Effects of bicuspid extractions and incisor retraction on upper airway of Asian adults and late adolescents: A systematic review

Abstract Objectives This systematic review aimed to assess the effects of bicuspid extractions and incisor retraction on airway dimension, hyoid position and breathing of adults and late adolescents. Methods The review was conducted according to PRISMA guidelines. Eight databases including PubMed, EMBASE, Web of Science and Scopus were searched to August 2018. Minimum age of participants was 16 years. The intervention was dual‐arch bicuspid extractions with incisor retraction. Outcomes were airway dimension, hyoid position and breathing assessment. Results All nine publications meeting inclusion criteria were from Asia. They were divided into three Asian subregions. All East Asian lateral cephalometric studies reported anteroposterior airway narrowing at the oropharynx and sometimes the hypopharynx. However, the narrowing was small, comparable to measurement errors, and highly variable. Two out of three East Asian computed tomography (CT) studies described reductions in airway dimensions. The single functional breathing study showed increased simulated flow resistance after incisor retraction in East Asians. South Asian studies had mixed findings, with some reporting significant airway narrowing. The single study from West Asia found no significant airway or hyoid changes. Conclusions Airway response to bicuspid extractions and incisor retraction varied substantially when assessed with cephalometry. CT measurements present larger effect sizes and smaller variations, providing stronger evidence of airway narrowing. Orthodontic extractions for incisor retraction may be more frequently indicated in Asia, and East Asians seem particularly susceptible to airway narrowing and postero‐inferior hyoid movement with incisor retraction. Better designed CT studies are needed for confirmation due to small effect size and large variability.


| INTRODUC TI ON
Obstructive sleep apnoea (OSA) is a condition characterised by repeated collapse of the upper airway during sleep, leading to oxygen desaturations, persistent respiratory effort, arousals and sleep fragmentation. 1 It is defined by the occurrence of daytime sleepiness, loud snoring, witnessed breathing interruptions or awakenings due to gasping or choking in the presence of at least five obstructive respiratory events per hour of sleep (apnoea-hypopnea index [AHI] > 5). 2 The prevalence of moderate to severe OSA with AHI ≥ 15 is as high as 30%-50%, with the majority of subjects not diagnosed. [3][4][5] Severe OSA is associated with increased mortality, cardiovascular diseases, stroke, diabetes, motor vehicle accidents, cognitive impairments and reduced quality of life. 6 Obstructive sleep apnoea is a heterogeneous disorder, with obesity, age, oropharyngeal and facial anatomy, 7 as well as nonanatomical and functional factors such as neuromuscular feedback and airway collapsibility playing pathogenic roles in OSA. [8][9][10] Anatomic factors are important contributors and have been correlated to OSA severity. [11][12][13][14] Some clinicians have suggested that tooth extractions predispose patients to OSA. The proposed mechanism is a reduced arch depth in the sagittal plane resulting in decreased oral cavity volume and posterior displacement of the tongue and soft palate. The reduction in arch depth may be more significant in certain skeletal types, particularly Class II subtypes, and the decrease in airway space may lead to possible aggravation of snoring and OSA. 15,16 Reopening of closed orthodontic extraction spaces was even recommended to resolve OSA. 17 Clinically, Fukuda et al 18 found higher AHI in orthodontic extraction patients compared with matched untreated controls.
Conversely, Larsen et al 19 found no difference in OSA prevalence between patients with orthodontic extractions and matched controls. As it is difficult to link orthodontic treatment performed in adolescence 20 with development of OSA in later adulthood, [21][22][23] changes in airway anatomy are often used as a proxy for OSA risk, as OSA severity is correlated to anteroposterior (A-P) airway dimension, cross-sectional airway area (CSA), pharyngeal airway length, hyoid bone position and airway resistance. [11][12][13][14] Decrease in airway space 24,25 and changes in hyoid bone position 25,26 after orthodontic extractions have been reported.
Conversely, other studies have found no change in airway space [27][28][29] and hyoid position 30 after orthodontic extractions. The lack of consensus could be attributed to differences in patient age and extraction indications. 31,32 Airway effects from orthodontic extractions in growing patients may be ameliorated by pharyngeal growth. 27,28,32 Different orthodontic mechanics can also have differing airway effects. 29,31,32 A prior systematic review 31

| MATERIAL S AND ME THODS
This systematic review was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines a . 33 The review was registered with the PROSPERO database (PROSPERO 2018 CRD42018102318) b .

