Nongenotoxic carcinogens: an extension of the perspective provided by Perera.

Perera recently discussed the very real problems that accompany any attempt to classify rodent carcinogens into two groups--genotoxic or nongenotoxic. Not the least of these problems is that no agreed definition of these two terms exist. Nonetheless, the current carcinogen databases, for example, that of the U.S. National Toxicology Program (NTP), clearly comprise two broad groups of carcinogens--DNA reactive, mutagenic and multiply carcinogenic chemicals, and others. The others appear to be nonreactive to DNA, are inactive in the primary mutagenicity assays, and usually elicit highly selective carcinogenic responses in animals. These two classes of carcinogen are illustrated by examples taken from the NTP database and are discussed within the possible context of the latter group not being active in humans or, if they are, only when a threshold dose has been exceeded, chronically.