Methodological issues of using observational human data in lung dosimetry models for particulates

Area of Science:

• Occupational Health
• Toxicology
• Biomathematics

Background:

• Physiologically based pharmacokinetic (PBPK) models require human data for calibration and validation.
• Human data can be sparse and heterogeneous, necessitating specialized modeling approaches.
• Evaluating variability and uncertainty in human lung dosimetry is critical for accurate risk assessment.

Purpose of the Study:

• To develop and validate a human lung dosimetry model using data from U.S. coal miners.
• To investigate sources of variability and uncertainty in lung particle dosimetry.
• To predict inter-individual differences in lung particle burdens based on exposure.

Main Methods:

• A multivariate optimization procedure fitted a dosimetry model to data from 131 U.S. coal miners.
• Model structure uncertainty was assessed by fitting various forms for particle clearance and sequestration.
• Sensitivity analysis identified key model parameters influencing output, and clearance parameters were estimated for individuals.

Main Results:

• The best-fit model incorporated first-order interstitialization and no dose-dependent decline in alveolar clearance.
• Fractional deposition was the most influential parameter; race and fibrosis severity predicted alveolar clearance rates.
• Slower estimated clearance correlated with higher observed lung burdens, indicating significant inter-individual variability.

Conclusions:

• The developed methods addressed uncertainty in model structure and variability in estimated clearance parameters.
• Individual clearance rates substantially influence predicted lung burden and potential disease risk.
• Findings aid risk assessment by estimating lung burden distributions under various exposure scenarios.