The objective of this systematic review was to evaluate the effect of different types of orthognathic surgery on the dimensions of the upper airways assessed using three-dimensional images. An electronic search was performed in Cochrane Library, Medline, Scopus, VHL, Web of Science, and the System for Information on Grey Literature in Europe, ending January 2015. Inclusion criteria encompassed clinical studies in humans, patient age >15 years, patients submitted to maxillary or mandibular advancement or setback surgery, isolated or in combination, and presentation of airway measures, specifically volume and/or minimum cross-sectional area (CSA), obtained from computed tomography or magnetic resonance imaging. Additional searches were conducted on the references of included articles and in the NLM catalogue. An assessment of the risk of bias was performed. A total of 1180 studies were retrieved, of which 28 met the eligibility criteria; one was later excluded as it presented a high risk of bias. A meta-analysis was performed. There is moderate evidence to conclude that the upper airway minimum CSA increases significantly (124.13 mm 2 ) after maxillomandibular advancement (MMA); the total volume increases significantly after MMA (7416.10 mm 3 ) and decreases significantly after maxillary advancement + mandibular setback (−1552.90 mm 3 ) and isolated mandibular setback (−1894.65 mm 3 ).
The upper airways are of increasing interest to the different medical professionals working in the head and neck region, primarily due to the associations between craniofacial development and morphology, the upper airway configuration, and respiratory disorders.
Although orthognathic surgeries are performed to correct bone discrepancies, they inevitably affect the relationship between the soft and skeletal tissues. Maxillary and/or mandibular surgical replacement can cause different changes in the area and volume of the oral and nasal cavities, depending on the magnitude and direction of correction, and subsequently may influence the quality of sleep of treated patients in the long term, when associated with risk factors.
According to Mattos et al., the airway anteroposterior length may be altered in the following ways: a decrease in the region of the soft palate and base of the tongue after isolated mandibular setback (MdS) surgery; an increase in the posterior nasal spine region and decrease in the soft palate, tongue, and vallecula regions after combined surgery of maxillary advancement with mandibular setback (MxA + MdS); and an increase in the soft palate region after maxillomandibular advancement (MMA) surgery. However, these results were based on cephalometric analyses.
Although cephalometry has been the recommended method for the analysis of craniofacial development for many years, the representation of the airways and other three-dimensional (3D) structures in two dimensions has its limitations. It is known that computed tomography (CT) and magnetic resonance imaging (MRI) allow linear, cross-sectional area (CSA), and volumetric assessment of the upper airways, providing the otherwise unavailable useful quantitative and qualitative information. Both of these methods have been studied extensively and are considered reliable for reproducible assessment of the upper airways when based on well-defined parameters.
No systematic reviews comparing changes in the airways resulting from different orthognathic surgeries exclusively using 3D examination have yet been reported in the literature. The systematic review by Mattos et al. compared different types of orthognathic surgery and their effects on the upper airway dimensions; however, the meta-analysis used data from two-dimensional images only, as the four articles using CT were not comparable. Fernández-Ferrer et al. assessed 3D images (CT) to investigate the results of one type of surgery (mandibular setback) only, and no meta-analysis was performed. It should also be noted that more than five new studies have been published since the completion of the literature search of these two previous reviews. Recently, an increase in this type of surgical assessment has been observed due to the introduction of these methods in the routine practice of surgeons and dentists, and also because of an increase in research in the field of OSAS (obstructive sleep apnoea syndrome). A search conducted on the Scopus database indicated an increasing number of publications on the subject, particularly since 2008.
The aim of this study was to assemble, through a systematic review, scientific evidence related to the effects of different types of orthognathic surgery on the minimum CSA and volume of the upper airway as assessed using CT or MRI.
Materials and methods
This review was based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. The review protocol for this study was registered in the PROSPERO database as CRD42014013323 ( www.crd.york.ac.uk/PROSPERO ).
