Characterizing uncertainty and population variability in the toxicokinetics of trichloroethylene and metabolites in mice, rats, and humans using an updated database, physiologically based pharmacokinetic (PBPK) model, and Bayesian approach

Abstract

We have developed a comprehensive, Bayesian, PBPK model-based analysis of the population toxicokinetics of trichloroethylene (TCE) and its metabolites in mice, rats, and humans, considering a wider range of physiological, chemical, in vitro, and in vivo data than any previously published analysis of TCE. The toxicokinetics of the “population average,” its population variability, and their uncertainties are characterized in an approach that strives to be maximally transparent and objective. Estimates of experimental variability and uncertainty were also included in this analysis. The experimental database was expanded to include virtually all available in vivo toxicokinetic data, which permitted, in rats and humans, the specification of separate datasets for model calibration and evaluation. The total combination of these approaches and PBPK analysis provides substantial support for the model predictions. In addition, we feel confident that the approach employed also yields an accurate characterization of the uncertainty in metabolic pathways for which available data were sparse or relatively indirect, such as GSH conjugation and respiratory tract metabolism. Key conclusions from the model predictions include the following: (1) as expected, TCE is substantially metabolized, primarily by oxidation at doses below saturation; (2) GSH conjugation and subsequent bioactivation in humans appear to be 10- to 100-fold greater than previously estimated; and (3) mice had the greatest rate of respiratory tract oxidative metabolism as compared to rats and humans. In a situation such as TCE in which there is large database of studies coupled with complex toxicokinetics, the Bayesian approach provides a systematic method of simultaneously estimating model parameters and characterizing their uncertainty and variability. However, care needs to be taken in its implementation to ensure biological consistency, transparency, and objectivity.

Introduction

Trichloroethylene (TCE) is a volatile organic solvent that has had widespread commercial and industrial use, particularly for industrial vapor degreasing of metal parts and as a chemical intermediate. TCE is also a common environmental contaminant at hazardous waste sites, in groundwater, and in ambient and indoor air. Because it is a dense nonaqueous-phase liquid, TCE is particularly difficult to remediate once it has entered groundwater [ATSDR, 1997, ATSDR, 2003].

Understanding TCE toxicokinetics is critical to both the qualitative and quantitative assessment of human health risks from environmental exposures and continues to be the subject of active research (Boyes et al., 2005, Chiu et al., 2007). Understanding TCE metabolism is especially important toxicologically because specific metabolites or metabolic pathways are associated with a number of endpoints of observed toxicity. In mice, TCE induces pulmonary toxicity after acute exposure (e.g., Forkert et al., 1985, Green et al., 1997, Odum et al., 1992) and lung tumors after chronic exposure (Fukuda et al., 1983; Maltoni et al., 1996), with available evidence suggesting that these effects are mediated by pulmonary oxidative metabolism (Forkert et al., 2005, Forkert et al., 2006). TCE and its oxidative metabolites trichloroacetic acid (TCA) and dichloroacetic acid (DCA) cause a number of similar effects in the liver of laboratory animals, including hepatomegaly in multiple rodent species, and hepatocarcinogenicity in multiple strains and sexes of mice (Bull, 2000). Moreover, the hepatomegalic effects in mice are more consistently associated with oxidative metabolism than with TCE itself (Buben and O'Flaherty, 1985, Evans et al., 2009). In the kidney, TCE causes tubular toxicity in mice and rats and is associated with small increases in the incidences of kidney tumors reported in multiple strains of TCE-exposed rats (Maltoni et al., 1986, National Cancer Institute NCI), (1976, National Toxicology Program NTP), (1988, National Toxicology Program NTP), (1990). These effects are thought to be mediated by products of glutathione (GSH) conjugation (Lash et al., 2000b, Lash et al., 2001). Human epidemiologic data also suggest that the kidney and liver are targets of TCE non-cancer toxicity (NRC, 2006). For carcinogenicity, the epidemiologic evidence is strongest for kidney cancer (NRC, 2006), with several well-designed and sufficiently powered studies reporting statistically significant increases in renal cell carcinoma incidence in groups occupationally exposed to TCE (e.g., Brüning et al., 2003, Charbotel et al., 2006, Zhao et al., 2005, Raaschou-Nielsen et al., 2003).

