Airway Changes after Cleft Orthognathic Surgery Evaluated by Three-Dimensional Computed Tomography and Overnight Polysomnographic Study

Cleft lip and palate is the most common congenital craniofacial anomaly. Up to 60% of these patients will benefit from cleft orthognathic surgery, which consists primarily of maxillary advancement and mandibular setback to address midface retrusion and relative mandibular protrusion, respectively. It is believed that maxillary advancement can enlarge the airway whilst mandibular setback can reduce the airway, but this has not previously been quantified for cleft patients undergoing orthognathic surgery. This unique longitudinal prospective study of 18 patients was conducted between April 2013 and July 2016. No significant changes occurred by six months postoperatively in body mass index, apnoea-hypopnoea index or lowest oxygen saturation (LSAT). There was a mean increase of 0.73 cm3 in velopharyngeal volume, a mean decrease of 0.79 cm3 in oropharyngeal volume, an improvement in snoring index, and no statistically significant change in hypopharyngeal volume. In conclusion, cleft orthognathic surgery that produced anterior advancement of the maxilla, setback of the mandible and clockwise rotation of the maxillo-mandibular complex resulted in increased velopharyngeal, decreased oropharyngeal and unchanged hypopharyngeal airways, and improved snoring, but did not significantly alter objective sleep-related breathing function.

. Movement of facial landmarks; Horizontal (data presented in mm, mean ± standard deviation).
Changes in airway dimensions are provided in Tables 5, 6 and 7. There were no statistically significant changes in the minimum LSAT measurements at velopharyngeal, oropharyngeal and hypopharyngeal airway levels. There was a mean increase of 0.22 cm in the minimum AP measurement at the velopharyngeal level (p = 0.005), a mean decrease of 0.15 cm at the oropharyngeal level (p = 0.007), and no statistically significant change at the hypopharyngeal level. There was a mean increase of 0.46 cm 2 in the minimum CSA at the velopharyngeal level (p < 0.001), a mean decrease of 0.81 cm 2 at the oropharyngeal level (p = 0.025), and no statistically significant change at the hypopharyngeal level. There was a mean increase of 0.73 cm 3 in velopharyngeal volume (p = 0.028), a mean decrease of 0.79 cm 3 in oropharyngeal volume (p = 0.038), and no statistically significant change in hypopharyngeal volume and total pharyngeal volume.

