Derivation of oral cancer slope factors for hexavalent chromium informed by pharmacokinetic models and in vivo genotoxicity data

Hexavalent chromium [Cr(VI)] is present in drinking water from natural and anthropogenic sources at approximately 1 ppb. Several regulatory bodies have recently developed threshold-based safety criteria for Cr (VI) of 30-100 ppb based on evidence that small intestine tumors in mice following exposure to >= 20,000 ppb are the result of a non-mutagenic mode of action (MOA). In contrast, U.S. EPA has recently concluded that Cr(VI) acts through a mutagenic MOA based, in part, on scoring numerous in vivo genotoxicity studies as having low confidence; and therefore derived a cancer slope factor (CSF) of 0.5 (mg/kg-day)(-1), equivalent to similar to 0.07 ppb. Herein, we demonstrate how physiologically based pharmacokinetic (PBPK) models and intestinal segment specific tumor incidence data can form a robust dataset supporting derivation of alternative CSF values that equate to Cr(VI) concentrations ranging from below background to concentrations similar to those derived using threshold approaches-depending on benchmark response level and risk tolerance. Additionally, we highlight weaknesses in the rationale EPA used to discount critical in vivo genotoxicity studies. While the data support a non-genotoxic MOA, these alternative toxicity criteria require only PBPK models, robust tumor data, and fair interpretation of published in vivo genotoxicity data for Cr(VI).

Automated summary

This article describes the presence of hexavalent chromium (Cr(VI)) in drinking water and the safety standards for its concentration. It explores the effects of different mechanisms on the carcinogenicity of Cr(VI). By using PBPK models and tumor incidence data, alternative safety standards for different Cr(VI) concentrations can be derived. Additionally, weaknesses in the rationale used by the U.S. EPA to discount in vivo genotoxicity studies are highlighted.