The authors have declared that no competing interests exist.
Conceived and designed the experiments: WZ YZ. Performed the experiments: WZ HCQ. Analyzed the data: MY YZ. Contributed reagents/materials/analysis tools: HCQ. Wrote the paper: WZ HCQ.
We conducted a comprehensive meta-analysis of 12 studies to examine whether maxillary protraction face mask associated with rapid maxillary expansion (FM/RME) could be an effective treatment for Class III malocclusion and to evaluate the effect of timing on treatment response. Patients with a maxillary deficiency who were treated with FM with or without RME were compared with those who had an untreated Class III malocclusion. In both treatment groups, forward displacement of the maxilla and skeletal changes were found to be statistically significant. In addition, posterior rotation of the mandible and increased facial height were more evident in the FM group compared with the control group. However, no significant differences were observed between the early treatment groups and late treatment groups. The results indicated that both FM/RME and FM therapy produced favorable skeletal changes for correcting anterior crossbite, and the curative time was not affected by the presence of deciduous teeth, early mixed dentition or late mixed dentition in the patient.
Mandibular or mandibular dentition prognathism, retrusive maxillary or maxillary dentition, and combinations of these components may lead to a Class III malocclusion[
Current non-surgical treatment methods for severe skeletal Class III malocclusions to correct maxillary discrepancies in young adolescents include rapid maxillary expansion (RME)[
We used [‘face mask’ or ‘FM’], [‘face mask/rapid maxillary expansion’ or ‘FM/RME’ (text word)], [Class III malocclusion or Angle Class III (MeSH)] and [‘maxillary protractor’] as search terms. The wide electronic search scope included PubMed, Cochrane Library, Web of Science, Springer Link, and ScienceDirect. In addition, we searched all these databases to avoid missing relevant studies published before October 6, 2014. We also evaluated studies that were cited in the reference lists of the included papers to ensure the inclusion of all relevant studies.
The publications had to reach the following standards to meet the strict inclusive criteria: i) the study concentrated on the treatment efficacy of FM or FM/RME and the relationship between timing factors and maxillary protraction; ii) all patients had clinical Class III malocclusion from the period of early mixed dentition to early permanent dentition, and their ages ranged from seven to fourteen years old; iii) the study provided the original data, or we were able to obtain the data from the primary data; and iv) the study was a case-control study or a randomized controlled trial (RCT). Moreover, the language of all included studies had to be English. We required complete, accurate, and useful data; consequently, reviews, abstracts, conference papers, case reports and letters were excluded without consideration.
We extracted information from the included research, such as author names, publication year, volume and issue; article design; number of cases and placebos, efficacy and safety assessment. Wei Zhang and Hong-Chen Qu independently checked the data from all the included studies. Subsequently, a third reviewer (Yang Zhang) discussed inconsistent evaluations and thereby helped to reach a final agreement.
Each publication’s quality was assessed by two reviewers (Wei Zhang and Hong-Chen Qu) according to a modified STROBE quality score system [
To acquire reliable and accurate results, two authors (Mo Yu and Yang Zhang) who were not involved in the data collection were in charge of extracting the data. The authors calculated the mean difference (MD) and 95% CI using Review Manager Version 5.3 software (provided by the Cochrane Collaboration). The I2 test was used to quantify the effect of heterogeneity. A higher result on the I2 test represented an increased possibility that heterogeneity contributed to the inter-study variability. Both fixed-effects and random-effects models were used: if the I2 test < 50% or P ≥ 0.05 (Q-test), we used the fixed-effects model; if there was significant heterogeneity among the included studies (I2 test > 50%), the random-effects model was employed. We used Funnel plots to detect publication bias; a symmetrical plot indicated little publication bias.
Based on the inclusion criteria, we included 12 satisfactory studies[
In this meta-analysis, 12 studies were selected for qualitative analysis.
