A Physiologically Based Pharmacokinetic Model for Trichloroethylene in the Male Long-Evans Rat

Abstract

A physiologically based pharmacokinetic (PBPK) model for trichloroethylene (TCE) in the male Long-Evans (LE) rat was needed to aid in evaluation of neurotoxicity data collected in this rodent stock. The purpose of this study was to develop such a model with the greatest possible specificity for the LE rat. The PBPK model consisted of 5 compartments: brain, fat, slowly perfused tissue, rapidly perfused viscera, and liver. Partition coefficients (blood, fat, muscle, brain, liver) were determined for LE rats. The volumes of the brain, liver, and fat compartments were estimated for each rat, with tissue-specific regression equations developed from measurements made in LE rats. Vapor uptake data from LE rats were used for estimation of Vmaxc. As blood flow values for LE rats were not available, values from Sprague-Dawley (SD) and Fischer-344 (F344) rats were used in separate simulations. The resulting values of Vmaxc were used to simulate tissue (blood, liver, brain, fat) TCE concentrations, which were measured during (5, 20, 60 min) and after (60 min of TCE followed by 60 min of air) flow-through inhalation exposures of LE rats to 200, 2000, or 4000 ppm TCE. Simulation of the experimental data was improved by use of F-344 blood-flow values and the corresponding Vmaxc (8.68 mg/h/kg) compared to use of SD flows and the associated Vmaxc (7.34 mg/h/kg). Sensitivity analysis was used to determine those input parameters with the greatest influence on TCE tissue concentrations. Alveolar ventilation consistently (across exposure concentration, exposure duration, and target tissue) had the greatest impact on TCE tissue concentration. The PBPK model described here is being used to explore the relationship between measures of internal dose of TCE and neurotoxic outcome.

Evaluation of physiologically based pharmacokinetic models for use in risk assessment

“…For example, if the sensitivity ratio for breathing rate is the highest of all input parameters, this indicates that it is the most sensitive model parameter for the dose metric. If, for example, the sensitivity ratio for breathing rate is 2, this signifies that a 1% change in the numerical value for breathing rate will result in a 2% change in the model output (Simmons et al, 2002). Sensitivity ratios greater than 1 (in absolute value) are of concern because this results in the amplification of input error (Allen et al, 1996).…”

“…Sensitivity ratios greater than 1 (in absolute value) are of concern because this results in the amplification of input error (Allen et al, 1996). In practice, PBPK models generally have normalized sensitivity coefficients ≤ |1.0|, except during transient phases (illustrated in Simmons et al, 2002), while biological modeling of pharmacodynamics may show much larger amplification effects (see examples in Allen et al, 1996). It is critical that fractional bloodflows sum to cardiac output when they are varied in sensitivity analyses or mass balance will be violated and normalized sensitivity ratios much larger than 1 may be obtained for blood flow parameters.…”

An Example of Model Structure Differences Using Sensitivity Analyses in Physiologically Based Pharmacokinetic Models of Trichloroethylene in Humans

“…Previous work has been done developing a PBPK model for TCE with a brain compartment in order to evaluate neurotoxicity data in the Long-Evans rat (Simmons et al, 2002). The incorporation of a brain compartment is important in humans since TCE can go across the blood brain barrier (BBB) and because TCE has a significant effect on the brain as a narcotic agent at higher concentrations.…”

A proposed methodology for selecting a trichloroethylene inhalation unit risk value for use in risk assessment

“…In addition, considerable information is known regarding potential mechanisms by which TCE, or more properly its metabolites, may cause liver tumors. Species-speciWc diVerences in TCE toxicokinetics have largely been worked out through modeling, albeit with notable uncertainty (Allen and Fisher, 1993;Barton and Clewell, 2000;Bois, 2000a,b;Clewell et al, 2000;Fisher et al, 1998;Keys et al, 2003;Simmons et al, 2002). Thus, the knowledge base supporting the liver tumor endpoint is fairly robust and the endpoint is assigned a score of "++.…”