Assessment of the Relevance of Toxicological Findings in the Development of an Oral Reference Dose for GenX
Author Block: C. M. Thompson1, C. Ring2, L. Pham3, G. A. Chappell3, and L. C. Haws2. 1ToxStrategies Inc., Katy, TX; 2ToxStrategies Inc., Austin, TX; and 3ToxStrategies Inc., Ashville, NC.
Detection of GenX, an alternative to the long-chain per- and polyfluoroalkyl substances (PFAS), in water sources near production facilities has led to the need for development of toxicity criteria protective of oral exposure. Assessment of available toxicity data (i.e., chronic, subchronic, reproductive, and developmental toxicity studies in rats and mice) demonstrated that although the liver is a primary target organ in both rats and mice, GenX has been reported to cause other effects such as reduced fetal and pup body weight, reduced pup survival, and altered mammary gland development. The strength of evidence and relevance for developmental effects was assessed, including an analysis of confounding effects such as reduced maternal feed intake in high dose groups and transcriptomic evidence for enrichment of altered glucose metabolism and gluconeogenesis. Liver effects in mice were determined to be consistent with peroxisome proliferator‐activated receptor‐α (PPARα) activation, including increased liver weight, liver hypertrophy, and increased peroxisomal enzyme activity. Re-evaluation of liver tissue from previously conducted studies in mice (2010) using more current histopathological diagnostic criteria demonstrated that lesions previously diagnosed as ‘single cell necrosis’ were in fact apoptosis. This was subsequently corroborated by immunostaining for activated caspase-3 as well as transcriptomic pathway enrichment analyses. GenX was also found to increase mitosis in the liver at the same doses as apoptosis, which is consistent of rodent-specific cell cycle changes induced by PPARα activators. The involvement of PPARα and absence of liver necrosis, fibrosis, and inflammation indicate that the mouse liver lesions have limited relevance for risk assessment. Overall, the analyses support development of an RfD based on liver lesions observed in rats exposed for 2 years. Modeling these lesions using both frequentist and Bayesian benchmark dose (BMD) models, resulted in deterministic and probabilistic RfD values for GenX of 0.02 and 0.01 mg/kg/ day, respectively. Incorporation of the more conservative of the two RfD values in standard equations for development of a maximum contaminant level goal yielded an acceptable concentration in drinking water of 70 ppb.