Open AccessEndotypic Mechanisms of Successful Hypoglossal Nerve Stimulation for Obstructive Sleep Apnea
Author(s) -
Sara Op de Beeck,
Andrew Wellman,
Marijke Dieltjens,
Kingman P. Strohl,
Marc Willemen,
Paul H. Van de Heyning,
Johan Verbraecken,
Olivier M. Vanderveken,
Scott A. Sands,
Star trial investigators
Publication year - 2021
Publication title -
american journal of respiratory and critical care medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.272
H-Index - 374
eISSN - 1535-4970
pISSN - 1073-449X
DOI - 10.1164/rccm.202006-2176oc
Subject(s) - medicine , polysomnography , obstructive sleep apnea , odds ratio , anesthesia , confidence interval , apnea , arousal , sleep apnea , logistic regression , physical therapy , psychology , neuroscience
Rationale: Approximately one-third of patients with obstructive sleep apnea (OSA) treated with hypoglossal nerve stimulation (HGNS) therapy are incomplete responders, despite careful patient selection based on baseline characteristics and drug-induced sleep endoscopy. Objectives: Here we use polysomnographic endotyping to assess the pathophysiological mechanisms underlying favorable versus incomplete responses to HGNS therapy. Methods: Baseline polysomnography data of the STAR (Stimulation Therapy for Apnea Reduction) trial were included. Raw baseline polysomnographic data from 91/126 patients were available for analysis. Traits-loop gain, arousal threshold, collapsibility, and muscle compensation-were calculated from the baseline polysomnography data according to Sands and colleagues (AJRCCM 2018, SLEEP 2018). Logistic regression assessed apnea-hypopnea index (AHI)-adjusted associations between HGNS response (>50% reduction in AHI to <10/h at 1 yr) and OSA traits. Measurements and Main Results: Overall, HGNS treatment reduced AHI from 30.7 (24.9-39.9) to 8.5 (4.0-19.5) events/h ( P < 0.0001; median [quartiles 1-3]); N = 53/91 were responders. In adjusted analysis, a favorable response to therapy was independently associated with higher arousal threshold (odds ratio [95% confidence interval]: 6.76 [2.44-23.3], P = 0.001), greater compensation (odds ratio: 4.22 [1.70-12.55] per SD, P = 0.004), and lower loop gain (in milder collapsibility, per significant interaction, P = 0.003). The higher arousal threshold was evident in responders before adjusted analysis. Predicted responders had an approximately fourfold lower treatment AHI versus predicted nonresponders (4.9 [2.7-8.5] vs. 20.7 [10.9-29.7], P < 0.0001; median [quartiles 1-3]); differences remained significant after cross-validation. Conclusions: Favorable responses to HGNS therapy are associated with the pathophysiological traits causing OSA, particularly a higher arousal threshold. Along with established criteria, individuals with favorable traits could potentially be prioritized for precision HGNS therapy.This analysis was a secondary analysis of the STAR trial registered with clinicaltrials.gov (NCT01161420).