Meta-analysis on the mandibular dimensions effects of the MARA appliance in patients with Class II malocclusions
To evaluate the short- and long-term effects of the mandibular anterior repositioning appliance (MARA) on mandibular dimensions in patients with Class II malocclusion and to assess the stability of the MARA results. Multiple electronic databases were searched for articles published in any language until March 2014. A manual search was also performed of reference lists of retrieved articles. The primary outcomes were the short-and long-term effects of the MARA appliance on mandible dimensions. The secondary outcome was postretention stability. Outcome measures were total mandibular unit length, corpus length, and ramus height. Two reviewers examined all articles independently and assessed their methodologic quality. Meta-analyses were conducted using random-effects models. The Cochrane test and the I2 statistic were used to assess heterogeneity. Sensitivity analyses were performed and publication bias was evaluated. Seven retrospective clinical controlled studies that compared MARA with controls were included. Three of the studies were medium quality; the rest were low quality. Meta-analysis of the short-term effects revealed a significant increase in total mandibular unit length (1.16 mm/y) and ramus height (1.58 mm/y) with MARA and a nonsignificant increase in corpus length (0.21 mm/y). Analyses of the long-term effects showed a statistically significant advantage of MARA over controls for all three variables, but the effect sizes were small. More high-quality studies are warranted. The MARA appliance produced statistically significant mandibular growth enhancement in the short- and long-term. These findings, however, may not be clinically significant.ABSTRACT
Objectives:
Materials and Methods:
Results:
Conclusions:
INTRODUCTION
Class II malocclusion is a prevalent form of orthodontic discrepancy that is commonly associated with mandibular retrognathia.1 A number of functional appliances are available to treat Class II malocclusions in growing subjects. Some are removable and others are fixed.2,3 The goal of most of these appliances is to stimulate mandibular growth by posturing the mandible forward into a Class I occlusion. The expectation is that the condylar processes will remodel superiorly and posteriorly in the condylar fossae.2
One of the many functional appliances available today is the mandibular anterior repositioning appliance (MARA). This appliance corrects a Class II malocclusion into a Class I by displacing the mandibular condyles anteriorly and inferiorly, thereby resulting in remodeling of the temporal fossae and the condyles.4
The efficacy of the MARA on mandibular growth has been evaluated in numerous studies, but contradictory results have been reported.5–7 Some studies found significant mandibular dimensional changes with MARA,6,8 whereas others found no significant differences between MARA and untreated control subjects with Class II malocclusion.5,9 Therefore, the objective of this review was to evaluate the short- and long-term effects of the MARA appliance on mandibular dimensions in growing patients (10–16 years old) and to assess the postretention stability of the MARA results.
MATERIALS AND METHODS
Study Inclusion/Exclusion Criteria
The PICO (patient, problem, or population; intervention; comparison; outcome) criteria were used to determine whether a study should be included or excluded (Table 1). Studies considered for inclusion were randomized controlled trials (RCTs), prospective controlled clinical trials, and retrospective controlled clinical trials on humans with no restriction on the sample size. Case reports, case series, review articles, editorials, and commentaries were all excluded. The primary outcomes were the short-term (measured after removal of the MARA functional appliance) and the long-term (measured after the completion of the comprehensive fixed orthodontic treatment) effects of the MARA on the mandibular dimensions. The secondary outcome was the postretention stability of the changes obtained with the MARA (measured at least 1 year postretention). Outcome measures were total mandibular unit length, corpus length, and ramus height. Only studies that used the cephalometric landmark condylion were included. Studies were excluded if the measurements used the constructed point articulare.
Data Sources
Multiple electronic databases were searched for articles published in any language until March 2014, including Ovid MEDLINE, EMBASE, Database of Abstracts of Reviews of Effectiveness (DARE), Educational Resources Information Center (ERIC), Cumulative Index to Nursing & Allied Health (CINAHL), Cochrane Controlled Trials Register, Cochrane Database of Systematic Reviews, PubMed, and the World Wide Web using Google Scholar. The System for Information on Grey Literature in Europe (SIGLE) was searched for grey literature. Medical subject headings (MeSH) or text words were used depending on the database (Table 2). A manual search was also performed of theses and dissertations as well as reference lists of retrieved articles. Two reviewers carried out the search for the studies independently and in duplicate. Full articles underwent quality assessment. Disagreements were resolved through discussion.
