Anterior open-bite (AOB) treatment is considered challenging because of difficulties in determining and addressing etiologic factors and the potential for relapse in the vertical dimension after treatment. In this review, we compiled evidence on the long-term stability of the major therapeutic interventions for correcting AOB. Our objective was to review and compile evidence for the stability of surgical and nonsurgical therapies for AOB malocclusion. Our data sources were PubMed, EMBASE, Cochrane Library, limited gray literature search, and hand searching.
A search was performed of the electronic health literature on the stability of AOB after treatment. Hand searching of major orthodontic journals and limited gray literature searching was also performed, and all pertinent abstracts were reviewed for inclusion. Full articles were retrieved for abstracts or titles that met the initial inclusion criteria or lacked sufficient detail for immediate exclusion. Studies accepted for analysis were reviewed and their relevant data retrieved for pooling. The long-term stability estimates were pooled into nonsurgical and surgical groups, and summary statistics were generated.
One hundred five abstracts met the initial search criteria, and 21 articles were included in final analyses. Rejected articles failed to exhibit follow-up times of 12 months or more, did not include measurements of overbite (OB), or did not meet inclusion criteria. All included articles were divided into a surgical group (SX) with a mean age of 23.3 years and a nonsurgical group (NSX) with a mean age of 16.4 years. All studies were case series. Random-effects statistical models were used to pool the mean OB measures before and after treatment and also at the long-term follow-up. The pretreatment adjusted means of OB were –2.8 mm for the SX and –2.5 mm for the NSX. AOB closures up to +1.6 mm (SX) and +1.4 mm (NSX) were achieved. Relapse in the SX group during the mean 3.5 years of follow-up reduced the OB to +1.3 mm; the NSX group relapsed to +0.8 mm in the mean 3.2 years of follow-up. Pooled results indicated reasonable stability of both the SX (82%) and NSX (75%) treatments of AOB measured by positive OB at 12 or more months after the treatment interventions.
In the included case series publications, success of both the SX and NSX treatments of AOB appeared to be greater than 75%. Because the SX and the NSX were examined in different studies and applied to different clinical populations, no direct assessment of comparative effectiveness was possible. The pooled results should be viewed with caution because of the lack of within-study control groups and the variability among studies.
Anterior open-bite (AOB) is historically considered a challenging malocclusion to treat, and its correction is prone to relapse. The etiology is complex, potentially involving skeletal, dental, respiratory, neurologic, or habitual factors. Open-bite treatment is usually targeted at obtaining a positive amount of overlap of the maxillary and mandibular incisors. Relapse or instability refers to the tendency for the AOB to recur after treatment; this can result in a decrease in incisor overlap or a frank return of interincisal space (negative overlap).
There is no consensus as to the optimal therapy for AOB. It might be treated dentally by moving the teeth in the alveolar bone and soft-tissue housing with interarch orthodontic mechanics. Behavior-modifying appliances might be indicated when digit sucking or interincisal tongue posture is identified. Attempts to gain stability of therapy have led to the evolution of dentoalveolar and surgical interventions. New techniques can involve minimally invasive osseous implant anchorage or extensive maxillomandibular repositioning surgery.
A systematic review and a meta-analysis seek to use existing evidence to produce an unbiased summary estimate of a quantifiable effect. A meta-analysis looks at the average effects of size and direction, precision, and the extent of differences between studies that can be explained by chance (heterogeneity). Even when the state of evidence is low or ambiguous, the technique provides a summary of current knowledge and can offer insight to direct future research.
Although there are many articles on the treatment of AOB in the orthodontic and surgical literature, most look only at the postintervention effects. Because of the propensity for relapse after any orthodontic treatment, it is important to look beyond the immediate posttreatment time point to assess long-term stability. Relapse after AOB treatment has been attributed to tongue posture, growth pattern, treatment parameters, and surgical fragment instability. Most skilled practitioners can obtain positive overlap of the teeth with orthodontic or surgical interventions, but retaining the vertical correction can be challenging once the appliances are removed.
The a priori objective of this study was to assess the scientific literature and compile the current state of the evidence for stability of surgical and nonsurgical therapies for AOB malocclusion.
