Development of a physiologically-based pharmacokinetic model for jet fuels in the rat

Abstract

The pharmacokinetic behavior of the majority of jet fuel constituents has not been previously described in the framework of a physiological-based pharmacokinetic (PBPK) model for inhalation exposure. Exposure-related effects have been reported in multiple organ systems, though exposure methods were different across studies, utilizing either vaporized or aerosolized fuels. Uncertainties in the kinetics associated with both exposure types previously represented a barrier to model development and interpretation of toxicity data. While assessments of pharmacokinetics and metabolic interactions have been conducted for binary to quaternary hydrocarbon mixtures, few models exist for higher-order fuel or solvent mixtures. The purpose of this work was first to elucidate the characteristics of the dynamic chamber environment and assess the effect on pharmacokinetic behavior of aerosolized jet fuels, in comparison to vaporized fuels. Using this information, individual PBPK models were developed to assess chemical behavior and then combined into the first PBPK model for petroleum-based and synthetic jet fuels. The resulting models were capable of predicting individual chemical and fuel kinetic behavior following both vaporized and aerosolized chemical exposures. To support model development, exposures to individual n-alkanes n-octane and n-tetradecane were conducted at 89 mg/m3 aerosol+vapor and 100-5000ppm vapor, respectively. Exposures to JP-8 and S-8 were conducted at ~900-1000 mg/m3, and to a 50:50 blend of both fuels at ~200 mg/m3 aerosol+vapor. A novel computational description of the respiratory tract was developed, with concentrations directed to either gas-exchange or respiratory tract tissue compartments, describing vapor and aerosol uptake respectively. Visceral tissue compartments were described using perfusion and diffusion-limited equations connected by blood. The model described the kinetics of individual chemicals and fuel constituents at multiple aerosol and vapor concentrations, utilizing a chemical “lumping” strategy to estimate parameters for unspeciated fuel fractions. The model more accurately simulated data for aromatic and lower molecular weight (MW) n-alkanes than for some higher MW chemicals. Metabolic interactions were more pronounced at high total fuel concentrations (~2700 - 1000 mg/m3) than at low concentrations (400 - 200 mg/m3). This model serves as the most detailed assessment of fuel pharmacokinetics to date.