A Bayesian Population PBTK Model for Multiroute Chloroform Exposures
Y.C. Yang, M. Ouyang, X. Xu, P.G. Georgopoulos
Environmental and Occupational Health Sciences Institute, UMDNJ - R.W. Johnson Medical School and Rutgers University
Physiologically based toxicokinetic (PBTK) modeling offers a rational basis for the extrapolation of toxicokinetic data from acute, high dose, experiments in animals, to chronic, low dose, exposures in humans. However, many PBTK model parameters are difficult to measure, and need to be extrapolated from related studies or optimized with biomarker data while fixing other model parameters at typical or mean literature values. However, using fixed parameters or developing the results as point estimates does not incorporate population variability in the toxicokinetics, and also does not consider the other uncertainties within the data. Bayesian methods can be applied to partially overcome these limitations by combining the prior and the data likelihood to produce posterior distributions of model parameters.
In this case study, time-series of exhaled breath measurements were used to assess inhaled and dermal exposures from use of chlorinated drinking water. An individual chloroform PBTK model with distributed parameter descriptions of skin transport is developed and then expanded into a population level model within a Bayesian hierarchical framework. Posterior distributions of physicochemical and biochemical parameters were estimated using the Markov Chain Monte Carlo (MCMC) method. Physiological and anatomical parameters were calculated using age and gender dependant deterministic equations and subject-specific information.
This work had been funded in part by the US Environmental Protection Agency under Cooperative Agreement # EPAR-827033 to the Environmental and Occupational Health Sciences Institute (EOHSI). The viewpoints expressed here are the responsibility of the authors and do not necessarily reflect the views of the USEPA or its contractors.