HSC Section 3 - Trauma, Critical Care and Sleep Medicine

be related to intermittent hypoxia, systemic inflamma- tion, and vibrational trauma from snoring. Therefore, aug- menting the efferent loop of this reflex with direct stimulation of the genioglossus muscle during sleep reduces airway instability and collapsibility. The intention of this article is to review the current state of the science of a novel therapy for OSA, the Inspire upper airway stim- ulation system (Inspire Medical Systems, Inc., Maple Grove, MN), and to present pertinent new observations since the Food and Drug Administration’s approval of this therapy in 2014. The institutions of the authors of this review have received research support ( R . S ., M . B . G .), and the authors have received a consulting fee from the manu- facturer of the upper airway stimulation device. UPPER AIRWAY STIMULATION OVERVIEW Initial Evidence Preclinical studies of OSA in animal models showed that electrical stimulation of the hypoglossal nerve pro- duced significant increases in airflow and airway stabili- ty. 29–31 The first human trials confirmed that hypoglossal nerve stimulation improves airflow and upper airway stability, without causing arousal from sleep or adverse neuromuscular side effects. 32,33 Naso- pharyngoscopy demonstrated that genioglossus stimula- tion resulted in enlargement and stabilization of not only the retrolingual airway but also the retropalatal space. Enlargement of the retropalatal airway is thought to be due to a direct coupling of the palate to the tongue base via the palatoglossus muscle. In addition, forward motion of the tongue dorsum reduces contact with the soft palate and allows the velum to drop away from the posterior pharyngeal wall. As OSA frequently involves multilevel collapsibility, this was a key observation, sug- gesting that hypoglossal nerve stimulation had the potential to impact multiple levels of the pharyngeal air- way rather than the tongue base alone. Human Pilot and Feasibility Studies Following the published technical limitations of early-generation implants, multiple investigators spent the next decade improving the components prior to the launching two larger trials with results available as of 2011. 34,35 A limited sample–size pilot study of the Inspire upper airway stimulation system that served as the basis for the larger phase III trial found that eight (42%) of 19 implanted patients with sufficient follow-up met criteria for success, with at least a 50% reduction in apnea-hypopnea index (AHI) and a final overall AHI < 20. 35 When these eight responders were compared to the 11 nonresponders, it was found that an AHI < 50, body mass index (BMI) 32 kg/m 2 , and absence of com- plete concentric pattern of collapse at the palate viewed during drug-induced sleep endoscopy (DISE) were pre- dictive of success of hypoglossal nerve stimulation. 35 Complete concentric collapse at the palate is likely due to increased parapharyngeal fat and creates a degree of collapsibility in the retropalatal airspace that is difficult to overcome with upper airway stimulation alone. Eight

additional participants were then enrolled prospectively using the above selection criteria and demonstrated sig- nificant AHI reduction from 38.9 6 9.8 to 10.0 6 11.0 ( P < .01) at the 6-month postimplant visit. 35 The proof of concept that targeted hypoglossal nerve stimulation can lower AHI was further validated in a separate phase II study of a different hypoglossal nerve stimulator (ImThera Medical, San Diego, CA). 36 In that study, 43 of 46 (93%) subjects demonstrated a significant reduction in AHI at 6 months, with 15 patients (33%) meeting criteria as AHI responders ( 50% decrease in AHI with overall AHI 20). When comparing AHI responders to nonresponders, it was found that responders were more likely to have a base- line BMI 35, a baseline AHI 65 and apnea index 30, and 15 events per hour where the oxygen saturation decreased > 10%. These responder variables are current- ly being applied as inclusion criteria for a phase III trial of the ImThera device. The STAR Trial The basic science experiments and feasibility stud- ies served as the basis for the multicenter prospective Stimulation Therapy for Apnea Reduction (STAR) trial. 37 Eligible implant participants had moderate–severe OSA (AHI 20–50), CPAP intolerance, BMI 32, and absence of a complete circumferential pattern of palatal obstruc- tion on DISE. After a rigorous clinical, polysomno- graphic, and DISE screening, the study met enrollment goals with 126 participants undergoing surgical implan- tation of the hypoglossal nerve stimulation system, fol- lowed by a 12-month assessment for effectiveness and adverse events. Devices were titrated in the sleep labo- ratory during full-montage attended polysomnography, similar to CPAP titration, to optimize comfort and effec- tiveness. Primary outcome measures (AHI, 4% oxygen desaturation index [ODI]) and secondary outcomes mea- sures (Epworth Sleepiness Scale [ESS], Functional Out- comes of Sleep Questionnaire [FOSQ]) all demonstrated clinically and statistically significant improvements at 12 months (median AHI reduction from 29.3 to 9.0, and median ODI reduction from 25.4 to 7.4). Two-thirds of the implanted participants were considered successful responders to therapy by previously published surgical success criteria (AHI decrease 50% and overall AHI < 20). Quality-of-life measures also improved significant- ly across the cohort, with ESS reduced from median 11.0 to 6.0, and FOSQ increased from 14.6 to 18.2 at the 12- month follow-up. Risk and morbidity data were favorable, with no permanent hypoglossal nerve weakness, no serious device-related infection requiring explantation, and sig- nificantly less postoperative discomfort (mean pain level 3/10 on visual analog scale [VAS] on postoperative day [POD] 1) compared to historical cohorts of traditional pharyngeal or skeletal sleep apnea surgeries (mean pain level 8/10 on VAS on POD 1). 38 One-third of the partici- pants reported intermittent and self-limited tongue dis- comfort due to stimulation itself or abrasion of the tongue on an adjacent tooth. Of this group of patients,

Laryngoscope 126: September 2016

Soose and Gillespie: Upper Airway Stimulation Therapy

174