Bioassay of trifluralin for possible carcinogenicity.

A bioassay for possible carcinogenicity of technical-grade trifluralin was conducted using Osborne-Mendel rats and B6C3F1 mice. Analysis of the technical product established the presence of 84 to 88 ppm dipropylnitrosoamine. The product was administered in the feed, at either of two concentrations, to groups of 50 male and 50 female animals of each species. Fifty animals of each sex were placed on test as controls for the rat bioassay, while 20 of each sex were utilized as controls for the mouse study. The time-weighted average high and low dietary concentrations of trifluralin were, respectively, 8,000 and 4,125 ppm for male rats, 7,917 and 4,125 ppm for female rats, 3,744 and 2,000 ppm for male mice, and 5,192 and 2,740 ppm for female mice. After a 78-week treatment period, there was an additional observation period of 33 weeks for rats and 12 weeks for mice. For female mice the association between increased dosage and elevated incidence of hepatocellular carcinomas was significant (0/20, 12/47, and 21/44 of the control, low dose, and high dose, respectively) as was the relationship between dose and incidence of alveolar/bronchiolar adenomas. Significance of incidence for both types of tumors was supported by tests for significance at each dose level. Squamous-cell carcinomas of the stomach were observed in dosed female mice, but not in controls. Although incidences of these tumors were not statistically significant, they are unusual lesions in B6C3F1 mice and are considered to be treatment-related. Neoplasms observed in treated rats were types that have occurred spontaneously in this strain and were apparently unrelated to trifluralin treatment. Evaluation of the results of this bioassay indicates that technical-grade trifluralin is a carcinogen in female B6C3F1 mice, being associated in these animals with an elevated incidence of hepatocellular carcinomas, alveolar/bronchiolar adenomas and squamous-cell carcinomas of the forestomach. Sufficient evidence was not provided for the carcinogenicity or tumorigenicity of trifluralin in male B6C3F1 mice or in Osborne-Mendel rats of either sex.