Dosimetric adjustments for interspecies extrapolation of inhaled poorly soluble particles (PSP)

Direct calculation of delivered dose in the species of interest potentially affects the magnitude of an uncertainty factor needed to address extrapolation of laboratory animal data to equivalent human exposure scenarios, thereby improving the accuracy of human health risk estimates. Development of an inhalation reference concentration (RfC) typically involves extrapolation of an effect level observed in a laboratory animal exposure study to a level of exposure in humans that is not expected to result in an appreciable health risk. The default dose metric used for respiratory effects is the average deposited dose normalized by regional surface area. However, the most relevant dose metric is generally one that is most closely associated with the mode of action leading to the response. Critical factors in determining the best dose metric to characterize the dose-response relationship include the following: the nature of the biological response being examined; the magnitude, duration, and frequency of the intended exposure scenario; and the mechanisms by which the toxicants exert their effects. Dosimetry models provide mechanistic descriptions of these critical factors and can compute species-specific dose metrics. In this article, various dose metrics are postulated based on potential modes of action for poorly soluble particles (PSP). Dosimetry models are used to extrapolate the internal dose metric across species and to estimate the human equivalent concentration (HEC). Dosimetry models for the lower respiratory tract (LRT) of humans and rats are used to calculate deposition and retention using the principle of particle mass balance in the lower respiratory tract. Realistic asymmetric lung geometries using detailed morphometric measurements of the tracheobronchial (TB) airways in rats and humans are employed in model calculations. Various dose metrics are considered for the TB and pulmonary (P) regions. Because time is an explicit parameter incorporated in species-specific constants such as mucociliary clearance rates used in the models, the impact of the application of optimal model structures to refine adjustments and assumptions used in default risk assessment approaches to address exposure duration are discussed. HEC estimates were found for particles ranging in sizes that corresponded to existing toxicity studies of PSP (0.3 to 5 microm). A dose metric expressed as number of particles per biologically motivated normalization factors (e.g., number of ventilatory units, number of alveoli, and number of macrophages) was lower than the current default of mass normalized to regional surface area for either deposited or retained dose estimates. Retained dose estimates were lower than deposited dose estimates across all particle sizes evaluated. Dose metrics based on the deposited mass per unit area in small and large airways of the TB region indicate HECs of 1 to 5 times those of rats: that is, an equivalent exposure to humans which would achieve the same internal dose as in the rat would be 1 to 5 times greater. HEC estimates in the TB region increase with an increase in particle size for particles from 0.3 to 2 microm in the small airways and >3 microm in the large airways. The HEC decreases with increase in particle size in the P region across all particle sizes studied, and the decrease has a more significant slope for those particles >2 microm due to the limited inhalability of particles this size in rats relative to humans. Our modeling results elucidate a number of important issues to be considered in assessing current default approaches to dosimetry adjustment for inhaled PSP. Simulation of realistic, polydisperse particle distributions for the human exposure scenario results in reduced HEC estimates compared to estimates derived with the experimental particle distribution used in the laboratory animal study. Consideration should be given also to replacing the default dose metric of normalized deposited dose in the P region with normalized retained dose. Chronic effects are more likely due to retained dose and estimates calculated using retained versus deposited mass are shown to be lower across all particle sizes. Because dose metrics based on normalized particle number rather than normalized mass result in lower HEC estimates, use of inhaled mass as the default should also be revisited, if the pathogenesis suggests particle number determines the mode of action. Based on demonstrated age differences, future work should pursue the construction of "lifetime" estimates calculated by sequentially appending simulations for each specific age span.