A physico-chemical screening test for chemical carcinogens: the ke test.

A pulse-conductivity technique was used to measure the rate at which excess electrons in liquid cyclohexane attach to carcinogens and non-carcinogens in order to determine if the electron attachment rate constant, ke, could be used to screen potential carcinogens. The keS of 114 chemicals are reported; these chemicals are among 182 that had previously been tested in a validation study of several short-term carcinogen-screening bioassays. The remaining 68 chemicals for which keS were not measured include chemicals that were unavailable, were not sufficiently stable or soluble in cyclohexane, or did not have a well-defined mol. wt. For the 114 chemicals that were tested, 35 are carcinogens, 50 are putative non-carcinogens and 29 have not been adequately tested or yielded equivocal responses in animal-test studies. Diffusion-controlled keS were measured for 27 of the 35 carcinogens tested whereas the keS of 45 of the 50 non-carcinogens were less than diffusion controlled. From these results, several measures of the predictive performance of using a diffusion-controlled ke to indicate a positive response to a carcinogen were calculated and compared with the Ames-test predictiveness in screening the same chemicals. The predictive criteria calculated were sensitivity, specificity, accuracy and predictive value, all of which were greater for the ke test than for the Ames test. Comparisons of the chemicals that yielded false-negative responses in the ke and Ames tests indicate a high degree of independence between the two which implies that the tests could be efficaciously used in a battery of short-term tests. Rationales are offered concerning the observed ke--carcinogenicity correlation and the apparent lack of the need for procarcinogens to be metabolically activated to yield a positive ke response.