Efficacy of Upper Airway Stimulation on Collapse Patterns Observed during Drug-Induced Sedation Endoscopy.

OBJECTIVE To describe upper airway collapse patterns observed on drug-induced sedation endoscopy (DISE) during screening for a clinical trial and to evaluate the impact of collapse patterns found on preoperative DISE on response rates to upper airway stimulation (UAS) therapy. STUDY DESIGN Retrospective review of an ongoing prospective multi-institutional cohort study. SETTING Twenty-two participating institutions of the STAR trial. SUBJECTS AND METHOD In total, 222 subjects were screened with DISE to determine eligibility for an implantable UAS device. Supine laryngoscopy was performed during moderate sedation (propofol and/or midazolam). Airway collapse pattern and severity were graded at 4 levels, including velum, oropharynx, tongue base, and epiglottis (VOTE classification). Patients with complete concentric collapse (CCC) at the velum were excluded from implantation. RESULTS The CCC at the velum was observed in 52 (23%) of screened subjects, and these subjects were subsequently excluded from implantation. Of the 170 subjects without CCC at the velum, 126 (77%) underwent implantation: 121 (96%) had multilevel collapse and 5 (4%) had single-level collapse. When comparing preimplantation DISE findings, UAS responders at 12 months had lower baseline VOTE scores compared with therapy nonresponders. CONCLUSION Drug-induced sedation endoscopy is an efficient and safe method for determining UAS eligibility and has the potential to identify UAS nonresponders. Most patients had multilevel airway collapse, illustrating the limitations of single-level upper airway surgery in treating obstructive sleep apnea. Upper airway stimulation is effective therapy for most patients with multilevel airway collapse; however, patients with complete anterior-posterior or lateral soft palate and/or epiglottic collapse may be at increased risk of therapy failure.