Maxillomandibular advancement (MMA) can be effective for managing obstructive sleep apnoea (OSA); however, limited information is available on the predictor surgical variables. This study investigated whether normalization of the mandibular occlusal plane (MOP) was a determinant factor in curing OSA. Patients with moderate or severe OSA who underwent MMA were evaluated by preoperative and postoperative three-dimensional (3D) scans and polysomnograms. The postoperative value of MOP and the magnitude of skeletal advancement were the predictor variables; change in the apnoea–hypopnoea index (AHI) was the main outcome variable. Thirty-four subjects with a mean age of 41 ± 14 years and 58,8% female were analysed. The Epworth Sleepiness Scale (ESS) was 17.4 ± 5.4 and AHI was 38.3 ± 10.7 per hour before surgery. Postoperative AHI was 6.5 ± 4.3 per hour ( P < 0.001) with 52.94% of the patients considered as cured, and 47.06% suffering from a mild residual OSA with ESS 0.8 ± 1.4 ( P < 0.001). 3D changes revealed a volume increase of 106.3 ± 38.8%. The mandible was advanced 10.4 ± 3.9 mm and maxilla 4.9 ± 3.2 mm. MOP postoperative value was concluded to be the best predictor variable. Treatment planning should include MOP normalization and a mandibular advancement between 6 and 10 mm. The maxillary advancement would depend on the desired aesthetic changes and final occlusion.
Obstructive sleep apnoea syndrome (OSA) is a medical condition with an estimated prevalence of 4–6% in the adult population as defined by an apnoea and hypopnoea index (AHI) of ≥5 events per hour . OSA is related to cardiovascular and metabolic diseases, among others, and is potentially life threatening. Campos-Rodríguez et al. identified a 10% absolute increase in mortality rate for OSA patients who did not use continuous positive airway pressure (CPAP).
The standard management of OSA is CPAP therapy , but empirical studies have suggested that CPAP compliance is low, ranging from 30% to 60% . Most patients report that they are uncomfortable using the device and the majority of patients would prefer alternative treatment options.
Maxillomandibular advancement (MMA) is the most effective surgical therapy for adult patients with OSA outside the tracheostomy treatment option. The treatment consists of osteotomies of the mandible and the maxilla, producing simultaneous advancement of the maxillomandibular complex. This procedure enlarges the pharyngeal space by expanding the skeletal framework, and thus reducing the risk of pharyngeal collapse during negative pressure inspiration .
It is currently used for patients with dentofacial deformities requiring orthognathic surgery. Patients with a severe obstructive disease, in whom all applicable conservative therapies have failed or proved intolerable can also benefit from MMA surgery. However, the procedures are difficult to perform, requiring very specific surgical experience. In addition, careful surgical planning and postoperative follow-up are two important elements to consider and ensure long-term stability.
MMA is a revolutionary alternative to conventional treatments for OSA, but there are no well-defined studies about the surgical variables that predict AHI normalization and/or the minimum amount of skeletal advancement that can lead to OSA resolution. This study reports on the surgical treatment outcomes of the 34 subjects with moderate to severe OSA after MMA, with special attention to treatment planning and 3D pharyngeal changes. The study’s primary objective was to determine the type and amount of skeletal movements as predictors to correct the AHI and pharyngeal airway space volume (PAS-VOL) changes.
In this study, the hypothesis was that normalization of the mandibular occlusal plane is mandatory to effectively enlarge the upper airway when MMA is performed and should thus be included in the treatment planning to cure the disease. In the cases where the amount of skeletal advancement is better defined, a more rational surgical planning regime will be possible enabling the scope of the technique that is not limited to treating retrognathic patients.
Material and methods
The study population comprised all the patients presenting for the evaluation and management of OSA. Data were collected between December 2007 and January 2015 at the Nisa Pardo de Aravaca Hospital’s Department of Oral and Maxillofacial Surgery in Madrid, Spain.
It was designed as a prospective cohort study that aimed to follow up the selected patients over 12 months after intervention. No patient received any concomitant surgical adjunctive procedure, before or after MMA. All patients signed an informed consent for surgery, and specific consent for 3D scan examination, photographs, and 3D virtual planning. Permission from Institutional Review Board Human Studies Committee was obtained.
Diagnosis of OSA was according to current guidelines, and severity was defined as the following: mild (AHI 5–14 per hour), Moderate (15–29 per hour), and severe (>30 per hour) .
The following criteria were used to select the study participants: patients had to be adults with moderate to severe OSA (AHI ≥ 15 per hour); refusal to begin or continue treatment with CPAP; health conditions compatible with this surgery; normal anatomy of the mandibular ascending ramus and the temporomandibular joint (TMJ).
A thorough review of the patients’ mandibular anatomy was completed prior to surgery. An ascending ramus length >50 mm (distance measured from the condylion to the gonion in a 3D scan) was required to perform a bilateral sagittal split osteotomy of the mandible. Patients with severe ramus hypoplasia and/or condylar resorption were considered as inadequate candidates for receiving conventional MMA. Intraoral mandibular distraction osteogenesis was indicated and performed in this group of patients as it has been suggested in previous studies and were excluded of this series.
Additional criteria in the cohort selection included acceptance of customized morphing software simulation; the patient’s acceptance of corrective orthodontic treatment pre- and postoperatively and complete retention after removing the braces during a minimum of a 2-year follow-up with a good dental hygiene track record.
Patients were excluded from the study if they had central sleep apnoea diagnosed by a neurophysiologist, were active smokers, had previous uvulopalatopharyngoplasty (UPPP), or complete edentulism (toothless or had extremely bad remaining teeth) and/or severe TMJ disease in stages IV–V of Wilkes classification.
