Multistage carcinogenesis: implications for risk estimation.

In undertaking a quantitative estimation of carcinogenesis risk, it is essential to keep in mind that carcinogenesis is a multistage process, and that each stage can be affected by different classes of risk factors. Furthermore, different mechanisms are involved in the various stages of carcinogenesis. Thus, a dose-response analysis of one given factor cannot provide an accurate estimation of carcinogenic risk. Carcinogenic risk estimation is usually undertaken for a specific chemical or group of chemicals; however, the concept of multistage carcinogenesis is based on biological processes and not on the mechanisms of action of the agents involved. It is therefore important to consider three related, but different, factors involved in carcinogenesis: stage, agent, and activity of agent. This is especially important in developing a short-term test for stage-related risk factors, such as tumor-promoting agents. For this reason, carcinogens should not be classified according to only one chemical activity. This article briefly reviews the cellular and molecular mechanisms involved in multistage carcinogenesis, and discusses their implications for risk estimation. Special consideration is given to the effect of treatment frequency on the response of tumor-promoting agents, as seen in long-term tests in experimental animals. It is proposed that exposure frequency be taken into account together with exposure dose.