Review and evaluation of the NCI/NTP carcinogenesis bioassays.

A comparison of the carcinogenesis bioassay results obtained by the National Cancer Institute (NCI) and the National Toxicology Program (NTP) indicates that approximately one-half of the bioassays directed by both institutions were positive for carcinogenicity. The more recent 85 bioassays completed by NTP reveal a higher proportion of studies interpreted as demonstrating no evidence of carcinogenicity than represented in the initial 198 bioassays conducted by NCI. Of the 100 NCI bioassays that were not positive for carcinogenicity 3 (3%) were classified in the category of "no evidence for carcinogenicity in two animal species." Of the 43 NTP bioassays that were not positive for carcinogenicity 36 (84%) were placed in the category of "no carcinogenic effects." The reason for this shift from a 33:1 positive to negative ratio in the NCI bioassays to an approximately 1:1 ratio in the NTP bioassays appears to be a difference in interpretation of the adequacy of the testing. For example, 6 of the 36 NTP negative bioassays involved testing in only one species. Uniform criteria for concluding that a bioassay is negative must be developed and the results of all existing and future carcinogenesis bioassays must be interpreted with these exclusive criteria. Other bioassay problems are explored, including the incomplete validation of the carcinogenesis bioassay protocol by confirmatory results with positive and negative reference agents, the apparent lack of bioavailability data for some orally administered negative compounds, the continued use of mouse hepatic neoplasia as a single discriminating parameter, the variability in the inter- and intrastudy incidence of spontaneous tumors, and the continued reliance on the maximum tolerated dose.