Other demographic variables collected included age, sex, body mass index (BMI), treatment with CPAP, and using a mandibular advancement device (MAD) as additional parameters for describing all patients selected for the study.
In terms of the OSA, AHI, mean blood oxygen saturation ( S aO 2 ), oxygen desaturation index (ODI), and the value of the Epworth Sleepiness Scale (ESS) were recorded for all patients before 6 six months after MMA. The ODI was observed to be 3% or more oxygen desaturation lasting 10 seconds or more per hour. Each patient underwent hospital-based nocturnal polysomnography (PSG) preoperative and 6 months postoperatively.
The PSG results from the standard PSG locations from electroencephalogram (F3, F4, C3, C4, O1, O2, A1, and A2), electro-oculogram, and electromyogram recordings were obtained with standard procedures by means of Cadwell Easy 2 (Cadwell, Kennewick, WA, USA) digital data acquisition system using a sampling frequency of 512 Hz. Airflow and thoracoabdominal effort were recorded quantitatively by nasal pressure cannula, thermistor, and respiratory inductance plethysmography. The studies were done following the American Academy of Sleep Medicine standard diagnostic guidelines by an experienced doctor in sleep medicine .
Each selected patient underwent a preoperative anaesthesia evaluation. Hypertension, obesity, and diabetes were not considered as absolute contraindications for this surgery.
Standardized preoperative aesthetic evaluation based on clinical facial analysis (CFA) was performed for all patients using Arnett principles .
A 3D computed tomography scan was captured with an i-CAT machine (Imaging Sciences International, Hatfield, PA) before surgery to evaluate both the superior airway and the skeletal components of this anatomical region. A waxbite registration (Delar, Lake Oswego, USA) was used to verify that the TMJs were seated in the glenoid fossa during the procedure. Precise indications were also given to place the tongue against the incisor teeth and to hold the breath during exhalation. Head orientation was checked as often as necessary to maintain a natural head position . All explorations were supervised personally by the same investigator (P.R.-B.).
The scan images were loaded into Dolphin Imaging 11.0 Premium 3D software (Dolphin Imaging and Management Solutions, Chatsworth, CA, USA) for the initial anatomical analysis and surgical planning. The superior upper airway boundary was defined at midsagittal plane, at the level of the basion (Ba, median point of the anterior margin of the foramen magnum), and the inferior boundary at the anteroinferior vertex of the fourth cervical vertebra (C4). The posterior nasal spine was discarded as the superior boundary because its position changes after the maxillary osteotomy. The lateral and posterior boundaries of the model consisted of the pharyngeal walls and the anterior boundary, the anterior wall of the pharynx, the base of the tongue, and the soft palate. All this defined volume was called the pharyngeal airway space volume (PAS-VOL), and was measured in mm 3 . The minimum axial area (MAA) is the minimum cross-sectional area (axial area) of the entire PAS, and it was measured in mm 2 .
A 3D surgical virtual planning with specific Software (CD Ortosan, Madrid, Spain) was performed to elaborate the intermediate surgical splint. The angle between the mandibular occlusal plane (MOP) and the true vertical line (TVL), the MOP–TVL angle, was especially considered to allow maxillomandibular counterclockwise (CCW) advancement (normal values between 92° and 96°).
The 3D superimposition of both pre- and postoperative scans were analysed with this software to measure the real skeletal spatial changes of pogonion (Pg), A point (A), and MOP–TVL angle after surgery, and all the airway changes ( Figs. 1 and 2 ).
The airway and skeletal variables considered for the study were both the preoperative and postoperative PAS-VOL and the MAA including the MOP–TVL angle, pogonion changes (Pg), and A point changes (A).
Surgery was performed under general anaesthesia with endotracheal intubation and controlled hypotension. No patient required blood transfusion. All procedures were performed intraorally with local bone grafts from the sinus walls and osseous nasal septum as well as plasma-rich growth factors as needed. Stabilization was achieved with titanium plates and screws and intermaxillary fixation was not needed. A continuous water-circulating external cooling device (Hilotherm (R) GmbH, Argenbuhl-Eisenharz, Germany) was used for all patients. Maximum hospitalization was 48 hours.
Postoperative follow – up
All patients were evaluated weekly during the first 8 weeks and monthly for the first year postoperatively.
The post-surgical assessment was performed 6 months after surgery. Consisting of the following clinical reviews: a radiological evaluation with panoramic and lateral cephalometric radiographs, a nocturnal PSG, and a new 3D computed tomography scan, with the same preoperative protocol.
Surgical success was defined as a postoperative AHI <20 and a 50% reduction in AHI after MMA procedure, and the “cure rate” was defined as a postoperative AHI <5, as described by Holt yang Guilleminault .
Data quality was checked through the assessment of all the variable ranges, outliers, and errors. First, the patient data formed the basis of a descriptive analysis, computed to provide an overview of the study population. Continuous variables were tested for normality applying the Shapiro test, allowing the researchers to proceed with further analysis with the usual parametric tests if normality was met, otherwise with non-parametric tests.
In the case of the quantitative variables, differences by sex were performed with the t -test for the variables in which normality could be assumed, and the Wilcoxon–Mann–Whitney test in cases where normality could not be assumed. Pre- and post surgery pair-wise comparisons were calculated with the Wilcoxon signed rank sum test. The quantitative variables were expressed as mean ± standard deviation, while the qualitative ones as counts and percentage of the total. Comparisons between the qualitative variables were done with χ 2 test. In order to assess association in a bivariate analysis, Pearson’s correlation was calculated. Given the observational, standard practice approach, no formal a priori sample size calculation was estimated. However, a posteriori power calculation indicated plenty of significance in all comparisons. For all analyses, P < 0.05 was considered statistically significant. Statistical software R version 3.2.3 was used .