bind to DNA, and is not mutagenic. Nevertheless, in two-generation experiments at high doses it induces a significant incidence of bladder tumors in male rats. It is also a strong tumor-promoting substance following chemical initiation or bladder ulceration. These complex protocols can be readily explained by the proliferative response induced in the urothelium following NaS administration. As expected for a nongenotoxic chemical, there is no effect on the probability of initiation or transformation. Unlike the sodium salt, high doses of the calcium and acid forms of saccharin do not increase cell proliferation significantly, and would not be anticipated to induce tumors. Also, there appears to be a threshold effect related to dose of NaS and the induction of urothelial proliferation. Since cell proliferation is the mechanism by which NaS induces bladder tumors in rats, it is expected that there is also a threshold with respect to carcinogenesis. By allowing for agents to be defined in terms of their ability to affect the genome directly or to act as cell proliferators, model-based analyses provide a rational basis for extrapolating from high doses in animal experiments to low doses in assessing risk for humans.
ASJC Scopus subject areas