Physiologically Based Kinetic Models for the Alkenylbenzene Elemicin in Rat and Human and Possible Implications for Risk Assessment

The present study describes physiologically based kinetic (PBK) models for the alkenylbenzene elemicin (3,4,5-trimethoxyallylbenzene) in rat and human, based on the PBK models previously developed for the structurally related alkenylbenzenes estragole, methyleugenol, and safrole. Using the newly developed models, the level of metabolic activation of elemicin in rat and human was predicted to obtain insight in species differences in the bioactivation of elemicin and read across to the other methoxy allylbenzenes, estragole and methyleugenol. Results reveal that the differences between rat and human in the formation of the proximate carcinogenic metabolite 1'-hydroxyelemicin and the ultimate carcinogenic metabolite 1'-sulfoxyelemicin are limited (<3.8-fold). In addition, a comparison was made between the relative importance of bioactivation for elemicin and that of estragole and methyleugenol. Model predictions indicate that compound differences in the formation of the 1'-sulfoxymetabolites are limited (<11-fold) in rat and human liver. The insights thus obtained were used to perform a risk assessment for elemicin using the margin of exposure (MOE) approach and read across to the other methoxy allylbenzene derivatives for which in vivo animal tumor data are available. This reveals that elemicin poses a lower priority for risk management as compared to its structurally related analogues estragole and methyleugenol. Altogether, the results obtained indicate that PBK modeling provides an important insight in the occurrence of species differences in the metabolic activation of elemicin. Moreover, they provide an example of how PBK modeling can facilitate a read across in risk assessment from compounds for which in vivo toxicity studies are available to a compound for which only limited toxicity data have been described, thus contributing to the development of alternatives for animal testing.