Effect of airway wall motion on particle deposition and delivery in the neonatal trachea

Abstract

Modeling pulmonary drug delivery in the airway using computational fluid dynamics (CFD) simulations tracks drug particles throughout the airway, providing valuable information on the deposition location of inhaled drugs. However, most studies simulate particle transport within static airway models that do not incorporate physiological airway motion; this choice limits accuracy since airway motion directly affects particle transport and deposition, notably in newborns with airway abnormalities such as tracheomalacia. The objective of this study is to determine the effect of airway motion on drug delivery in neonates with and without airway disease. For this study, two control subjects without any airway disease and three subjects with tracheomalacia (dynamic tracheal narrowing) were enrolled. Each subject was imaged at approximately 40-weeks post-menstrual age using magnetic resonance imaging (MRI). MRI data were retrospectively reconstructed to obtain static airway images gated to different time points of the breath (i.e., end expiration and end inspiration) and an image representing combined data from all timepoints (ungated). Virtual airway surfaces (pharynx to main bronchi) were made from each MR image. A moving airway surface was created from surface registration of these surfaces and used as the boundary for a CFD simulation of one inhalation, along with subject-specific inspiratory flow waveforms. To assess the effect of airway wall motion on particle deposition, static-walled simulations, based on the airway surfaces at end inspiration, end expiration, and the ungated airway surface, were also performed using the same flow boundary conditions. Particle transport (particles diameter range 0.5–15 μm) was compared between the simulations during the inhalation. Airway surface motion affected particle transport into the small airways by 65% on average (0.5–5 μm– 22%, 5-15 μm– 86%) compared to static-walled simulations, while comparison between static end expiration and other static-walled simulations using geometries acquired during different phases of breathing differed by more than 500% on average (0.5–5 μm– 45%, 5-15 μm– 741%). For particle deposition, airway surface motion affected by 43% on average (0.5–5 μm– 86%, 5-15 μm– 21%) compared to static-walled simulations and comparison between static end expiration and other static-walled simulations differed by 47% on average (0.5–5 μm– 58%, 5-15 μm– 41%). Differences between dynamic and static deposition results and between static simulations from different timepoints occurred in patients with and without airway disease. This study suggests the importance of using airway wall motion in CFD simulations to model aerosolized drug delivery in the airway. If a CFD simulation is limited to only a static airway image without physiological motion, particle deposition mapping may yield markedly inaccurate results, potentially resulting in higher or lower drug dosing than intended.