Material and methods
Electronic searching was performed to identify all eligible studies for inclusion in the review and meta-analysis, according to criteria described below. A health sciences librarian was consulted, and wide electronic searches in PubMed (1949-May 2009), EMBASE (1988-May 2009), and the Cochrane Library were performed to return the greatest number of hits. Keywords used are shown in Table I . All languages were searched, and pertinent articles were translated and reviewed.
|Database||Key words||Results ∗||Selected||% of 21 total selected abstracts|
|PubMed||(open bite OR openbite) AND (recurr ∗ OR treatment outcome OR follow-up studies OR stability OR instabil ∗ OR retreat ∗ OR relaps ∗ )||389||21||100%|
|EMBASE||(open bite OR openbite) AND (recurr ∗ OR treatment outcome OR follow-up studies OR stability OR instabil ∗ OR retreat ∗ OR relaps ∗ )||73||4||19%|
|Cochrane Library||(open bite or openbite):ti,ab,kw AND (recurr ∗ OR treatment outcome OR follow-up studies OR stability OR instabil ∗ OR retreat ∗ OR relaps ∗ ):ti,ab,kw||18||2||10%|
Inclusion criteria for the final selection were (1) human subjects, (2) stability of outcome assessed at the posttreatment follow-up ≥1 year, (3) negative overbite (OB) or open-bite preintervention as defined by vertical measures, and (4) corrective therapy for open-bite malocclusion adequately described.
The exclusion criteria were (1) case reports with ≤5 subjects, (2) editorials or opinion or philosophy articles with no new data, (3) subjects with other craniofacial pathologies or anomalies potentially influencing stability or complicating treatment (syndromes, periodontal disease, cleft lip or palate, trauma), and (4) mixed measures of open bite (combining horizontal and vertical measures).
Two orthodontic experts (G.M.C. and J.C.) independently reviewed the list of titles and abstracts for inclusion. All articles that appeared to meet the inclusion criteria were reviewed, and differences were resolved by consensus. Hand searching was performed in the major journals in the field: American Journal of Orthodontics and Dentofacial Orthopedics and Angle Orthodontist . Hand searching of reference lists was also performed on included studies. A limited search of the gray literature (unpublished) was performed by using the University of Washington’s library of orthodontic theses. Articles with the same data set were combined and the most recent article reported. The last search was performed in April 2009.
No study with a long-term follow-up had a control group to demonstrate the efficacy of the intervention. All studies meeting the inclusion criteria were follow-up studies of a series of patients who received 1 form of treatment (surgical or nonsurgical) and therefore did not allow inferences about the comparative effectiveness of alternative treatments or comparisons with no treatment. Sample sizes were low, with 1 exception. Most studies did not describe their methodology for selecting subjects, and most did not address dropouts. A methodologic quality-assessment list was developed after the study of Nguyen et al by analyzing study design, study content, statistical analysis, and conclusions ( Table II ). Each study was scored by the same 2 investigators, and discrepancies were resolved by discussion and consensus. The maximum quality score possible was 20.
|Study design||Study conduct||Statistical analysis||Conclusion|
|Study||Population described (2)||Selection criteria (2)||Sample size (2)||Controls used (2)||Follow-up definition & length (2)||Dropouts mentioned (1)||Measurement defined (2)||Reliability/ error testing
|Appropriate statistics (1)||Confounders analyzed (2)||Presentation of data (2)||Reasonable conclusion for study power (1)||Total (20 possible)|
|Arpornmaeklong and Heggie||2||1||2||0||2||0||2||1||0||0||1||0||11|
