By Sree Roy

A PhD student at the University of Antwerp in Belgium, Karlien Van den Bossche, MD, is working on predicting the outcomes of oral appliance therapy in patients with obstructive sleep apnea (OSA) as part of her graduate thesis. At the university’s hospital, drug-induced sleep endoscopy (DISE)—which investigates the characteristics of sleep apnea airway collapse while a patient is under sedation—is a common practice that informs therapeutic decisions. There, about 250 people with OSA undergo DISE each year. 

For mandibular-advancement-style oral appliances in particular, the clinicians regularly employ DISE with and without the use of a simulation bite to evaluate the effect of the devices on upper airway collapse. But research studies to predict treatment efficacy of oral appliances using DISE are typically based on “a subjective interpretation of the endoscopic video-footage, rather than exact quantitative measurements,” says Van den Bossche, who is also an ENT-head and neck surgery resident at the Antwerp University Hospital. “Hence, our interest grew to investigate this topic in more detail, by quantitatively measuring the effect of an MAD on upper airway dimensions during DISE.”

Results of their quantitative study, published in the July 2022 issue of the Journal of Dental Sleep Medicine, won a student excellence award at the 2022 American Academy of Dental Sleep Medicine Annual Meeting. The findings provide more insight into the mechanism of action of oral appliances on upper airway morphology during sleep. 

Drug-induced sleep endoscopy images from the study “Quantitative effect of mandibular advancement devices on upper airway dimensions during drug-induced sleep endoscopy in obstructive sleep apnea” showing baseline airway opening versus opening with an oral appliance in place.

A total of 336 images from DISE video footage—168 baseline and 168 with an oral appliance in place—in 56 OSA patients were scored. To digitally measure cross-sectional areas at both levels before and after mandibular advancement, the airway lumen was followed transversally using a polygon selection tool in an image processing program with consideration of light-dark interfaces to ensure a constant depth. 

The retroglossal cross-sectional area was defined by following the lumen according to the posterior boundary of the lingual tissue around the pharyngeal walls at a constant anatomical level. At a retro-epiglottic level, the cross-sectional area was determined in a transverse plane by following the posterior border of the epiglottis and continuing around the hypopharyngeal walls. Expansion ratios were calculated by dividing cross-sectional areas during mandibular advancement by areas at baseline. Treatment response was defined as reduction in AHI ≥50%.

The investigators saw a significant difference between retroglossal cross-sectional areas at baseline (47,823.36±2,357.58 pixels) and with the oral appliance (55,818.52±2,357.58 pixels) (p≤0.0001). “This is in line with previous research,” Van den Bossche says. “Interestingly, after adjusting the MAD [mandibular advance device]-treatment response definition for body position during sleep, our study shows that patients who have a higher overall retroglossal upper airway expansion are more likely to be treated successfully with a MAD.”

They saw no significant difference at the retro-epiglottic area (p=0.1074), which Van den Bossche says is in line with previous studies indicating that the occurrence of epiglottic collapse during DISE does not affect oral appliance outcome.

The study population was predominantly supine-dependent (60.7%). “It is important to mention that body position during the night may play a role while comparing OSA severity,” Van den Bossche says. “To overcome this problem, a correction according to body position is applied in our additional analyses to evaluate MAD treatment response.”

Greater expansion ratios for retroglossal area were seen in responders (1.31±0.49) compared to non-responders (1.12±0.31), although non-significant (p=0.0876). No significant association was found between treatment response and retro-epiglottic area.

Van den Bossche says, “The current research highlights the additional value of exact quantitative measurements of the upper airway during DISE in predicting treatment outcome with an MAD. Moreover, the utility of performing the DISE-procedure in evaluating MAD treatment response in clinical practice is emphasized.”

Sree Roy is editor of Sleep Review.


Van den Bossche K, Van de Perck E, Vroegop AV, et al. Quantitative effect of mandibular advancement devices on upper airway dimensions during drug-induced sleep endoscopy in obstructive sleep apnea. JDSM. 10 Jul 2022;9(3):31-2 (Abstract #015).

Illustration 51794020 © Juliatimchenko |