All articles were reviewed, and data were extracted using a customized data abstraction sheet. The quality of the studies was independently assessed using the Cochrane collaboration's tool for assessing risk of bias.10 The criteria for the assessment were (1) random sequence generation, (2) allocation concealment, (3) blinding of outcome assessors (blinding of participants was not possible), (4) completeness of outcome data, (5) evaluation of selective reporting, and (6) no other sources of bias. The possible responses for each criterion were yes, unclear, or no. For this study, the score categories were two points for “yes,” one point for “unclear,” and no point for “no.” The quality of the studies was considered low if it received 1–4 points, medium if it received 5–8 points, or high if it received 9–12 points. Disagreement was resolved through discussion.
Statistical Analysis
Meta-analyses were conducted for the continuous outcomes assessed in the studies. The sample size, the mean difference (MD), and standard deviation of the difference (SD diff) of the changes in total mandibular unit length, corpus length, and ramus height were all obtained for MARA and controls from each study. Two studies5,9 did not report the standard deviations of the mean changes; thus, they were estimated on the basis of the MD, sample size, and P values as explained in the Cochrane Handbook for Systematic Reviews of Interventions.10
Review Manager software, version 5.2 was used to analyze the data.11 The statistical effect size was calculated using weighted MD (WMD) and 95% confidence Interval (CI); studies were combined using the inverse variance random-effects models method. Effect size was considered significant if P < .05. Heterogeneity between the studies was assessed using the Cochrane test (P < .1 was considered significant) and the I2 statistic.12 An I2 cutoff of 25%, 50%, and 75% was used to represent low, moderate, and high heterogeneity, respectively.12,13 Sensitivity analyses were performed to assess the robustness of the meta-analyses based on study quality, sample size, and matching by skeletal maturational age. Publication bias was assessed using the Egger's linear regression,14 using the Begg and Mazumdar15 rank correlation, and by visually checking the symmetry of the funnel plots. A symmetrical funnel plot and nonsignificant two-tailed Egger's and Begg's tests indicate no publication bias.
RESULTS
The search identified 37 relevant articles. After exclusion at the title and abstract stages, 10 studies underwent qualitative assessment. Three were excluded because one was not available,16 one did not use the landmark condylion,17 and one used a dissimilar follow-up period for the MARA and controls.18
Therefore, only seven retrospective clinical controlled studies met the inclusion criteria.5,6,8,9,19–21 The participants were growing subjects with Class II division 1 malocclusion in all studies except for one in which the type of Class II malocclusion was not specified.6 The age ranges were 10.11–12.3 years for the MARA subjects (N = 180) and 11.63–11.9 years for the control subjects (N = 126). One study20 did not specify the age range or the sample size for the control group. Descriptions of the included studies are presented in Table 3.
One of the studies evaluated the effects of treatment timing with MARA and fixed appliances in patients with Class II malocclusions compared with untreated historical controls.20 They evaluated patients at the prepubertal, pubertal and postpubertal skeletal developmental stages before treatment (T1) and after comprehensive treatment (T2). For the purpose of this meta-analysis, only the group treated during the peak growth spurt was included. Four other studies5,8,9,21 included subjects who were treated at the beginning of or during the growth spurt.
Outcomes assessed were the short- and long-term effects of MARA on mandibular dimensions (Table 4). Outcome measures included the mean changes in total mandibular unit length using Co-Gn; corpus length using Go-Gn, Go-Me, and Go-Pg; and ramus height using Co-Go. No studies were found for the postretention stability outcome. Overall mean treatment duration ranged from 10.7 to 18 months for MARA and from 27 to 54 months for the total comprehensive orthodontic treatment/observation period. The MARA treatment protocols reported in the studies included MARA alone, MARA with concurrent 2 × 4 fixed appliances, or MARA with full upper and lower fixed orthodontic appliances bonded at the start of the treatment.
Three of the studies8,9,21 were medium quality; the rest were low quality. Data from all the studies were used in the quantitative analyses of the short- and long-term effects on mandibular dimensions. All values were annualized to account for the different mean treatment times.