|Denison et al||1||1||2||1||1||0||2||1||1||0||1||1||12|
|Ding et al||1||1||1||0||2||0||2||1||1||0||0||1||10|
|Espeland et al||2||2||2||0||2||0||2||1||1||0||1||1||14|
|Fischer et al||1||1||2||0||2||0||1||1||1||0||1||1||11|
|Hoppenreijs et al||2||1||2||0||2||0||2||1||1||2||1||1||15|
|Huang et al||2||1||2||1||2||0||2||1||1||0||0||1||13|
|Janson et al||2||1||2||1||2||0||2||1||1||1||2||1||16|
|Kahnberg et al||1||1||1||0||1||1||1||0||0||0||2||0||8|
|Katsaros and Berg||1||1||1||1||1||1||2||1||1||0||2||1||13|
|Kim et al||2||1||2||0||1||1||0||1||0||0||2||0||10|
|Kucukkeles et al||1||1||1||0||1||0||1||1||0||0||2||0||8|
|Lawry et al||1||1||1||0||1||0||0||1||0||0||1||1||7|
|McCance et al||1||1||1||0||1||0||0||1||0||0||2||1||8|
|Moldez et al||1||1||1||1||2||0||1||1||1||0||1||0||10|
|Nelson et al||2||0||1||0||1||0||2||0||1||0||1||1||9|
|Remmers et al||2||2||2||0||2||0||2||1||1||0||2||1||15|
|Sugawara et al||1||1||1||0||1||0||2||0||0||0||2||1||9|
|Swinnen et al||2||2||2||1||2||0||2||1||0||0||1||1||14|
Data collection forms and electronic spreadsheets were used for data abstraction. There were some variations in the methodology used to measure OB. Of the 16 studies included, 14 measured OB on cephalometric radiographs either perpendicular to the occlusal plane (10 studies) or from S’ to N (7° down from sella-nasion) (2 studies), and along the nasion-menton line (2 studies). Two articles used direct measurements on dental casts or patients to quantify OB. Three surgical articles did not report postintervention estimates or variances for OB. These data were calculated from reported change scores between study time points and the variance of the reported change used.
Despite the lack of high-level evidence, summarizing the available data with a forest plot has value. Mean OB data at each time point were pooled by using a random-effects model. Chi-square tests of homogeneity and the I statistic were computed to evaluate the heterogeneity of the included studies. The random-effects model partly accounted for the heterogeneity among the articles when estimating the precision of summary estimates by allowing for a distribution of the true parameter being estimated among studies, rather than assuming only 1 true parameter value; 95% CI values were calculated to indicate the precision of the pooled means.
A secondary analysis looked at dichotomous success as a percentage of stable subjects at the long-term follow-up. The percentages of reported patients with positive overlap at the longest follow-up point were calculated.
The search strategy returned 428 potential articles for inclusion. Table I outlines the search results, the number of studies selected for inclusion from each database, and the percentage contribution to the included articles.
No studies were identified with control or comparison groups so that standardized mean-difference statistics could be developed. Consequently, a meta-analysis using effect size as described by the Cochrane Collaboration could not be performed. Sixteen case-series studies identified in the search had reports allowing extraction of mean OB data for the preintervention condition (T1), the posttreatment result after therapeutic intervention (T2), and long-term stability follow-up (T3). These data were pooled to enable the primary evaluation of long-term surgical and nonsurgical open-bite treatment outcomes. Figure 1 is the flow diagram outlining the process leading to the included articles. Table III lists articles considered for inclusion but later rejected and the reasons for exclusion.
|Author and year||Limitations||Author and year||Limitations|
|Aarnes 1974||2||Kiliaridis et al 1990||1|
|Bailey et al 1994||2||Klocke et al 2002||3|
|Bazzucchi et al 1999||1||Kloosterman 1985||2|
|Beane 1999||1, 2||Kuroda et al 2007||1|
|Beckmann and Segner 2002||2||Kuster and Ingervall 1992||1|
|Bell et al 1977||2||Lello 1987||2|
