Publications : 2017

Thompson C, Kirman C, Suh M, Proctor D, Haws L, Harris M, Hays S. 2017. Risk assessment of oral exposure to Cr(VI): Integration of mode of action, pharmacokinetics, and dose-response modeling. Poster presented at Society of Toxicology Annual Meeting, March 14, Baltimore, MD.


In 2008, the NTP finalized a 2-year drinking water study in F344 rats and B6C3F1 mice exposed to up to 180 ppm hexavalent chromium [Cr(VI)]. To better inform risk assessment of the neoplastic and nonneoplastic outcomes observed in the 2-year study, a research effort was undertaken to investigate the mode of action (MOA) and pharmacokinetics of orally ingested Cr(VI). MOA data, including that from genotoxicity assays in target tissues, provides support for nonlinear risk assessment of neoplastic lesions. In addition, newly developed rodent and human PBPK models indicate nonlinear pharmacokinetics for Cr(VI) disposition also supports nonlinear risk assessment approaches for both neoplastic and nonneoplastic lesions. To facilitate critical effect selection, we modeled nonneoplastic lesions selected in the U.S. EPA draft assessment on Cr(VI). The two most sensitive tissues identified by EPA were the liver (in mice and rats) and the small intestine (mice). Based on MOA information and PBPK models, these lesions were modeled using internal dose metrics relevant for each tissue (i.e., the amount of Cr(VI) entering the intestinal mucosa and blood chromium levels). BMD/L10 values were derived based on internal dose metrics in rodents, and human PBPK models capable of accounting for lifestage and health-related differences in Cr(VI) disposition were used to compute the lifetime average daily dose leading to the same internal dose metric. After applying a 30-fold UFC, preliminary RfDs based on chronic inflammation in the rat liver and histiocytic infiltration in the mouse liver were 0.001 and 0.002 mg/kg-day, respectively. The preliminary RfD based on diffuse epithelial hyperplasia in the mouse small intestine was 0.003-0.004 mg/kg-day. These RfD values roughly equate to 32-128 ppb Cr(VI) in drinking water. Notably, the BMDL10 values for intestinal hyperplasia were more than 10-fold lower than for intestinal cancer regardless of whether male and female mice were modeled individually or combined. These findings are in agreement with the MOA indicating that protection against nonneoplastic lesions protects against intestinal cancer.