You are here:
Integrating Exposure, Pharmacokinetics, And Dosimetry With In Vitro Dose-Response Data To Evaluate Chemical Risk
Citation:
Leonard, J., H. El-Masri, C. Stevens, K. Mansouri, AND C. Tan. Integrating Exposure, Pharmacokinetics, And Dosimetry With In Vitro Dose-Response Data To Evaluate Chemical Risk. 2018 SOT Annual Meeting and ToxExpo, San Antonio, TX, March 11 - 15, 2018.
Impact/Purpose:
High throughput toxicity testing holds the promise of providing data for tens of thousands of chemicals that currently have no data due to the cost and time required for animal testing. Interpretation of these results require information linking the perturbations seen in vitro with adverse outcomes in vivo and requires knowledge of how estimated exposure to the chemicals compare to the in vitro concentrations that show an effect. This abstract describes methods for generating exposure and ADME predictions at differing levels of detail to provide these connections.
Description:
High throughput in vitro toxicity testing of hundreds to thousands of chemicals across any number of biological endpoints allows for rapidly assessing human and ecosystem health impacts, thus reducing resources associated with traditional animal testing. In order to apply these in vitro data in risk assessment, it is necessary to compare chemical concentrations sufficient to produce bioactivity in vitro with those capable of perturbing a relevant in vivo target, as determined by exposure factors and pharmacokinetic (PK) properties. This work presents a variety of computational approaches, in case studies, to demonstrate how uncertainty in exposure potentials and PK properties can influence the applicability of in vitro dose-responses data in risk assessment. First, qualitative screening of a large assemblage of chemicals with in vitro acetylcholinesterase inhibition activity was accomplished through in silico prediction of ADME properties to determine if such properties are favorable for reaching the target enzyme. This analysis resulted in a 33% reduction in the number of candidates, and the remaining compounds were further ranked based on quantitative estimates of in vivo bioactivities derived using a PK/pharmacodynamic (PD) model. Next, with sufficient exposure and PK data, external points of departure were predicted for six thyroid peroxidase inhibitors using a physiologically based PK/PD model. It was found that investigating exposure, potency, or PK individually is insufficient for evaluating chemical risk. Finally, potential metabolites were predicted for chemicals shown to be inactive across eighteen estrogen receptor (ER)-related in vitro assays, and ER binding activity for these potential metabolites was investigated using a qualitative structure activity relationship model. Of the 1,400 inactive parents investigated, 20% were found to have at least one metabolite predicted to be active. The integration of in vitro testing with computational approaches will help to improve confidence in evaluation of chemical risk by identifying those chemicals requiring no further testing and those inactive parents capable of producing potentially active metabolites.
