The brain-tumour issue in long-term toxicity studies in rats.

Problems arise in the measurement of the neurocarcinogenic potential of chemical substances in chronic toxicity studies because of the spontaneous occurrence of neoplasms in the brain and other organs of rats from the age of about 2 yr. Statistical analysis may be equivocal and must be accompanied by a thorough biological analysis, to determine the presence or absence of the following characteristic effects of neurocarcinogenic agents: a reliable and consistent increase in brain-tumour incidence beyond the expected control level, a decrease in the age at which tumours appear and/or in survival, a dose-effect relationship, a greater effect on embryonal and foetal brain cells than on those of adults, a shift to more anaplastic types of tumour and the finding of preneoplastic lesions. These criteria have been met in chronic tests on the neurocarcinogens ethyl- and methylnitrosourea and some have been met in results obtained with the weak carcinogen methyl methanesulphonate. However, none of these criteria were met in the case of a test compound subjected to two 2-yr studies (one involving transplacental exposure), although brain tumours occurred in the controls and in all the experimental groups. The test substance is not considered to be a neurocarcinogenic agent.