| Search Strategy
Eight databases were systematically searched from their inception up to August 2018 (using the search terms detailed in Table 1

| Selection criteria
The following inclusion and exclusion criteria were defined a priori.

| Study types
Randomised clinical trials, quasi-experimental studies, prospective and retrospective cohort studies, case-control studies and case series were included, while all other study designs were excluded.

| Study language
Studies were restricted to those reported in the English language.

| Study participants
Studies where the subjects were above 16 years old were included.

| Study intervention
The intervention was orthodontic treatment with dual-arch bicuspid extractions plus upper and lower incisor retraction. The intervention must be accompanied by examination with two-dimensional (2D) or three-dimensional (3D) radiographic examination before and after orthodontic treatment or retraction of incisors. Studies with singlearch extractions or extractions without mention or measurement of incisor retraction were excluded. Studies with subjects undergoing growth modification or orthognathic surgery were also excluded, as these may produce airway changes independent of the extraction treatment. 34,35

| Study comparison
Treated subjects were compared with untreated controls or non-extraction controls where applicable.

| Study outcome measures
The outcome variables evaluated were as follows: • Linear upper airway measurements.
• Vertical and horizontal changes in hyoid bone position.
• Functional assessment of breathing.

| Data collection and synthesis
The titles and abstracts of identified studies were screened independently by two authors (NJH and SYL), followed by an independent checking of their full texts for eligibility by both authors. Any conflicts at either stage were resolved by full-text screening and moderation by a third author (YAU). Final decisions were made after consensus was reached.

| Assessment of methodological quality
The Joanna Briggs Institute's Critical Appraisal Checklist was used to assess methodological quality of the selected studies (Table 2).
This was assessed independently by two authors (NJH and SYL), and conflicts between them were resolved by the third author (YAU).

| Yield of search
The search strategy yielded a total of 1652 articles and abstracts, of which 441 were duplicates. Screening of the titles and abstracts of the remaining 1211 articles resulted in 24 articles selected for full-text assessment. However, the full texts of three articles were inaccessible. [36][37][38] After full-text appraisal, 12 articles were excluded due to the following reasons: 1. Full text not in English. [39][40][41][42] 2. Treatment group below 16 years old. 24,27,43,44 3. Single-arch extraction. 45,46 4. Incisor retraction not uniformly applied. 29 5. Unclear inclusion criteria, no email response from authors. 47 Nine eligible articles were selected for this systematic review and narrative synthesis ( Figure 1).

| Study characteristics
The selected studies were all from Asia and were divided by subregions based on United Nations' classification of macrogeographic subregions. 48 The studies were further organised into lateral cephalometric and computed tomography (CT) studies, with one study using both CT and a CT-derived midsagittal lateral cephalometric image for airway measurements. 49 All studies were uncontrolled before-after case series. Two of the studies 50,51 reported data from multiple patient groups, from which only the study group with inci-  Table 3.
Outcome measures and landmarks used in lateral cephalometric studies are shown in Table 4.