Discussion
Cleft lip and palate is the most common congenital facial anomaly. There is a need for multidisciplinary care for these patients during many stages of their lives [9][10][11] . This study is typical of complete cleft care, involving multidisciplinary treatment and analyses of their treatment outcomes.     Table 7. Changes in the pharyngeal airway; Volume measurements at three pharyngeal airway segments and total pharyngeal airway.
When cleft lip/palate patients grow, maxillary retrusion is frequently noted. The main cause of reduced anterior-posterior maxillary growth is the effect of scar tissue formation from surgical cleft lip repair, gingivoperiosteoplasty, hard palate closure and alveolar bone grafting [12][13][14][15][16][17][18] . It has been estimated that up to 60% of cleft patients require Le Fort I osteotomies to correct maxillary hypoplasia 2 .
Growth restriction causes the maxilla to become deficient and retrusive, leading to a prognathic facial appearance 19 . The patients in this current study all presented some degree of mandibular prognathism, demonstrated by increased SNB (Table 1). Such discrepancies in maxillary and mandibular development can cause severe class III malocclusion. The surgical correction of maxillary hypoplasia and cleft malocclusion with conventional orthognathic surgery (Le Fort I and bilateral sagittal split osteotomy) is generally conducted after skeletal maturity. Adjunctive orthodontic treatments are essential. Computer-aided 3D surgical planning using CT scans is performed by the orthodontic team in our institution and has improved surgical outcomes 20,21 . Orthognathic surgery for cleft patients is intended to improve dental occlusion, skeletal relationships and facial appearance 22 . Of the 18 patients studied, 14 (78%) reported a marked improvement and four (22%) reported some improvement in their facial appearance with orthognathic surgery.
To correct a hypoplastic maxilla with inadequate incisor show, typically the maxilla is advanced and moved inferiorly in the anterior portion (demonstrated by movement of the A point), and rotated clockwise at the PNS relative to its vertical position. The mandibular body is conversely moved posteriorly. Overall, the maxilla-mandibular complex is clockwise rotated and the mandibular body moved inferiorly. This is demonstrated as setback and inferior movement of the B points, POGs, GGTs and LCIs (Tables 2 and 3; Fig. 1).
The 2-D and 3-D influences on the airway of maxillary advancement and mandibular setback plus clockwise rotation of the maxilla-mandibular complex in cleft orthognathic surgery have never been reported in the literature. Mandibular setback surgery for skeletal class III malocclusion or mandibular prognathism caused decreases in the airway dimensions 23,24 , and maxillary advancement produced increased upper airway dimensions 25,26 . However, cleft orthognathic surgery requires both maxillary advancement and mandibular setback. The present study demonstrated a postsurgical increase in velopharyngeal volume from 5.67 cm 3 to 6.40 cm 3 , a decrease in oropharyngeal volume from 4.45 cm 3 to 3.65 cm 3 , and no significant changes in hypopharyngeal and total pharyngeal volumes (Fig. 2). These three-dimensional changes were also observed in the minimum cross-sectional area measurements at the three levels of the airway. There was an increase of minimum cross-sectional area at the velopharyngeal airway (Fig. 3), a decrease at the oropharyngeal airway (Fig. 4) and no significant change at the hypopharyngeal airway (Fig. 5). These airway changes were mainly due to AP distance rather than LAT distance changes (Figs 1, 2, 3, 4 and 5). However, these airway changes, as measured by 3D CT scans, did not translate into objective respiratory functional changes during sleep.
Although cleft lip-palate patients frequently complain of snoring and respiratory difficulties during sleep, overnight polysomnographic studies are not commonly performed as a presurgical assessment. Our findings suggest that cleft orthognathic surgery did not significantly alter these patients' AHIs or LSAT readings six months after surgery.
Strengths and limitations of this study. This study has several strengths. First, our Craniofacial Center is a major teaching hospital referral centre that has treated most of the cleft patients in our nation. All patients were consecutive consenting patients from the first and corresponding authors' cleft practices and both surgeons exercise the same surgical approach to cleft orthognathic surgery; thus, selection bias is minimised. Second, the study group has extensive experience in airway measurement techniques, and have already published several peer-review articles regarding airway measurement using 3 dimensional CT scan technology 5,23,25 . Third, given that Taiwan is a small island, complete six months follow up for most of our patients was achieved.  Three-dimensional airway model before (left) and after (right) surgery. Notice that the velopharyngeal airway has increased, the oropharyngeal airway has decreased, and the hypopharyngeal airway has not changed.  The limitations of this study merit emphasis. First, although the Craniofacial Center treats several hundred patients each year with orthognathic surgery, most patients declined to be included in the present investigation due to the need for two overnight polysomnographic studies. Second, we included only unilateral complete cleft lip and palate patients, but patients with bilateral complete cleft lip and palate tend to have more severe midface retrusion and would warrant further study. Third, in our institution more than 90% of patients proceeding with cleft orthognathic surgery have bimaxillary surgery for aesthetics. Airway and polysomnographic changes in patients electing only for maxillary advancement invites further study. Fourth, most of our patients did not have, or only had mild, obstructive sleep apneoa before surgery. Further study in cleft patients with moderate to severe obstructive sleep apneoa warrants future investigation.
We consider patients with objectively small airway, obesity, and/or symptoms, signs or family history of obstructive sleep apnea, as high risk for airway compromise following mandibular setback; their presence may lead us to discourage mandibular setback. Other unfavourable characteristics for airway risk following mandibular setback appear to be male sex and middle agedness. Careful surgical planning is important so as not to compromise basic airway needs after mandibular setback 23 .
In conclusion, cleft orthognathic surgery that produced anterior advancement of the maxilla, setback of the mandible and clockwise rotation of the maxillo-mandibular complex resulted in increased velopharyngeal, decreased oropharyngeal and unchanged hypopharyngeal airways, and improved snoring, but did not significantly alter objective sleep-related breathing function.

