Effects of Hypoglossal Nerve Stimulation on Upper Airway Structure and Function Using Moving Wall Computational Fluid Dynamics Simulations: A Pilot Study

• DOI: 10.1111/jsr.70040
• Publication Date: 2025-03-18T08:34:07Z
• Authors: Qiwei Xiao; Daniel Ignatiuk; Chamindu Gunatilaka; Keith McConnell; Christine Schuler; Ann Romaker; Stacey Ishman; Robert Fleck; Raouf Amin; Alister Bates

ABSTRACTHypoglossal nerve stimulation (HGNS) is an innovative alternative treatment option for obstructive sleep apnoea (OSA) in patients unable to tolerate continuous positive airway pressure. However, the success rate of HGNS is variable, but the reasons underlying variation in treatment efficacy are not well understood. In this pilot study of three male subjects, we propose an innovative, non‐invasive method to quantify the structural and functional changes to the upper airway that occur with HGNS. We used four‐dimensional computed tomography (4DCT) and computational fluid dynamics (CFD) simulations of respiratory airflow to quantify how HGNS changes: (1) airway cross‐sectional area (CSA), (2) work done by muscles and air pressure in dilating and collapsing the airway and (3) airway resistance. Subjects underwent 4DCT under natural stage non‐REM 2 (N2) sleep with and without HGNS. Each patient had concurrent electroencephalograms and airflow measurements. CFD simulations were performed based on anatomy and airway motion from 4DCT images and airflow data. HGNS was associated with an increase in neuromuscular work done in dilating the airway (up to 490%); airway CSA increased by up to 300%. Most motion with HGNS occurred in the oropharynx; changes in the nasopharynx and hypopharynx varied between subjects. Minute ventilation increased in all subjects (15%–36%). Airway resistance decreased across the three subjects (73%–97%). Quantifying the parameters measured in this study may help explain variable responses to HGNS as a treatment for OSA. These procedures may, in future, help predict non‐responders to HGNS, isolate reasons for poor responses, or inform device titration.