RESUMO
STUDY OBJECTIVES: Acquiring a better comprehension of obstructive sleep apnea physiopathology can contribute to improving patient selection for surgical treatments. We hypothesize that maxillary transverse deficiency restricts the space available for the tongue, leading to upper airway obstruction during sleep. Our primary hypothesis was that maxillary transverse deficiency increases the prevalence of tongue collapse during drug-induced sleep endoscopy (DISE). The secondary hypothesis was that maxillary transverse deficiency will also increase the prevalence of circumferential collapse at the velopharynx. The exploratory hypothesis was that maxillary transverse deficiency is associated with increased obstructive sleep apnea severity. The objectives of this study were to correlate maxillary morphometric measurements with (1) the anatomic level of obstruction during DISE and (2) the apnea-hypopnea index on polysomnography. METHODS: We made a cross-sectional analysis of patients with obstructive sleep apnea undergoing DISE in search of positive airway pressure alternative treatment. Maxillary measurements were collected from a computed tomography scan (interpremolar distance, intermolar distance [IMD] and sella-nasion A point angle), findings from DISE, and sleep study variables from polysomnography. Correlation between computed tomography, DISE, and polysomnography data was assessed using Pearson's correlation, and receiver operating characteristic curves were determined for each facial measurement. RESULTS: Sixty-nine patients were included in the study. The group with velopharyngeal circumferential collapse had mean IMD = 26.30 mm (25.5-31.45), and the group with anteroposterior collapse had mean IMD = 29.20 mm (26.8-33.10; P = .040). The group with complete tongue-base obstruction had mean interpremolar distance = 26.40 mm (25.1-28) and IMD = 26.30 mm (25.6-28.4), and the group without obstruction had mean interpremolar distance = 28.7 mm (27.2-30; P = .003) and IMD = 34.06 mm (32.1-37; P < .001). The receiver operating characteristic curve determined an IMD cutoff of 29.8 mm for predicting tongue-base obstruction. CONCLUSIONS: The maxillary transverse deficiency, identified by reduction in interpremolar distance and IMD, predicted the occurrence of complete tongue-base obstruction, complete concentric collapse at the velopharynx, and multilevel obstruction during DISE. We did not find an association between the maxillary measurements and obstructive sleep apnea severity. These associations hold some promise in ultimately supplanting insights previously available only through DISE.