ABSTRACT
Toxaphene is a liver tumor promoter in B6C3F1 mice but not in F344 rats or hamsters. Recent studies demonstrate that key events leading to the mouse liver tumor response for toxaphene are mediated by activation of the constitutive androstane receptor (CAR). Benchmark dose modeling was conducted on available data for five endpoints in B6C3F1 mouse liver tissue or cultured liver cells (tumor response, cytotoxicity, proliferation, gap junction intercellular communication inhibition, and CAR-mediated CYP2B10 induction) and for CAR activation in human HepG2 cells, all reported in previous studies. The available evidence supports a nonlinear CAR-mediated mode of action (MOA) for toxaphene-induced mouse tumors including demonstration of a J-shaped dose-response pattern for human CAR activation, indicating that linear risk extrapolation at low doses is not supported for this MOA. Based on analysis of benchmark dose lower confidence limits at 10% response (BMDL10) and no observed effect levels (NOELs) for potential key events in the mouse liver tumor MOA for toxaphene, an RfD of 0.13 mg/kg-d is proposed based on a the BMDL10 for human CAR activation in human HepG2 cells. This value is below candidate RfD values based on BMDL10 estimates for both mouse liver tumors and mouse hepatocyte proliferation and therefore can be considered protective for human risk of liver tumor promotion and other CAR-mediated adverse health effects based on available data.
Acknowledgments
The authors thank Dr. James E. Klaunig and his research team including Dr. Zemin Wang and colleagues at the Department of Environmental Health, Indiana University, Bloomington, Indiana for their assistance in providing raw data for our benchmark dose modeling.
Conflict of interest
The authors are employees of Exponent, Inc., a United States firm that provides scientific research and consulting services under contract to private sector and governmental clients.