Effectiveness of Hypoglossal Nerve Stimulation Therapy in Positional Obstructive Sleep Apnea: A Retrospective...

Background: Obstructive sleep apnea (OSA) is a prevalent disorder characterized by airway collapse during sleep. Continuous positive airway pressure (CPAP) is first-line treatment but adherence can decay over time due to intolerance. Hypoglossal nerve stimulation (HNS) has emerged as an alternative, especially for CPAP-intolerant patients. OSA can be classified into position-dependent (PD-OSA) and non-position-dependent (NPD-OSA) subtypes based on apnea-hypopnea index (AHI) variation by sleep posture. Study Objectives: This study aims to evaluate polysomnographic changes following HNS therapy and compare treatment outcomes in PD-OSA and NPD-OSA patients. Methods: A retrospective observational study of 30 patients treated with HNS at a single center between January 2022 and March 2025 was conducted. The primary endpoint was change in overall apnea-hypopnea index (AHI) from baseline to first post-implant in-laboratory polysomnography (PSG). Secondary endpoints included changes in phenotype-specific (supine and non-supine) AHI, Epworth Sleepiness Scale (ESS), and sleep architecture parameters. Subgroup comparisons were performed between PD-OSA and NPD-OSA phenotypes. Results: Thirty patients (median age 69.5 years; 73% male; median BMI 28.9 kg/m²) were included; 27 had sufficient positional data for phenotype classification (66.7% PD-OSA, 33.3% NPD-OSA). Median AHI decreased from 23.5 to 4.8 events/h (p < 0.0001), with reductions in both REM and supine AHI. PD-OSA patients demonstrated the greatest improvement in supine AHI, whereas NPD-OSA patients uniquely improved in non-supine AHI. ESS decreased by a median of 1.5 points overall (p = 0.0015) and met the minimal clinically important difference in NPD-OSA. Sleep architecture showed minimal change, except for a reduction in supine sleep percentage (p = 0.0114). Conclusions: HNS therapy improved AHI and subjective sleepiness across OSA phenotypes, with distinct positional responses. These findings support the clinical utility of HNS in both PD-OSA and NPD-OSA and suggest phenotype-specific treatment effects warrant further investigation.