The inclusion criteria were as follows: prospective or retrospective clinical studies in humans; patient age >15 years; patients submitted to surgeries of maxillary or mandibular advancement or setback, isolated or combined; measurements of the upper airways, including volume and/or the minimum CSA, from the whole upper airway, retropalatal and/or retrolingual regions (pre- and post-surgical, or the difference between these times, with the standard deviation, P -value, or any other variability measures) obtained from CT or MRI. The exclusion criteria were the following: case reports, case series, review articles, editorials, reviews, and books; articles on reliability and/or comparison of methods or programmes of assessment; articles presenting only the axial area of specific levels; studies concerning patients who had craniofacial anomalies, lip and/or cleft palate, or patients who were systemically compromised; and studies concerning individuals who underwent orthognathic surgery involving transverse corrections or distraction osteogenesis.
Eligible studies that answered the PICO question ( Table 1 ) were identified by an electronic search conducted in the following databases: Cochrane Library, Medline (via PubMed), Scopus, VHL (Virtual Health Library–Lilacs and BBO), Web of Science, and the System for Information on Grey Literature in Europe (OpenGrey). The end-point of the search period was January 9, 2015. Specific search strategies were developed for each database with the guidance of a librarian (DMTPF); the PubMed strategy is presented in Table 2 . Details of the searches for all databases are provided in a supplementary file (Supplementary Material, Table S1). A complementary search was performed of journals referenced in the National Library of Medicine (NLM) catalogue (via PubMed) containing the entries of journals referenced in the NCBI database using the term ‘oral and maxillofacial surg*’. The journals with their title in English that were once indexed in PubMed but are no longer indexed were selected for this additional search. A manual search of the reference lists of studies included in this systematic review was also performed.
|P – Population||Patients submitted to orthognathic surgery|
|I – Intervention||Surgical correction involving the anteroposterior aspects of the maxilla and/or mandible|
|C – Comparison||Between the different types of orthognathic surgery|
|O – Outcome||Dimensional changes of the upper airway (minimum cross-sectional area and volume) measured using CT or MRI images|
|Question||What are the effects of orthognathic surgery for anteroposterior correction of the maxilla and/or mandible on the dimensions of the upper airways assessed using 3D images?|
|((((((((((((((((((((((((Orthognathic Surgery[MeSH Terms]) OR “Orthognathic Surgery”[Title/Abstract]) OR Orthognathic Surgical Procedures[MeSH Terms]) OR “Orthognathic Surgical Procedures”[Title/Abstract]) OR Surgery, Oral[MeSH Terms]) OR Maxillofacial Abnormalities[MeSH Terms]) OR Maxillofacial Abnormalities[Title/Abstract]) OR Maxillofacial Development[MeSH Terms]) OR Mandibular Advancement[MeSH Terms]) OR Orthodontics[MeSH Terms]) OR Mandible[MeSH Terms]) OR Maxilla[MeSH Terms]) OR “jaw surgery”[Title/Abstract]) OR “bimaxillary surgery”[Title/Abstract]) OR “maxillo mandibular advancement”[Title/Abstract]) OR “surgical orthodontic treatment”[Title/Abstract])))) OR (((Advancement[Title/Abstract] OR setback[Title/Abstract] OR surger*[Title/Abstract]) AND (Mandibular[Title/Abstract] OR Maxillary[Title/Abstract]))))) AND (((((((((((((((((((pharynx[MeSH Terms]) OR pharyn*[Title/Abstract]) OR nasopharynx[MeSH Terms]) OR nasopharyn*[Title/Abstract]) OR hypopharynx[MeSH Terms]) OR hypopharyn*[Title/Abstract]) OR oropharynx[MeSH Terms]) OR oropharyn*[Title/Abstract]) OR airway obstruction[MeSH Terms]) OR “airway obstruction”[Title/Abstract])OR sleep apnea syndromes[MeSH Terms]) OR “Sleep Disordered Breathing”[Title/Abstract]) OR laryngopharyn*[Title/Abstract]) OR “posterior airway space”[Title/Abstract]) OR “air space”[Title/Abstract]) OR “upper airway”[Title/Abstract]) OR “oral airway”[Title/Abstract]) OR “nasal airway”[Title/Abstract]))) AND (((((((((((((((((((((((((((Tomography, X-Ray Computed[MeSH Terms]) OR “Cone-Beam Computed Tomography”[Title/Abstract]) OR CBCT[Title/Abstract]) OR Multidetector Computed Tomography[MeSH Terms]) OR “Multidetector Computed Tomography”[Title/Abstract]) OR Imaging, Three-Dimensional[MeSH Terms]) OR “Three-Dimensional Image”[Title/Abstract]) OR “3-D Imaging”[Title/Abstract]) OR Magnetic Resonance Imaging[MeSH Terms]) OR “Magnetic Resonance”[Title/Abstract]) OR “Tomography MR”[Title/Abstract]) OR “Tomography NMR”[Title/Abstract]) OR “minimum axial area”[Title/Abstract]) OR volume[Title/Abstract]) OR area[Title/Abstract]) OR linear[Title/Abstract]) OR “airway cross section”[Title/Abstract]) OR “CBCT Scans”[Title/Abstract]) OR “CBCT Scan”[Title/Abstract]) OR “CAT Scans”[Title/Abstract]) OR “CAT Scan”[Title/Abstract]) OR “Cone Beam CT”[Title/Abstract]) OR “NMR Imaging”[Title/Abstract]) OR “MRI Scan”[Title/Abstract]) OR Invivo[Title/Abstract]) OR Dolphin[Title/Abstract]))))|
Specific search strategy for each database.
After the exclusion of duplicate articles, two reviewers (IOC and COL) independently examined the list of titles and abstracts according to the eligibility criteria. The article was reviewed in full if the title and abstract did not provide sufficient information. Disagreements between the two reviewers were resolved by seeking a third reviewer’s (CTM) opinion in a consensus meeting. The authors of some studies were contacted by e-mail or social media to check the eligibility criteria and to provide missing data or information on sample overlap, including patient age. In cases where the articles did not present the minimum age of the patients and the authors did not respond to attempts at contact, these were included only if they mentioned that the patients were adults and/or presented the average age ( x ) and standard deviation ( σ ), and [ x − 2 ( σ )] > 15 years.
After checking the full text of the articles for eligibility, two reviewers (IOC and COL) analyzed the studies for risk of bias based on the quality assessment method reported by Mattos et al. Some adjustments were made (the items corresponding to the control group and blinding assessment were removed and the sample size calculation was scored) and the selected papers were evaluated in six categories, as described in Table 3 . Each article was classified as presenting a low risk of bias (score ≥4.5), moderate risk of bias (score >2 and <4.5), or high risk of bias (score ≤2). High-risk studies were excluded from the review. Any disagreement during this step was resolved by consulting a third reviewer (AAC).
|Component||Risk of bias||Points||Definition|
|1. Eligibility criteria for participants described||Low||1.0||Inclusion and/or exclusion criteria described|
|Moderate||0.5||No description of criteria, but selection done at least by age and type of surgery|
|High||0||No description of criteria for selection|
|2. Statistical analysis performed||Low||1.0||Statistical analysis described fully, including sample size calculation, and adequate|
|Moderate||0.5||Statistical analysis not described fully or inadequate|
|High||0||No statistical analysis applied|
|3. Reliability of measures tested||Low||1.0||Measures repeated and statistical tests applied|
|Moderate||0.5||Measures repeated and inadequate or no statistical tests applied|
|High||0||Measures not repeated|
|4. Drop-outs reported||Low||1.0||Drop-outs reported with an explanation|
|Moderate||0.5||Drop-outs reported with no explanation|
|High||0||No description of drop-outs|
|5. Follow-up period reported||Low||1.0||Follow-up period reported|
|High||0||No description of the follow-up period or unclear follow-up period|
|6. Potential bias and trial limitations addressed||Low||1.0||Description of potential bias and trial limitations acknowledging them|
|High||0||No description of potential bias or trial limitations|
The data from the selected articles were extracted and presented in a table. A meta-analysis was performed using Comprehensive Meta-Analysis software (version 3.2.00089; Biostat, Inc., Englewood, NJ, USA) with a fixed-effects model. The correlation coefficient for comparison was estimated from the studies that presented pre- and post-surgical means for the volume and minimum CSA and the post-surgical minus pre-surgical mean difference with the standard deviation. In the case of two or more options, the longest follow-up period was considered. Heterogeneity was tested using the Q -value, I 2 index, and Tau 2 , and a sensitivity analysis was performed in the case of high heterogeneity.
