Feasibility of inline respimat soft mist inhaler delivery via high-flow nasal cannula: In vitro study of adapter design and flow rate

High-flow nasal cannula (HFNC) delivers heated, humidified oxygen at high flow rates and is widely used as a respiratory support. However, effective delivery of inhaled drugs during HFNC remains a challenge, with current options limited to nebulizers and pressurized metered-dose inhalers (pMDIs). Respimat soft mist inhalers (SMIs), which generate slow-moving, respirable aerosols, and are less dependent on patient coordination, have not been adapted for HFNC due to the absence of suitable inline interfaces. This study aimed to develop and evaluate a novel inline adapter to enable SMI aerosol delivery during HFNC therapy. Four inline adapter prototypes were designed and fabricated using 3D printing. Computational fluid dynamics (CFD) simulations were employed to evaluate flow distribution within each design. In vitro assessments were conducted to determine the emitted dose and aerosol droplet size distribution across HFNC flow rates ranging from 5 to 60 L/min. Based on these results, the most efficient adapter was selected for further evaluation using an in vitro system comprising a breathing simulator and an adult nasal airway replica (NAR). This setup was used to investigate inhaled dose (drug collected throughout the in vitro system) and regional deposition profiles across the same range of HFNC flow rates. Among the four adapter designs, Adapter 2 consistently demonstrated the highest emitted dose and favorable droplet size across different flow rates. In vitro testing showed that the emitted dose from the adapter increased with flow rate from 44.1 ± 1.9 % at 5 L/min to 58.6 ± 1.7 % at 10 L/min, and then remained stable up to 60 L/min (60.0 ± 4.3 %). However, the fine particle fraction was 37.5 ± 6.0 % at 5 L/min and 37.8 ± 3.2 % at 10 L/min, followed by a progressive decline to 21.6 ± 4.0 % at 20 L/min, 16.9 ± 2.7 % at 40 L/min, and 16.1 ± 2.3 % at 60 L/min (p < 0.05). In parallel, the volumetric median diameter increased from 6.46 ± 0.99 μm at 5 L/min and 6.36 ± 0.50 μm at 10 L/min to 13.50 ± 4.22 μm, 17.49 ± 2.06 μm, and 24.27 ± 0.99 μm at 20, 40, and 60 L/min, respectively (p < 0.05). Further evaluation using an in vitro system revealed a clear inverse relationship between HFNC flow rate and inhaled dose. The inhaled dose peaked at 0.74 ± 0.14 μg (29.6 % of the 2.5 μg label dose per actuation of tiotropium) at 10 L/min. Upper airway deposition remained relatively constant (∼0.14 μg, approximately 5.6 % of the label dose) across all flow rates, while lung dose increased significantly at lower flows. These findings suggest that combining optimized adapter geometry with reduced HFNC flow may offer a practical strategy to enhance the delivery efficiency of SMI aerosols.