| Linear changes
All three East Asian lateral cephalometric studies reported linear airway narrowing in the A-P dimension with incisor retraction. This was reported at the retropalatal, 25,49,50 retroglossal 25,49 and hypopharyngeal levels. 25,49 No changes were seen at the level of the nasopharynx. Airway length was measured by only one study and was found to be increased after incisor retraction. 25 Of the three South Asian studies, one reported no significant changes in airway dimensions 51 while two studies showed linear dimensional reduction at the retropalatal 52 and retroglossal levels. 52,53 Nasopharyngeal airway dimensional increase was reported by one study and attributed to lymphoid mass regression. 53 No significant change in airway length was found.
The West Asian study 30 found no significant change in airway dimensions from anterior retraction and arch dimension reduction in the treatment of bimaxillary proclination.

| Cross-sectional changes
All three CT studies 26 54 Zhang et al 49 found no change in the CSA from incisor retraction, but reported a cross-sectional shape change with decreased A-P dimension but increased lateral width, which maintained the overall CSA for the airway.

| Volumetric changes
Two studies reported airway volume changes after incisor retraction. Zheng et al 54

| Airway changes in relation to incisor retraction
Changes in airway dimension with respect to incisor retraction were investigated by four out of the five East Asian studies 25

| Hyoid changes
One out of the five East Asian studies did not study hyoid bone changes. 54 Of the four that did, three studies reported an inferior movement of the hyoid, 25 The West Asian study 30 found no significant change in hyoid bone position.

| Functional measures of breathing
Flow resistance was reported by Zheng et al 54 and was ascertained by computational fluid dynamics on 3D reconstructed airway models. There was no significant change in nasopharynx resistance.
Airflow resistance was significantly increased by 87.43% at the oropharynx, 27.14% at the hypopharynx, and 78.14% across the entire airway with incisor retraction.
Changes in airway dimension, hyoid bone position and functional breathing are summarised in Table 5.

| General remarks
The effects of bicuspid extraction and incisor retraction on airway dimension, hyoid position and breathing of adults and late adolescents were systematically reviewed in the present work. The PRISMA guideline was adopted to improve reporting transparency. 33 The review was restricted to those published in English due to  30,49,54 None of the studies reported growth assessment prior to commencing orthodontic treatment, but one author clarified that cervical growth maturation staging and hand-wrist radiographs were used for growth assessment. 51 Skeletal growth has been reported to cease at an average age of 17.5 years for females and 19.2 years for males. 55 For the upper airway, the major growth phases have been reported to be from 0 to 5 years, 6 to 9 years and 12 to 16 years old. 56,57 Quiescence of airway growth has been noted from 9 to 12 years and 15 to 18 years. 56 However, airway size and length have also been reported to increase until age 20. 58   The single West Asian study 30 found no significant change in airway dimensions or hyoid bone position.

| Hyoid measurements
The majority of studies use H-MP, HH1 and H-RGN to measure vertical and horizontal changes in the hyoid bone. 25,30,49,52,53 These measurements, however, rely on mandible position, which may rotate backwards during the normal course of orthodontic treatment. 62 The use of a stable horizontal or vertical reference plane 26,50 or an independent landmark unaffected by orthodontic treatment 51 would provide more accurate changes in hyoid bone positions.

| Changes in relation to incisor retraction
Airway dimensional change was reported to be correlated to upper incisor retraction distance by two studies, 26

| Long-term changes
All the studies lacked long-term follow-up. Partial reversion of the hyoid bone position and partial re-establishment of airway dimensions twelve months after posterior surgical setback of the mandible has been reported. 63,64 Whether the same effect exists in orthodontic extraction cases is still unknown.

| Individual variability
Standard deviations often exceeded the magnitude of mean A-P linear changes in East Asian lateral cephalometric studies, suggesting that A-P dimensional reduction from incisor retraction was highly inconsistent. Standard deviations exceeding mean effect size was also found in South Asian studies 52 and in hyoid bone movements measured on lateral cephalograms. 25 East Asians, as a group, appear to experience a decrease in airway A-P linear dimension at the oropharynx and hypopharynx, a modest proportion of patients may be entirely unaffected or may even experience an increase in airway dimension.

