Examination of low-incidence brain tumor responses in F344 rats following chemical exposures in National Toxicology Program carcinogenicity studies.

Neoplasms in the brain are uncommon in control Fischer 344 (F344) rats; they occur at a rate of less than 1% in 2-yr toxicity/carcinogenicity studies. Furthermore, only 10 of nearly 500 studies conducted by the National Toxicology Program (NTP) showed any evidence of chemically related neoplastic effects in the brain. Generally, the brain tumor responses were considered equivocal, because the characteristics of potential neurocarcinogenic agents (such as statistically significant increased incidences, decreased latency and/or survival, and demonstration of dose-response relationships) were not observed. A thorough examination, including comparisons with a well-established historical database, is often critical in evaluating rare brain tumors. Chemicals that gave equivocal evidence of brain tumor responses were generally associated with carcinogenicity at other sites, and many chemicals were mutagenic when incubated with metabolic activating enzymes. Other factors that were supportive of the theory that marginal increases in brain tumor incidence were related to chemical exposure were that (a) some of the tumors were malignant, (b) no brain neoplasms were observed in concurrent controls from some studies, and/or (c) brain tumors were also seen following exposure to structurally related chemicals. In 2-yr studies in F344 rats (studies conducted by the NTP), equivocal evidence of carcinogenicity was observed for the following 9 chemicals: isoprene, bromoethane, chloroethane, 3,3'-dimethylbenzidine dihydrochloride, 3,3'-dimethoxybenzidine dihydrochloride, furosemide, C.I. direct blue 15, diphenhydramine hydrochloride, and 1-H-benzotriazole. Glycidol was the only chemical evaluated by the NTP with which there was clear evidence of brain tumor induction in F344 rats. Clarification of the potential neurocarcinogenic risks of chemicals that produce equivocal evidence of a brain tumor response in conventional 2-yr rodent studies may be aided by the use of transgenic mouse models that exhibit genetic alterations that reflect those present in human brain tumors as well as by the use of in utero exposures.