LONG-TERM CHEMICAL CARCINOGENESIS EXPERIMENTS FOR IDENTIFYING POTENTIAL HUMAN CANCER HAZARDS - COLLECTIVE DATABASE OF THE NATIONAL-CANCER-INSTITUTE AND NATIONAL-TOXICOLOGY-PROGRAM (1976-1991)

The carcinogenicity database used for this paper originated in the late 1960s by the National Cancer Institute (NCI) and since 1978 has been continued and made more comprehensive by the National Toxicology Program (NTP). The extensive files contain, among other sets of information, detailed pathology data on more than 400 long-term (most often 24-month) chemical carcinogenesis studies, comprising nearly 1600 individual experiments having at least 10 million tissue sections that have been evaluated for toxicity and carcinogenicity. Using this data set we have a) determined the concordance in carcinogenic responses between rats and mice to be 74% and between sexes to be 85% (rats) to 87% (mice); b) discovered that using male rats and female mice would have identified correctly 95% of the positive or no evidence chemical carcinogenicity results obtained using the mom extensive protocol; c) established a historical control file of tumor incidence data; d) evaluated the false positive rate in the interpretation of carcinogenesis studies, concluding that this rate is probably no more than 7 to 8%; e) compiled listings of chemicals having like carcinogenic target sites for each of the 37 organs or systems for which histopathology diagnoses have been recorded routinely; f) demonstrated that evaluation of site-specific carcinogenic effects are preferable to doing analyses based on overall (all sites combined) tumor rates; g) learned that few chemicals cause only benign tumors or only liver tumors, the most common target site for chemically induced cancers; h) identified key sources of variability in tumor rates in long-term carcinogenesis studies; i) ascertained that corn oil gavage or gavage per se exhibits little, if any, adverse impact on long-term studies; j) showed that the Salmonella multistrain assay was as good as a battery of four short-term in vitro tests for predicting in vivo carcinogenicity, yet was only 66% concordant with an 89% positive predictivity and a 55% negative predictivity; k) investigated the relationship between chemically induced toxicity and chemically associated carcinogenicity, finding that few chemicals cause tumors only at the highest exposure concentration. These (and other) derived compilations are most useful for maintaining a historic and objective perspective when evaluating the carcinogenicity of contemporary experiments and for identifying potential carcinogenic hazards to humans.