Publications : 2013

Thompson CM, Kirman CR, Proctor DM, Suh M, Hays SM, Haws LC, Harris MA. 2013. A chronic oral reference dose for hexavalent chromium. Presented at the Society of Toxicology’s 52nd Annual Meeting, March 10-14, San Antonio, TX.


Intestinal tumors have been observed in mice (but not rats) following chronic exposure to high concentrations of hexavalent chromium [Cr(VI)] in drinking water. Mice (but not rats) also exhibit histological lesions consistent with intestinal wounding, specifically villous blunting and crypt hyperplasia—collectively termed diffuse hyperplasia. Recent mode of action studies support that these tumors were indeed the result of chronic wounding and regenerative hyperplasia to repair the intestinal mucosa. Herein, we develop an oral reference dose (RfD) that is protective of the tumor precursor lesion (diffuse hyperplasia), and therefore is protective of intestinal cancer. A rodent physiologically based pharmacokinetic (PBPK) model was used to predict internal dose measures for chromium in the duodenum, jejunum, and ileum of mice under the conditions of the 2-year bioassay. These internal dose metrics together with corresponding incidences for diffuse hyperplasia in each intestinal segment were used to characterize the dose-response relationship for the small intestine in a single plot containing a robust dataset with as many as 24 data points. Points of departures (PODs) were derived using benchmark dose modeling and global nonlinear regression, with models providing acceptable fits differing <3- fold. Human equivalent lifetime average dose values were estimated for each POD using two different methods of extrapolation with the human PBPK model for chromium. Dividing the PODs by uncertainty factors (UFs) of 10-30 yields a range of 8 RfD values (2 modeling approaches × 2 human equivalent dose methods × 2 UF values). The resulting RfD range is protective against diffuse hyperplasia, and is therefore protective of both noncancer and cancer effects in the small intestine. This range of RfD values leads to acceptable Cr(VI) concentrations in drinking water that are greater than those typically found in drinking water sources (≤5 μg/L).