Short-Term Effects of MARA
Changes in total mandibular unit length
Six studies5,6,8,9,19,21 evaluated this outcome using the cephalometric variable Co-Gn. The meta-analysis showed a 1.16 mm/y increase in the total mandibular length of the MARA group over controls (WMD = 1.16; 95% CI = 0.61, 1.72; χ2 test = 7.32; 5 df; P = .20; I2 = 32%; test for overall effect, Z = 4.11 and P ≤ .0001) (Figure 1). Sensitivity analysis excluding the three low-quality studies5,6,19 showed comparable findings (WMD = 1.12; 95% CI = 0.60, 1.63) (Table 5).
Citation: The Angle Orthodontist 85, 4; 10.2319/052814-378.1
Changes in mandibular corpus length
Different cephalometric variables were used among the studies. Two of them used Go-Me5 and Go-Po,6 respectively; the other two studies8,9 used Go-Gn. Statistical pooling of the four studies showed no significant differences between the groups and a medium heterogeneity (WMD = 0.21; 95% CI = −0.16, 0.57; χ2 test = 9.73; 3 df; P = .02; I2 = 69%; test for overall effect, Z = 1.11 and P = 0.27). Analysis of the two studies8,9 that used the same cephalometric measurements showed a similar effect size.
Changes in mandibular ramus height
Five studies5,6,8,9,21 assessed the MD in the ramus height between MARA and controls using Co-Go. The difference was statistically significant favoring MARA (WMD = 1.58; 95% CI = 1.12, 2.05; χ2 test = 4.75; 4 df; P = .31; I2 = 16%; test for overall effect, Z = 6.69 and P ≤ .0001). A similar finding was observed after excluding two low-quality studies5,6 (WMD = 1.59; 95% CI = 0.99, 2.19).
Evidence of publication bias was only noted for the ramus height outcome measure. The funnel plot was asymmetrical, and the Egger's and Begg's tests were statistically significant (P = .018 and P = .03, respectively). The Trim and Fill method was used to trim two studies to obtain symmetry in the funnel plot and the Random effect estimate was reduced to 1.37 (95% CI = 0.89, 1.83).
Long-Term Effects of MARA
Changes in total mandibular unit length
This outcome was assessed in five studies.5,8,9,20,21 The MARA produced a statistically significant increase in the total mandibular length (WMD = 0.83; 95% CI = 0.59, 1.07; χ2 test = 4.24; 4 df; P = .34; I2 = 6%; test for overall effect, Z = 6.75 and P ≤ .0001) (Figure 2). A similar result was revealed when one low-quality study5 and one medium-quality study with a small sample size20 were excluded from the analysis (WMD = 0.59; 95% CI = 0.34, 0.83) (Table 6).
Citation: The Angle Orthodontist 85, 4; 10.2319/052814-378.1
Changes in mandibular corpus length
This outcome was evaluated in three studies5,8,9 that used different cephalometric measurements. The analysis showed a small advantage of MARA over controls (WMD = 0.61; 95% CI = 0.16, 1.06; χ2 test = 6.62; 2 df; P = .04; I2 = 70%; test for overall effect, Z = 2.65 and P = .008). Sensitivity analysis using two studies8,9 revealed similar findings (WMD = 0.82; 95% CI = 0.52, 1.12).
Changes in mandibular ramus height
Four studies5,8,9,21 assessed the ramus height, and a 0.7 mm/y difference was detected (WMD = 0.70; 95% CI = 0.30, 1.11; χ2 test = 3.37; 3 df; P = .34; I2 = 11%; test for overall effect, Z = 3.40 and P = .0007). Sensitivity analysis excluding one5 study revealed a 0.76 mm/y difference (WMD = 0.76; 95% CI = 0.22,1.31).
No evidence of publication bias was detected for any of the long-term mandibular growth outcome measures.
DISCUSSION
The aims of this review were to evaluate the short- and long-term effects of MARA on mandibular dimensions in patients with Class II malocclusions and to assess the postretention stability of the MARA results.
All the studies that met the inclusion criteria were of low or medium quality. In several studies, the sample was not randomly selected or the random selection method was unclear. Only one study19 used a concurrent untreated control group; the rest used historical controls. Historical controls might not be similar to the MARA group in the growth pattern. Although five studies5,8,9,20,21 matched the MARA and control subjects by the cervical vertebral maturational age, two studies6,19 used only chronological age and dentoskeletal characteristics to match the samples. All but three of the studies reported similar baseline skeletal features among the MARA and the control samples. One6 reported a significant difference in the mean total mandibular unit length between the study groups, but the other two20,21 did not discuss the results of the baseline cephalometric measurements.