| Comparison of lateral cephalogram and computed tomography
The results of CT studies appear to show a larger percentage airway reduction with smaller individual variation. Although lateral cephalometric airway measurements have been reported to be reliable, 59 2D radiographs may not accurately reflect the 3D structure of the airway. [66][67][68][69] The semi-automated quantitative software assessment of the airway may minimise measurement errors on CT, and oropharyngeal airway volume measurements on CT have excellent reliability. 70

| Changes in functional breathing
Almost all the studies measured only morphological changes. As studies about post-orthodontic airway narrowing are primarily concerned with an increase in OSA predisposition, the use of morphological change as a surrogate for respiratory function is not ideal. 31 Although a close relationship between pharyngeal narrowing, hyoid bone position and OSA has been reported, 14,71 airway narrowing may not uniformly increase predisposition to OSA for all patients as functional and non-anatomic aetiologies are an important factor in up to 56% of OSA cases. 9,10 Polysomnography (PSG) is the diagnostic reference standard for OSA, but it is impractical to perform pre-and post-orthodontic PSG due to access limitations. 72 Functional breathing is more closely associated with OSA severity than morphologic changes 73

| Geographic and racial differences
Although geographic region was not specified in the search protocol, all studies meeting the criteria originated from Asia. This may be attributed to the fact that orthodontic extractions are more common in Asian populations. 20,74 In addition, East Asian populations present more frequently with bimaxillary proclination and lip protrusion 75,76 that is orthodontically treated with premolar extractions and incisor retraction. 77

| Review level
The review was restricted to studies published in the English language and is therefore subject to language and possibly publication bias. 87 The inclusion of non-English language studies may, however, not significantly change the results of this systematic review. [88][89][90][91] From the preliminary literature review, the decision was made a priori not to limit the types of clinical studies. The majority of the clinical studies on this topic were case series, before-after studies and other uncontrolled or poorly controlled observational study designs.
Uncontrolled before-after studies are deemed as case series by the Cochrane network 92,93 and at risk of bias, 94,95 but can provide sufficient information to calculate treatment effects, although not relative risk. 96

| Study level
All selected studies were uncontrolled and observational in design and were at increased risk of bias. 95 All but two 26

| Recommendations for future work
Learning from the inadequacies of prior studies, future research in this area should incorporate: 1. Detailed reporting of racial demographics, age, growth status, gender, horizontal and vertical skeletal subtypes, gender, oropharyngeal soft tissues and other possible confounders, as well as the intervention received, such as extraction pattern.
2. CT imaging for airway assessment.
3. Use of stable reference points for hyoid positional change assessment.
4. Functional assessment of breathing including polysomnography.

Use of untreated or non-extraction matched controls imaged at
pre-and post-treatment. 6. Appraisal of both upper and lower incisor changes and correlating this to airway dimensional changes.
7. Long-term follow-up to monitor for adaptive reversions of airway dimensions.

| CON CLUS IONS
Within the limitations of this systematic review, the following conclusions could be made: 1. Linear airway response to incisor retraction measured on lateral cephalograms varied substantially, while linear, cross-sectional and volumetric measurements of posterior airway space using CT showed larger effect sizes and smaller variations, providing stronger evidence of airway narrowing with bicuspid extractions and incisor retraction.
2. Hyoid bone positional changes in response to bicuspid extractions and incisor retraction varied substantially.
3. Functional breathing response to bicuspid extractions and incisor retraction was not adequately studied.
4. Orthodontic extractions for incisor retraction may be more frequently indicated in Asia, and East Asians seem particularly susceptible to airway narrowing and postero-inferior hyoid movement with bicuspid extractions and incisor retraction.

5.
Better designed CT studies are needed before definitive conclusions can be drawn due to small effect size and large variability.

ACK N OWLED G M ENTS
This work was supported by the National Dental Centre of Singapore.
The study received no external funding to conduct this research through any of the authors involved.

CO N FLI C T O F I NTE R E S T
The authors have stated explicitly that there are no conflicts of interest in connection with this article.