|Bennett et al 1999||1||Lentini-Oliveira et al 2007||1, 2|
|Bishara and Chu 1992||1, 2||Lopez-Gavito et al 1985||3|
|Champagne 1992||1, 2||Lugstein and Mossbock 1988||2|
|Cinsar et al 2007||1||MacIntosh 1981||2|
|Cozza et al 2007||1||Martis 1980||2|
|Cozza et al 2006||1||McSherry et al 1997||2|
|Cozza et al 2005||1, 2||Meral and Yuksel 2003||1|
|Dattilo et al 1985||2||Meyer-Marcotty et al 2007||1|
|De Frietas et al 2004||4||Ng et al 2008||1, 2|
|Dellinger 1996||2||Nwoku 1974||2|
|Emshoff et al 2003||2||Oliveira and Bloomquist 1997||2|
|Epker and Fish 1977||2||Pedrin et al 2006||1|
|Erbay et al 1995||1||Proffit et al 2000||2|
|Ermel et al 1999||2||Reitzik et al 1990||2|
|Frankel and Frankel 1983||2||Reyneke and Ferritti 2007||2|
|Goncalves et al 2008||2||Rittersma 1981||2|
|Gottlieb et al 2006||1, 2||Schmidt and Sailer 1991||2|
|Greebe and Tuinzing 1987||2||Schrems and Schrems-Adam 1982||2|
|Hayward 1978||2||Seres and Kocsis 2008||1|
|Hoppenreijs et al 2001||4||Shpack et al 2006||3|
|Hoppenreijs et al 1996||2, 4||Spens 1981||2|
|Iannetti et al 2007||2||Steiner and Gebauer 1985||3|
|Iscan et al 2002||1||Stella et al 1986||2|
|Janson et al 2008||1||Teuscher et al 1983||2|
|Janson et al 2003||4||Torres et al 2006||1|
|Johanson et al 1979||2||Turvey et al 1976||2|
|Joos et al 1984||2||Turvey et al 1988||2|
|Kahnberg and Widmark 1988||2|
Quality scores for the studies meeting the inclusion criteria were relatively low and as a whole averaged 10.3 quality points of a possible 20. Most articles had shortcomings in the reporting of subject selection methods and dropouts, the analysis of confounders, and the lack of controls or comparison groups. Table II lists the quality scores for the articles included in the primary and secondary analyses.
The intervention to close AOB was the primary discriminator used to divide the included articles into the 2 samples. Subjects in the surgical group all had maxillary impaction surgery, with 7 studies reporting mandibular surgery as well. Nearly all patients in the surgical studies had presurgical orthodontic treatment, with the largest study reporting that 64 of 267 patients had no orthodontic treatment before surgery. The subjects in the nonsurgery group all had fixed appliance therapy with or without appliances for anteroposterior correction (headgear or functional appliances). Five of the nonsurgical studies explicitly stated that vertical elastics were used to close the bite, 3 articles included patients with extractions, and 2 stated a recommendation for speech or myofunctional therapy.
The mean and standard deviation for the subjects’ ages in the surgical studies were 23.3 ± 1.6 years. This postadolescent age was not unexpected, since orthognathic surgery is not routinely performed on skeletally immature patients, and adults with open bite are more likely to have surgery recommended. The mean age of the nonsurgical group was 16.4 ± 4.9 years, more typical of most orthodontic patients. Three studies in the nonsurgical group did include adults, so this sample was not entirely composed of growing subjects. This fact is partly adjusted for in examining median age, which was the same as the mean age in the surgical group at 23.3 years, but only 13.5 years for the nonsurgical group. Both surgical and nonsurgical samples were predominantly female at 71.7% and 75.4%, respectively. Seven of the 11 surgical studies reported presurgical cephalometric values, rather than pretreatment values.
The mandibular plane angle (SN-MP) was examined as an indication of case difficulty. Average presurgical SN-MP for the surgical studies was 42.9°; pretreatment SN-MP for the nonsurgical studies was 39.2°. Both values were higher than the general population average, indicating challenging treatment.
Tables IV and V outline data abstracted from studies included in the primary analysis.
