Quantitative airway analysis during drug-induced sleep endoscopy for evaluation of sleep apnea

Objectives/hypothesis: To quantitatively measure changes in airway caliber at multiple anatomical levels during drug-induced sleep endoscopy (DISE) for evaluation of sleep apnea. We hypothesize that patients undergoing DISE will show: 1) collapse at multiple upper airway regions (retropalatal, retroglossal, and retroepiglottic), with greater collapse in the retropalatal region; and 2) greater anterior-posterior dimensional narrowing than the lateral.

Study design: Case series.

Methods: Patients underwent DISE employing propofol as part of a nonrandomized prospective trial assessing candidacy for transoral robotic surgery intervention for sleep apnea. Images of the retropalatal, retroglossal, and retroepiglottic regions were captured during an initial period of light sedation and again in a period of deep sedation. Images were analyzed using software to measure the percent change in regional airway measurements as a result of DISE.

Results: Thirty-seven sleep endoscopy videos were analyzed from patients with obstructive sleep apnea (apnea-hypopnea index: 42.9 ± 27.0 events/hour). Analyzable images were in the retropalatal (n = 24), retroglossal (n = 27), and retroepiglottic (n = 29) regions. The patients demonstrated mean reductions in airway area in the retropalatal (84.1 ± 18.7%), retroglossal (39.3 ± 37.5%), and retroepiglottic region (44.6 ± 42.8%). No statistically significant differences were found between lateral and anterior-posterior airway dimensional changes.

Conclusions: Patients undergoing DISE had significant reductions in airway area at multiple regions under deep sedation with propofol. We conclude that collapse in the retropalatal region is greater than the hypopharyngeal region. This method can be used to quantitatively measure DISE upper airway changes, which could potentially be used as a means for understanding surgical outcomes in patients with sleep apnea.