(1305) Battery Life and Predictors of Depletion in Hypoglossal Nerve Stimulation for Obstructive Sleep Apnea

Introduction: Hypoglossal nerve stimulation (HGNS) is an effective surgical therapy for obstructive sleep apnea (OSA), yet limited data exist regarding real-world implantable pulse generator (IPG) battery longevity and factors associated with earlier depletion. Improved understanding of battery life may enhance patient counseling and long-term device management.

Methods: Patients with HGNS implants who experienced IPG battery depletion were retrospectively identified. Collected variables included demographics, preoperative apnea–hypopnea index (AHI), device utilization, stimulation amplitude, optimal titration voltage, and Sher response (≥50% AHI reduction and postoperative AHI < 20). Battery life, defined as days from implantation to IPG battery depletion, was analyzed using Mann–Whitney U tests and Spearman correlation coefficients.

Results: Ten patients (mean age 58.3 years, BMI 31.0 kg/m2, 80% male) with moderate preoperative OSA (mean AHI 29.88 events/hr) were included. The mean battery life was 2887.1 ± 185.72 days (7.91 ± 0.51 years). Battery life showed a weak-to-moderate positive trend with pre-implant age (r=0.243, p=0.498) and moderate negative trends with pre-implant BMI (r=-0.479, p = 0.166) and AHI (r=-0.394, p=0.263). Post-operatively, battery life trended negatively with average utilization (r = -0.233, p=0.552), set amplitude (r=-0.28, p=0.432), and titration study optimal voltage (r=-0.188, p=0.604). Battery life did not differ significantly between Sher responders and non-responders (2986.5 ± 126.57 days, N=2 vs 2935 ± 122.34 days, N=4; p=0.817).

Conclusions: In this pilot cohort, HGNS battery life averaged approximately eight years, providing independent institutional validation of manufacturer estimates and prior trial-based data. Trends suggest that younger patients, greater disease severity, and higher stimulation requirements may be associated with earlier battery depletion. Further validation in larger cohorts is warranted, as elucidating these predictors may help guide expectations for device longevity and future replacement planning.