Short-Term Effects of MARA
This review found a total mandibular length increase of 1.16 mm/y with MARA compared with control subjects with untreated Class II malocclusion. The effect size was sustained (1.11 mm/y) when two low-quality studies6,19 were excluded and the associated heterogeneity was eliminated, which confirms a functional orthopedic effect of this appliance. A previous meta-analysis22 evaluated the effectiveness of functional appliances on mandibular growth in the short term. In that study, 32 RCTs fulfilled the inclusion criteria for the review, but only four were selected for the quantitative analysis. Annualized results using the random-effect model revealed statistically significant mandibular growth of 1.79 mm over untreated control subjects. The high quality of studies included in the previous meta-analysis may explain the 65% difference in the increase of total mandibular length between the reviews.
Johnston23 reported that nearly 4 to 6 mm of molar correction is needed to treat a Class II malocclusion. Thus, the 1.16 mm/y mandibular growth enhancement in this review contributes only partially to the correction of the Class II malocclusion.
Statistical pooling of the MDs of the mandibular corpus length showed a nonsignificant increase with MARA and a great heterogeneity. The small number of included studies and the differences in methodologies (eg, cephalometric measurements, treatment durations, and treatment mechanics) can all explain the heterogeneity.
Ramus height increased by 1.58 mm/y in the MARA group compared with the untreated control subjects. This finding suggests that the total mandibular dimensional change was more a result of mandibular vertical development than horizontal growth. Although the MARA appliance includes stainless steel crowns that cover both the maxillary and mandibular first molars, which might restrict the natural eruption of these teeth, this did not seem to affect the skeletal elongation of the ramus.
Long-Term Effects of MARA
The increase in total mandibular unit length, corpus length, and ramus height were all statistically significantly larger in the MARA group than in the control group, but the effect sizes were small and clinically insignificant if analyzed independently of other changes. The annualized mean differences achieved in the short term diminished over time as the sample surpassed the circumpubertal growth spurt.
Clinical Significance
In summary, this meta-analysis found a statistically significant mandibular growth enhancement with MARA in both the short and long term, but the findings may not be clinically significant. In a recent study, Brignardello-Petersen et al.24 suggested that for a result to be clinically significant, three criteria must be met: (1) the change/difference in the outcome between the groups has to be of interest to clinicians, (2) the change/difference must occur in an important outcome, (3) and the findings must be statistically significant. In this meta-analysis, the outcomes assessed are important and the changes in the mandibular growth are statistically significant. However, the pooled effect sizes are very small, when considered individually, to affect clinical decision-making.
Limitations
This meta-analysis was constrained by several limitations. The quality of the included studies was low or medium, therefore limiting the external validity of the results. Also, the varying methodologies and MARA treatment protocols may have reduced the homogeneity of the studies and weakened the pooled effect sizes. Future studies with strong designs (eg, RCTs) and rigorous methodologies (eg, similar dentoskeletal features at baseline, untreated control subjects with Class II malocclusion, and blinding of outcome assessors) are warranted. None of the studies evaluated the postretention stability effects of MARA. It is recommended that future studies evaluate all possible outcomes for changes in mandibular growth.
CONCLUSIONS
-
The MARA appliance produced statistically significant mandibular dimensional changes. These findings, however, are unlikely to be of clinical significance when analyzed individually.
-
Meta-analyses of the short-term effects revealed a significant increase with MARA over untreated control subjects in total mandibular unit length (1.16 mm/y), ramus height (1.58 mm/y), and a nonsignificant increase in corpus length (0.21 mm/y).
-
Statistical pooling of the effects after use of fixed appliances showed a statistically significant advantage of MARA over controls in total mandibular unit length (0.83 mm/y), corpus length (0.61 mm/y), and ramus height (0.70 mm/y).
Forest plots of MARA vs controls for the short-term mandibular dimensions outcome.
Forest plots of MARA vs controls for the long-term mandibular dimensions outcome.
Contributor Notes