Methods
This prospective longitudinal study was designed to investigate skeletal changes, two and three dimensional airway changes, polysomnographic changes and patient reported satisfaction with facial appearance after cleft orthognathic surgery. with the upper margin of the velopharynx as a horizontal plane that is perpendicular to the sagittal plane and passes through the posterior nasal spine; and the lower margin of the velopharynx as a horizontal plane that is perpendicular to the sagittal plane and passes through the tip of the uvula; (B) Oropharyngeal airway with its upper margin as the tip of the uvula and its lower margin as a horizontal plane perpendicular to the sagittal plane that passes through the tip of the epiglottis; (C) Hypopharyngeal airway with its upper margin at the tip of the epiglottis and its lower margin as a horizontal plane that is perpendicular to the sagittal plane that passes through the vocal cord. Data sharing statement, the data can be assessed: https://www1.cgmh.org.tw/intr/intr2/c32540/en/clinical_ trials_studies.html; extra data is available by emailing plastreconst@gmail.com.
Twenty-two complete unilateral cleft lip patients with class III malocclusion were treated with orthognathic surgery at the Craniofacial Center of Chang Gung Memorial Hospital from 1992-1996, and were included in the present study. Eighteen patients completed the study; all underwent overnight polysomnography and computed tomography (CT) scans before and at 6 months after orthognathic surgery. Demographic data was recorded (age, sex, body mass index; BMI). The BMI was recorded at the time of polysomnography, before surgery and at 6 months after surgery.
The inclusion criteria were: 1. unilateral cleft lip/palate patients; 2. patients with midface retrusion and malocclusion that required orthognathic surgery; 3. such patients who provided consent for inclusion; 4. patients who had reached skeletal maturation.
Exclusion criteria were: 1. presence of other craniofacial anomalies; 2. soft tissue airway surgery within one year; 3. presence of medical conditions that would preclude orthognathic surgery.
Orthognathic Surgery. Bimaxillary osteotomies were executed according to three-dimensionally planned surgical movements 21 . All patients underwent single splint orthognathic surgery 3 . All patients received combined Le Fort I maxillary osteotomies to advance the maxilla and bilateral sagittal split ramus osteotomies to set back the mandible. All surgical complications were recorded.

Measurements.
All measurements and data analyses were performed by a single investigator. CT scan data was stored in digital imaging and communications in medicine (DICOM) format and was transferred to a computer station with AVIZO version 7.01 software (VSG, France). The 3-dimensional CT images were first oriented as follows: axial plane was the Frankfort Horizontal (FH) plane (defined by the plane passing through bilateral orbitale and bilateral porion); coronal plane was perpendicular to the FH plan; sagittal plane was perpendicular to the FH and coronal planes, this passed through the midpoint of the bilateral orbitale. Three dimensional CTs (before and six months after surgery) were superimposed in cranial structures before measurements.
Facial skeletal movements before and after surgery were measured using vertical and horizontal distances from the landmarks to the sella (Fig. 6). The occlusal plane angle was that angle between the occlusal plane and the S-N plane. Airway changes, before and after surgery, were measured using 3-dimensional airway modelling as follows. The upper airway was divided into 3 segments: velopharynx, oropharynx and hypopharynx (Fig. 7). The minimum anterior-posterior dimension (AP), lateral dimension (LAT), and cross-sectional area (CSA) were measured at each of these three segments (velopharynx, oropharynx and hypopharynx; Fig. 8). The CT slices with the minimum available measured AP distances for each respective airway level were defined as the CT slices for preoperative and postoperative measurements.
Each patient's reported satisfaction regarding their facial appearance was obtained 6 months after surgery and was graded as (1) Marked improvement; (2) Some improvement; (3) No improvement and (4) Worse.
Statistical analyses. The paired t-test was used to compare facial skeletal movement, upper airway measurements and polysomnographic data. Fisher's exact test was used to compare NSD scores. All statistical analyses were performed with SPSS 17.0 (SPSS, Inc., Chicago, Ill). A p value of less than 0.05 denoted a statistically significant difference. Mean results are provided with standard deviations (mean ± standard deviations).