A simplified metabolism scheme for TCE is shown in Fig. 1. Briefly, as reviewed by Lash et al., 2000a, Chiu et al., 2006, metabolism of TCE occurs through two main irreversible pathways: oxidation via the microsomal mixed-function oxidase system (i.e., cytochrome P450s) and conjugation with GSH by glutathione S-transferases. For TCE oxidation, the P450 isoform CYP2E1 is thought to be most important in vivo. TCE is first oxidized to several unstable intermediate products (metabolites 2, 3, and 5), some of which are quickly transformed to trichloroethanol (TCOH-metabolite 6) (a reversible reaction) and trichloroacetic acid (TCA-metabolite 7). TCOH is glucuronidated to form TCOH-glucuronide (TCOG) (metabolite 8), which undergoes enterohepatic recirculation (excretion in bile with regeneration and reabsorption of TCOH from the gut). Both TCA and TCOG are excreted in urine, but other metabolism of TCA and TCOH has not been well characterized, although DCA (metabolite 9) has been hypothesized to be among the metabolism products (Lash et al., 2000a). TCE-oxide (metabolite 3) may also form DCA, among other species (Cai and Guengerich 1999). With respect to the conjugative pathway, dichlorovinyl glutathione (DCVG) (metabolite 4) is further processed to form the cysteine conjugate S-dichlorovinyl-l-cysteine (DCVC) (metabolite 10), which can undergo either bioactivation by beta-lyase or flavin-containing monooxygenases (FMO) to reactive species (Anders et al., 1988, Krause et al., 2003, Lash et al., 2003) or (reversible) N-acetylation to the mercapturate N-acetyl dichlorovinyl cysteine (NAcDCVC) (metabolite 11) excreted in urine or sulfoxidated by CYP3A to reactive species (Bernauer et al., 1996, Birner et al., 1993, Werner et al., 1995).

TCE has an extensive number of both in vivo pharmacokinetic and PBPK modeling studies in mice, rats, and humans (see Chiu et al. 2006, and Supplementary Material therein, for a review). Hack et al. (2006) represents the most recent published PBPK model attempting to integrate this vast database, using the Bayesian population approach first applied to TCE by Bois, 2000a, Bois, 2000b). A key question, then, is why should we develop yet another PBPK model for TCE and its metabolites? Indeed, our original intention was not to develop a completely revised model but rather to perform a minor “update” to the Hack et al (2006) effort. However, in the process of conducting a detailed evaluation of the Hack et al. (2006) model, the number of issues that required adjustment grew beyond what could reasonably be considered an application of the original model. The main conclusions of this evaluation, and their implications for PBPK model development, are summarized in Supplementary Materials (Table S-1). A deterministic analysis motivated the updated respiratory tract model and is reported separately (Evans et al., 2009).

In this article, we attempt to address these issues in conducting a new Bayesian, PBPK model-based, population toxicokinetic analysis of the currently available database of in vitro and in vivo toxicokinetic data in mice, rats, and humans. This analysis includes a much larger database of studies than has ever been published in a Bayesian PBPK analysis. Particular attention is paid to several important issues: (i) a fuller accounting of biologically plausible sources of uncertainty and variability; (ii) a clear separation between data used for developing prior estimates for model parameters and data used to update these estimates and generate posterior distributions, to avoid using the same data “twice”; (iii) explicit delineation of the type of variability being characterized in the population analysis, i.e., interstudy for mice and rats versus interindividual for humans; and (iv) a systematic evaluation of model convergence, posterior parameter estimates, and comparisons of model predictions with data. The resulting model with its parameter distributions is then used to characterize the uncertainty and variability in a number of important toxicokinetic processes, and can subsequently be used in probabilistic dose–response analyses (in preparation).