SIGNIFICANCE OF NATURALLY-OCCURRING TUMORS IN EVALUATING THE CARCINOGENICITY OF A TEST COMPOUND: A REVIEW AND AN IMPROVED CARCINOGENICITY BIOASSAY FOR CHEMICALS

Abstract

Naturally-occurring tumors, which generally increase in incidence with age in control rats or mice in 2-year carcinogenicity bioassays, show occasionally an increased or decreased incidence in animals dosed with chemical substances. Analyses of these tumors showing dose-related increases or decreases in their incidence rates in the chronic toxicity and oncogenicity studies of chemicals including 11 tests using F344 rats and 14 tests using B6C3F1 mice carried out in our Center for the past 8 years, revealed that they were exclusively naturally-occurring tumors commonly seen in these animals. These lesions included pituitary tumor, C-cell tumor of the thyroid, and mononuclear cell leukemia in F344 rats and bronchiolar/alveolar adenoma of the lung, hepatocellular adenoma, hepatocellular adenoma plus carcinoma, malignant lymphoma, and forestomach papilloma in B6C3F1 mice. One approach to understand the changes in the incidence rates of the naturally-occurring tumors is to elucidate the biological effects of the test chemical. Another approach is to identify sets of modifying factors, including nutritional effects, hormonal-imbalances, and other age-associated lesions, as seen in the roles of chronic nephropathy, myeloproliferative changes, and hyperplasias of the various endocrine organs in the development of tumors in corresponding tissues. The profile approach to induced rodent tumors in the past has revealed several characteristics. Among them, first, induced tumors show early development, mostly within a period less than 52 or 78 weeks of the treatment. Second, the dose levels of each chemical appear likely to have a minimum as well as an upper limit in the amount needed to induce tumors. Consequently, the current carcinogenicity tests using unusually high amounts of chemicals in a pure form will be able to demonstrate tumor induction, if they are really carcinogenic, before the onset of the spontaneous tumors, which are mostly seen mixed with a variety of non-neoplastic lesions in the same groups of animals after 78 weeks of the treatment. This means that the detection of possible chemical carcinogens would be more efficient and accurate by shortening the test period from 104 to 78 weeks in tests using F344 rats and B6C3F1 mice. Addition of a satellite test for interim observation of the pathological findings at 52 weeks, to this new way of testing will give more useful information on early changes and the morphogenesis of tumors induced by chemical substances.