Studies were compared (with ‘type of surgery’ as a subgroup) for changes in the following measurements: total volume, retropalatal volume, retrolingual volume, minimum CSA, retropalatal minimum CSA, and retrolingual minimum CSA. A comparison of total volume change between MdS and MxA + MdS was performed using data collected from studies that examined these two groups (subtracting the changes in the MdS group from the changes in the MxA + MdS group).
A flow diagram of the study selection process is shown in Fig. 1 . An additional search was performed in the following journals identified from the NLM catalogue: Journal of the Korean Association of Oral and Maxillofacial Surgeons , Journal of Maxillofacial and Oral Surgery , and Atlas of the Oral and Maxillofacial Surgery Clinics of North America ; one article was found. One article was excluded as we were unable to contact the authors. The total number of articles included was 28.
The risk of bias assessment is shown in Table 4 . Only eight studies were classified as having a low risk of bias, and the majority of the rest had a moderate risk of bias. One study had a high risk of bias and was excluded. The least scored items and the main cause of high risk of bias in the articles was blinding, followed by report of drop-outs. Although most studies presented adequate statistical analysis, only Brunetto et al. performed a sample size calculation.
|Study||Eligibility criteria for participants described||Statistical analysis performed||Reliability of measures tested||Drop-outs reported||Follow-up period reported||Potential bias and trial limitations addressed||Total points||Risk of bias a|
|Abramson et al., 2011||1||0.5||0||1||0||1||3.5||Moderate|
|Bianchi et al., 2014||1||0.5||0||0||1||1||3.5||Moderate|
|Brunetto et al., 2014||1||1||1||0||1||1||5.0||Low|
|de Souza Carvalho et al., 2012||1||0.5||0||0||1||0||2.5||Moderate|
|Faria et al., 2013||1||0.5||0||0||1||0||2.5||Moderate|
|Gokce et al., 2014||1||0.5||1||0||1||0||3.5||Moderate|
|Gordina et al., 2013||1||0||0||0||1||0||2.0||High|
|Hernández-Alfaro et al., 2011||0.5||0.5||0||0||1||1||3.0||Moderate|
|Hong et al., 2011||1||0.5||1||0||1||1||4.5||Low|
|Hsieh et al., 2014||1||0.5||1||0||1||0||3.5||Moderate|
|Jakobsone et al., 2010||1||0.5||1||1||1||1||5.5||Low|
|Kim et al., 2010||1||0.5||0||0||1||0||2.5||Moderate|
|Kim et al., 2013||1||0.5||1||0||1||0||3.5||Moderate|
|Kim et al., 2014||1||0.5||1||0||1||1||4.5||Low|
|Kim et al., 2014||1||0.5||0||0||1||0||2.5||Moderate|
|Kochel et al., 2013||1||0.5||1||0||1||0||3.5||Moderate|
|Lee et al., 2009||1||0.5||1||0||1||1||4.5||Low|
|Li et al., 2014||1||0.5||1||0||1||0||3.5||Moderate|
|Panou et al., 2013||1||0.5||1||0||1||1||4.5||Low|
|Park et al., 2010||1||0.5||0.5||0||1||1||4.0||Moderate|
|Park et al., 2012||1||0.5||1||0||1||0||3.5||Moderate|
|Raffaini and Pisani, 2013||1||0.5||0||0||1||1||3.5||Moderate|
|Schendel et al., 2014||1||0.5||0||0||1||1||3.5||Moderate|
|Uesugi et al., 2014||1||0.5||1||1||1||0||4.5||Low|
|Wang et al., 2012||1||0.5||0||0||1||0||2.5||Moderate|
|Wang et al., 2012||1||0.5||0||0||1||0||2.5||Moderate|
|Zinser et al., 2013||1||0.5||1||0||1||1||4.5||Low|