Integrated dosimetry, toxicokinetic, and toxicodynamic modeling of inhaled cerium oxide nanoparticles for human risk characterization.

Cerium oxide (CeO2) nanoparticles are widely used in industrial and commercial applications, resulting in increasing concerns regarding potential inhalation-related health risks. However, quantitative translation of inhalation exposure into human risk metrics remains limited due to the lack of integrated mechanistic frameworks linking respiratory deposition, toxicokinetics, toxicodynamics (TD), and risk assessment. The present study developed an integrated multiple-path-particle-dosimetry (MPPD)-physiologically-based-toxicokinetic (PBTK)-TD modeling framework for CeO2 nanoparticles inhalation and applied it to model-based human risk characterization. A rat MPPD-PBTK model was first established using published inhalation biodistribution data and successfully reproduced CeO2 concentrations in lung, serum, liver, kidney, spleen, gastrointestinal tract, urine, and feces (97.26% and 98.63% of observations within 2-fold and 3-fold-error, respectively). The model was subsequently extrapolated to humans by incorporating human-specific respiratory and physiological parameters. A reduced-order TD model describing reactive-oxygen-species generation, glutathione depletion, caspase-3 activation, and cell viability reduction was developed using published BEAS-2B toxicity data (100% of observations within 2-fold-error). Sensitivity analyses identified exposure concentration, pulmonary deposition fraction, pulmonary effective-exposure kinetics, and viability-related parameters as major determinants of model outputs. Monte-Carlo uncertainty propagation revealed substantial dispersion in pulmonary effective concentration and viability responses arising from combined plausible physiological variability and parameter/model uncertainty. Within the BEAS-2B-based TD framework, viability-based criteria yielded lower model-derived exposure thresholds than the selected ROS- or caspase-3-based criteria. Among the evaluated exploratory cellular-response criteria, the lowest model-derived risk-based human-equivalent concentration was obtained for viability < 95% at 180 h, yielding 0.0138 mg/m3 for the 24-h exposure scenario at a 10% model-based threshold-exceedance level when calculated using direct concentration-dependent reverse dosimetry.