|Study||T1 (n)||Female (%)||Age (y) at T1||SD age (y)||T1 SN-MP (°)||SD T1 SN-MP||T1 OB mean (mm)||T1 SD||T2 OB mean (mm)||T2 SD||T3 F/U time (y)||T3 OB mean (mm)||T3 SD||Open bite measured||Intervention||Surgical fixation|
|Lawry et al||19||68.4||21.40||35.0||5.1||−2.53 ∗||1.50||1.89||1.83||1.54||2.71||0.98||Cephalogram, possibly parallel occlusal plane||Maxillary impaction ± BSSO, fixed appliances||Wire, rigid|
|McCance et al
|10||NR||46.9||3.9||−4.60 ∗||4.90||−1.60||1.90||1.00||−1.70||1.90||Cephalogram, possibly parallel occlusal plane||Maxillary impaction ± BSSO, fixed appliances||Wire, rigid|
|McCance et al
|Kahnberg et al||19||57.9||23.30||−4.90 ∗||2.30||1.40||0.60||1.50||1.10||1.00||Clinical measurement in mouth or casts||Maxillary impaction, fixed appliances||Wire, rigid|
|Hoppenreijs et al||259||78.7||23.60||−1.24 ∗||2.48||1.86 †||2.74 †||5.75||1.24||1.45||Cephalogram, parallel to S’-N||Maxillary impaction ± BSSO, fixed appliances in 76%||Wire, rigid|
|Arpornmaeklong and Heggie Mx surgical only||17||76.5||21.40||42.5||7.6||−1.70 ∗||1.20||0.90 †||1.80 †||2.00||0.90||1.80||Cephalogram, parallel to occlusal plane||Maxillary impaction ± BSSO, fixed appliances||Rigid|
|Fischer et al||58||69.0||23.00||45.7||7.2||−0.80||2.80||1.30||1.10||2.00||0.80||1.40||NR, possibly parallel occlusal plane||Maxillary impaction ± BSSO, fixed appliances||Wire, rigid|
|Moldez et al Mx impaction||13||84.6||46.8||5.5||−2.20||2.10||2.00||0.90||5.00||1.10||0.90||Cephalogram, parallel to occlusal plane||Maxillary impaction ± BSSO, fixed appliances||Wire, rigid|
|Moldez et al Mx rotation||10||70.0||44.5||7.2||−3.80||1.70||2.20||0.70||5.00||1.70||0.80|
|Ding et al||10||80.0||24.42||41.4||5.2||−3.20||2.10||1.60||0.68||15.00||1.50||0.90||Cephalogram, parallel to occlusal plane||Maxillary impaction ± BSSO, fixed appliances||Wire, rigid|
|Espeland et al||40||60.0||25.80||9.50||39.9||7.1||−2.60 ∗||1.70||1.50 †||3.60 †||3.00||1.10||1.20||Cephalogram, parallel to S’-N||Maxillary impaction, fixed appliances||Rigid|
|Study||T1 (n)||Female (%)||Age (y) at T1||SD age (y)||T1 SN-MP (°)||SD T1
|T1 OB mean (mm)||T1 SD||T2 OB mean (mm)||T2
|T3 F/U time (y)||T3 OB mean||T3
|Open bite measured||Intervention|
|Nelson and Nelson||23||73.9||19.90||−2.60||1.20||1.40||1.20||2.00||−0.10||1.60||Vertical average distance between 4 incisors||Fixed appliances, elastics, possibly speech therapy|
|Katsaros and Berg||20||85.0||11.80||2.50||39.0||5.8||−1.90||1.80||1.20||2.00||2.00||1.20||1.80||Cephalogram, perpendicular to Na-Me line||Fixed appliances, functionals, extractions|
|Kucukkeles et al||17||70.6||19.35||40.5||5.8||−4.05||2.92||1.75||1.16||1.00||0.50||1.76||Cephalogram, parallel to occlusal plane||Fixed appliances, elastics, reverse curve of Spee archwires|
|Kim et al growing||29||72.4||13.50||1.92||37.7||4.6||−2.27||2.10||1.41||0.75||2.00||1.18||1.01||NR, possibly parallel occlusal plane||Fixed appliances, elastics, reverse curve of Spee archwires|
|Kim et al nongrowing||26||80.8||26.08||2.25||39.7||6.5||−2.23||2.10||1.90||0.57||2.00||1.55||1.09|
|Sugawara et al||9||77.8||19.30||40.1||2.1||−2.80||1.80||2.10||0.80||1.00||1.20||0.80||Cephalogram, parallel to occlusal plane||Fixed appliances, miniplate anchors|
|Janson et al nonextraction||21||76.2||12.40||36.9||5.7||−1.75||0.66||1.43||0.50||5.22||0.07||0.62||Cephalogram, parallel to occlusal plane||Fixed appliances, elastics|
|Janson et al ext||31||74.2||13.22||39.1||4.2||−2.73||1.80||1.09||0.94||8.35||1.02||1.62||Fixed appliances, elastics, extractions|
|Remmers et al||52||67.3||12.40||2.70||40.9||6.1||−3.20||1.90||0.40||1.10||5.00||0.20||1.80||Cephalogram, perpendicular to Na-Me line||Fixed appliances, elastics, headgear, functionals, extractions|