Carcinogen prediction in the laboratory: a personal view

Abstract

Although carcinogens can be divided into various categories, i. e. viruses, physical agents and synthetic and naturally occurring chemicals, it is the latter that give rise to the greatest concern because of their number, quantity and distribution. Present methods of testing chemicals for potential carcinogenicity rely in the main on administration of these at maximally tolerated dose levels to animals, usually rodents, for the animals’ lifetime. Such tests would be economically impractical for all chemicals to which man is exposed. New methods have recently been introduced to screen large numbers of chemicals quickly and cheaply which rely on the unifying hypothesis that all carcinogenic chemicals are electrophiles or must be converted to such by metabolism. These methods will be reviewed and compared with traditional methods of carcinogenicity testing, particularly as to their role in attempting to predict hazard to man.