Development of an infant complete-airway in vitro model for evaluating aerosol deposition.

A complete-airway in vitro model would be very useful for toxicological dosimetry testing and for developing targeted inhaled medications in cases where conducting in vivo experiments are exceedingly difficult, as with infants. The objective of this study was to determine whether packed bed in vitro models, which contain spheres as the primary repeating unit, provide a realistic representation of aerosol deposition in the tracheobronchial region of infant lungs based on computational fluid dynamics (CFD) predictions. The packed bed (PB) CFD model contained an inlet consistent with airway bifurcation B3 (∼lobar bronchi) leading to a spherical array with voids between the spheres forming a divided flow pathway. The hydrodynamic diameter of the voids was approximately matched to the diameter of bifurcations in various lung regions. For comparison, a CFD stochastic individual pathway (SIP) geometry with realistic bifurcations extending from B4-B15 (terminal bronchioles) was selected as an anatomically accurate model. The CFD-SIP model predictions were benchmarked with existing algebraic correlations for aerosol deposition in the lungs and found to be reasonable. Unfortunately, the CFD-PB model did not provide a good representation of aerosol deposition in the tracheobronchial region of human lungs. Through careful selection of the PB sphere size and inlet conditions, total deposition in the CFD-PB model matched CFD-SIP deposition within 10% absolute error across a range of relevant aerosol sizes. However, regional deposition within the CFD-PB model was very different from the CFD-SIP case. Therefore, the PB approach cannot be recommended for determining spatial or temporal distribution of aerosol transport and impaction deposition through the lungs.