You are here:
Development and analysis of high throughput physiologically based pharmacokinetic/toxicokinetic (PBPK/TK) dermal exposure model
Citation:
Meade, A., M. Evans, AND J. Wambaugh. Development and analysis of high throughput physiologically based pharmacokinetic/toxicokinetic (PBPK/TK) dermal exposure model. SOT, Nashville, TN, March 19 - 23, 2023. https://doi.org/10.23645/epacomptox.22637536
Impact/Purpose:
The purpose of this poster presentation is to present the httk dermal exposure model at the Society of Toxicology (SOT) 2023 Annual meeting in Nashville, TN. With this presentation, we hope to make the generalized dermal exposure PBPK/TK model found in the httk R package available to toxicologists and scientists interested in dermal exposure for high throughput toxicokinetics. This model would be useful to anyone who needs to perform toxicokinetics for dermal exposure and has little or no coding experience.
Description:
Dermal absorption of chemicals represents an important route of exposure in pharmaceutical, occupational, and environmental settings. There are thousands of chemicals in use with little or no toxicity or toxicokinetic data, and dermal absorption data are lacking, unsurprisingly, for many of these chemicals. One alternative is to estimate human toxicokinetics using high-throughput methods. Accordingly, the aim of this study was to develop a generalized physiologically based pharmacokinetic/toxicokinetic (PBPK/TK) dermal exposure model for the R package httk. For dermal exposures this model can be used in a high-throughput manner to estimate human blood and tissue concentrations and estimate risk for many chemicals. The structure of the dermal PBPK/TK model was based on Campbell, Clewell, Gentry, Anderson, and Clewell, Computational Toxicology, 929, 439 (2012). Chemical-specific metabolism and protein binding data were obtained from the literature as collected by R package httk. The physiochemical property data needed for the model were obtained from the EPA CompTox Chemicals Dashboard. Here we have compared three different in vitro methods for estimating chemical-specific and vehicle-specific permeability in the skin: 1) Potts-Guy [Potts and Guy, Pharm Res, 9, 663 (1992)], 2) Chen-Lian [Chen, Han, Saib, and Lian, Pharm Res, 32, 1779 (2015)], and 3) UK Surrey [based on Wang, Chen, Lian, and Han, Int J Pharm, 398, 114 (2010) and available as open source in Python code]. The model was constructed to allow comparison of dermal, oral, and intravenous exposures. Over 26 exposure scenarios across 14 chemicals were modeled and compared to published concentration-time in vivo data from the EPA CvT database [Sayre, Wambaugh, and Grulke, Scientific Data, 7, 122 (2020)]. Of these 14 chemicals, four are pharmaceuticals while the other 10 are occupationally or environmentally relevant, 10 of the 14 are lipophilic (log of the octanol to water partition coefficient, logP > 2), three of the 14 have low water solubility (< 10 mg/L), four of the 14 have high water solubility (> 1,000 mg/L), and eight of the 14 are considered volatile. The Root Mean Squared Error (RMSE) between log-transformed simulated and observed concentrations was calculated to determine how well the model captures the behavior of this data when using the three methods for dermal permeability. Based on these results, the Chen-Lian and UK Surrey methods outperformed the Potts-Guy method of calculating dermal permeability. We also found that model simulations of highly volatile chemicals tend to have worse fits to the given data with larger RMSE. This suggests that evaporation of chemical from the skin is an important component in the model, since highly volatile chemicals evaporate more readily. The views expressed in this abstract are those of the authors and do not necessarily represent the views or the policies of the U.S. Environmental Protection Agency.
URLs/Downloads:
DOI: Development and analysis of high throughput physiologically based pharmacokinetic/toxicokinetic (PBPK/TK) dermal exposure model
SOT2023_POSTER_AMEADE.PDF (PDF, NA pp, 7960.376 KB, about PDF)