Toxicokinetics of acrylamide in rats and humans following single oral administration of low doses

In vivo doses of acrylamide and glycidamide in humans after intake of acrylamide-rich food

For assessment of cancer risk from acrylamide (AA) exposure through food, the relation between intake from food in humans and the in vivo doses (area under the concentration-time curve, AUC) of AA (AUC-AA) and of its genotoxic metabolite glycidamide (GA) (AUC-GA) is used as a basis for extrapolation between exposure levels and between species. In this study, AA-rich foods were given to nonsmokers: a high intake of 11 μg AA/kg body weight (bw) and day for 4 days or an extra (medium) intake of 2.5 μg AA/kg bw and day for a month. Hemoglobin (Hb)-adduct levels from AA and GA, measured in blood samples donated before and after exposures, were used for calculation of AUC-AA and AUC-GA using reaction rate constants for the adduct formation measured in vitro. Both AA- and GA-adduct levels increased about twofold after the periods with enhanced intake. AUC for the high and medium groups, respectively, in nanomolar hours per microgram AA per kilogram bw, was for AA 212 and 120 and for GA 49 and 21. The AA intake in the high group was better controlled and used for comparisons with other data. The AUCs per exposure dose obtained in the present human study (high group) are in agreement with those previously obtained at 10(2) times higher exposure levels in humans. Furthermore, the values of AUC-AA and AUC-GA are five and two times higher, respectively, than the corresponding values for F344 rats exposed to AA at levels as in published cancer bioassays.

Development of a physiologically-based toxicokinetic model of acrylamide and glycidamide in rats and humans

Physiologically-based toxicokinetic ("pharmacokinetic") (PBPK or PBTK) modeling can be used as a tool to compare internal doses of acrylamide (AA) and its metabolite glycidamide (GA) in humans and rats. An earlier PBTK model for AA and GA in rats was refined and extended to humans based on new data. With adjustments to the previous parameters, excellent fits to a majority of the data for male Fisher 344 rats were obtained. Kinetic parameters for the human model were estimated based on fit to available human data for urinary metabolites of AA, and levels of hemoglobin adducts of AA and GA measured in studies in which human volunteers ingested known doses of AA. The simulations conducted with the rat and human models predicted that rats and humans ingesting comparable levels of AA (in mg/kg day) would have similar levels of GA in blood and tissues. This finding stands in contrast to the default approach that assumes a 3.2-fold increase in human risk due to pharmacokinetic differences between rats and humans. This model was used in a companion paper to estimate safe levels of ingested AA.