Three-dimensional assessment of anatomical balance and oral appliance treatment outcome in obstructive sleep apnoea.

PURPOSE: Mandibular advancement splints (MAS) are an effective treatment for obstructive sleep apnoea (OSA). However, MAS are not equally efficacious across all patients and the reasons are not well understood. Craniofacial and upper airway structure individually influence MAS response. We aimed to assess anatomical balance, defined as the ratio of upper airway soft tissue (ST) volume to maxillomandibular enclosure volume, between MAS treatment responders and non-responders. METHODS: OSA patients (apnoea-hypopnea index (AHI) >10 h(-1)) were recruited for MAS treatment. Magnetic resonance imaging of the upper airway was performed during wakefulness without and with MAS in situ. Images were processed for volumetric analysis of upper airway soft tissues (tongue, soft palate, paraphayrngeal fat pads and lateral pharyngeal walls) and three-dimensional cephalometry to acquire intra-mandibular space area (IMA) and total maxillomandibular (Mm) volume. Anatomical balance ratios were compared between MAS treatment responders (AHI <10 h(-1) + 50 % reduction) and non-responders. RESULTS: Image analysis was completed in 69 patients (68 % male, age 50.5 +/- 10.1 years, BMI 29.6 +/- 5.0 kgm(2), AHI 27.0 +/- 14.7 h(-1)) including 36 responders. Soft tissue volumes did not differ between MAS responders and non-responders. Non-responders had increased ST/IMA compared to responders (4.9 +/- 0.6 vs. 4.6 +/- 0.6, p = 0.031). In multivariate logistic regression with AHI and BMI, ST/IMA was the only predictive variable (p = 0.036, ROC AUC 0.7). However, changes in ST/Mm did not directly relate to treatment response. CONCLUSIONS: Anatomical imbalance assessed by intra-mandibular space area was associated with poor MAS treatment response. However, changes in anatomical balance with mandibular advancement did not reflect treatment outcome as static imaging may not adequately capture improvements in upper airway function.