Influence of genetic composition of test-animal populations on chronic toxicity studies used for risk estimation.

A lifespan exposure of mice to benzidine dihydrochloride was conducted for 33 m using both sexes of two populations of mice with the same gene pool. One population was the genetically homogeneous F1 hybrid produced by crossing BALB/cStCrlC3Hf/Nctr males with C57BL/6jfC3Hf/Nctr females. The second population consisted of genetically heterogenous monohybrid cross (MC) offspring produced by mating the F1 hybrids inter se. Data comparisons were made to determine if gene distribution among members of a population affects the response to a toxic insult. Endpoints tested consisted of mortality, liver tumor incidence and time of tumor onset, mortality from reticulum-cell sarcoma, and body weights. In most instances it was noted that among animals not dosed (controls), the F1 population had lower background incidence of lesions and lived longer than the MC population. However, among the dosed animals, the F1 mice were generally more susceptible to the toxic agent and developed higher incidences of the chemically induced lesions than did the MC population. The F1 hybrid population gave a more conservative estimate of risk than did the MC population. The calculation of the liver tumor risk for these two populations showed that lifespan exposure to benzidine would be predicted to result in a larger number (higher risk) when using the F1 data. A 4.5-fold difference in the toxic response was observed between the F1 females and the MC males. This emphasizes the importance of gene distribution in risk estimation studies.