ABSTRACT
I. INTRODUCTION For cancer risk assessment of chemicals, efforts have been made to distinguish between genotoxic mechanisms (involving direct DNA changes, such as alkylation or chromosome breakage) and nongenotoxic mechanisms (acting at some other site in the cell that does not involve genetic material). The rationale for this distinction is the assumption that genotoxic actions may have no effective threshold (a linear low-dose response), whereas nongenotoxic effects require a suprathreshold dose to disturb a homeostatic system sufficient to elicit a toxic response. This also implies that different scientific methods may be appropriate to assess cancer risk at low doses, depending on the chemical’s mechanism of action. Methods that assume no threshold imply a potential risk at any dose, whereas the opposite assumption is that a safe dose can be estimated. Thus, the question of potential genotoxicity is important for risk assessment of a chemical shown to be carcinogenic in experimental animals or humans. For those chemicals for which one or more bioassays show an increase in tumors at one or multiple sites, the question of potential genotoxicity-mediated carcinogenesis becomes the overriding factor in the toxicity evaluation. The significance is that if a chem-
ical is assumed to be genotoxic, the de minimis risk level might be estimated at a level much lower than the estimated safe level based on a non-linear doseresponse. This difference may, in turn, make an enormous difference in regulation of the chemical in air, water, food, or toxic waste. The approach for cancer risk assessment has been to assume a linear low-dose response for carcinogens unless there is sufficient evidence to prove otherwise. In the past 10 years, increased efforts have focused on the study of carcinogens that may act by way of a nongenotoxic mechanism. This chapter discusses recent developments in the area of cancer risk assessment based on nongenotoxic or epigenetic modes of action, with various interpretations of the adequacy of data on low-dose responses for risk assessment.
