How mandibular advancement splints alter genioglossus airway dilation patterns during inspiration in awake people with obstructive sleep apnoea may help predict treatment outcomes

Introduction: The efficacy of mandibular advancement splints (MAS) in obstructive sleep apnoea (OSA) is variable between patients, and this cannot be predicted reliably. Using tagged magnetic resonance imaging (MRI), we have previously observed that narrowing of the upper airway was associated with greater dilatory movement of the genioglossus during inspiration in people without OSA, and that the dilatory pattern was related to OSA status. Because mandibular advancement enlarges the airway by holding the mandible forward, we hypothesised MAS would alter how the genioglossus dilates the airway during wakefulness, and that motion may be related to MAS treatment outcome. Materials and methods: 87 untreated OSA participants (20 women, apnoea hypopnoea index (AHI) 7-102 events/hr, body mass index (BMI) 18-51 kg/m2, aged 19-76 years) underwent a MRI scan (Achieva 3TX, Philips) wearing a MAS. Mid-sagittal tagged MRI images were collected to quantify tongue dilatory movement with the jaw in neutral position (baseline) or advanced at 70% of the maximum, using harmonic phase methods. Imaging parameters: TR/TE = 400/16 ms, FOV = 220x196 mm, slice thickness = 10 mm, in-plane spatial resolution = 0.86x0.86 mm, tag spacing = 8.6 mm. Genioglossus dilation pattern was quantified over 3 inspirations using the antero-posterior movement of the back of the tongue. Treatment outcome was determined after approximately 12 weeks of therapy. Polysomnograms were scored following the criteria of the American Academy of Sleep Medicine v2.4. Results: 22 participants had mild OSA (5 < AHI ≤ 15 events/h), 30 had moderate OSA (15 < AHI ≤ 30 events/h) and 35 had severe OSA (AHI > 30 events/h). Sixty-two percent of OSA subjects had a minimal dilation pattern (< 1mm) at baseline and this proportion increased to 74% when the mandible was advanced. MAS altered tongue dilatory patterns for 39% (34/87) of participants. MAS was more likely to alter the genioglossus dilatory pattern of people with a tendinous pterygomandibular raphe (Fisher's exact test, P=0.04). Subjects with a beneficial (anterior movement > 1 mm) or non-beneficial dilatory patterns (posterior movement > 1 mm) at baseline were more likely to change (72% and 63%, respectively) than those who had minimal patterns (80% did not change). Seventy-two participants completed the study, and 34 were responders (AHI< 5 or AHI≤10 events/h with change in AHI>50%), 9 were partial responders (change in AHI >50%), and 29 non-responders (change in AHI< 50% and AHI ≥10 events/h). When MAS did not alter tongue dilatory pattern, the presence of pterygomandibular raphe was associated with poorer treatment outcomes (Fisher's exact test, P=0.02). When MAS altered tongue dilatory pattern, 80% of those who changed for non-beneficial pattern were non-responders, and 71% of those who changed for beneficial were full or partial responders. Conclusions: Changes in inspiratory tongue dilatory motion during wakefulness with MAS is associated with MAS treatment outcomes. The mechanism of action of MAS on upper airway dilator muscles differs between subjects, likely reflecting differences in the underlying causes of OSA.