Author | Year | Number | Study design | Quality score |
---|---|---|---|---|
Gencer | 2014 | FM group(n = 15); Control(n = 15) | Case-Control | 28 |
Chen | 2012 | FM group(n = 22); Control(n = 17) | Case-Control | 24 |
Ucem | 2004 | FM group(n = 14); Control(n = 14) | Case-Control | 25 |
Kilicoglu | 1998 | FM/RME group(n = 16); Control (n = 10) | Case-Control | 23 |
Masucci | 2011 | FM/RME group(n = 22); Control (n = 16) | Case-Control | 26 |
Sar | 2011 | FM/RME group(n = 15); Control (n = 15) | Case-Control | 24 |
Yuksel | 2001 | FM/RME group(n = 34); Control (n = 17) | Case-Control | 26 |
Kajiyama | 2000 | FM/RME group(n = 29); Control (n = 25) | Case-Control | 22 |
Lee | 2010 | Early treatment group (n = 26); Late treatment group (n = 23) | Case-Control | 27 |
Franchi | 2004 | Early treatment group (n = 33); Late treatment group (n = 14) | Case-Control | 24 |
Kajiyama | 2004 | Early treatment group (n = 63); Late treatment group (n = 57) | Case-Control | 24 |
Baccetti | 2000 | Early treatment group (n = 32); Late treatment group (n = 28) | Case-Control | 25 |
FM: face mask therapy. FM/RME: face mask and rapid maxillary expansion therapy
This meta-analysis demonstrated the difference in FM-treated Class III malocclusion patients and controls (
The comparison was performed using five indices. 1: Angle SNA. The FM treatment group presented a greater increase in SNA than the controls (SMD = 1.78, 95% CI = 1.57–1.99, P < 0.00001). 2: Angle SNB. The FM treatment group presented a greater decrease in SNB than the untreated controls (SMD = -1.75, 95% CI = -2.28–-1.23, P < 0.00001). 3: Angle ANB. The FM treatment group presented a greater increase in ANB than the controls (SMD = 3.64, 95% CI = 3.10–4.19, P < 0.00001). 4: ANS-Me length. The FM treatment group presented a greater increase in ANS-Me length than the controls (SMD = 2.92, 95% CI = 2.61–3.23, P < 0.00001). 5: Angle SN/GoGn. The FM treatment group presented a greater increase in SN/GoGn than the controls (SMD = 1.67, 95% CI = 0.63–2.71, P = 0.002). The FM groups exhibited significant improvement in skeletal retrognathism of the maxilla.
FM/RME-treated Class III malocclusion patients and untreated controls were also compared according to the five most revealing parameters. There was a significant improvement of the sagittal skeletal index in the FM/RME groups (
The comparison was performed using five indices. 1: Angle SNA. The FM treatment group presented a greater increase in SNA than the controls (SMD = 1.39, 95% CI = 0.85–1.94, P < 0.00001). 2: Angle SNB. The FM treatment group presented a greater decrease in SNB than the untreated controls (SMD = -2.54, 95% CI = -3.08–-2.01, P < 0.00001). 3: Angle ANB. The FM treatment group presented a greater increase in ANB than the controls (SMD = 3.25, 95% CI = 2.06–4.44, P < 0.00001). 4: ANS-Me length. The FM treatment group presented a greater increase in ANS-Me length than the controls (SMD = 2.08, 95% CI = -0.21–4.36, P = 0.07). 5: Angle SN/GoGn. The FM treatment group presented a greater increase in SN/GoGn than the controls (SMD = 3.26, 95% CI = 2.34–4.18, P < 0.00001). The FM/RME groups exhibited significant improvement in skeletal retrognathism of the maxilla.
A comparison of the early treatment and late treatment groups is shown in
The comparison was performed using five indices. 1: Angle SNA. The FM treatment group presented a greater increase in SNA than the controls (SMD = 1.09, 95% CI = -0.70–2.88, P = 0.23). 2: Angle SNB. The FM treatment group presented a greater decrease in SNB than the untreated controls (SMD = -1.42, 95% CI = -1.95–-0.90, P < 0.00001). 3: Angle ANB. The FM treatment group presented a greater increase in ANB than the controls (SMD = 1.72, 95% CI = -0.76–4.19, P = 0.17). 4: ANS-Me length. There were no significant differences in ANS-Me length between the early treatment group and the late treatment group (SMD = 0.50, 95% CI = -1.87–2.86, P = 0.68). 5: SN/GoGn angle. There were no significant differences in SN/GoGn angle between the early treatment group and the late treatment group (SMD = 0.5, 95% CI = -0.14–1.14, P = 0.13). 6: Co-Gn length. There were no significant differences in Co-Gn length between the early treatment group and the late treatment group (SMD = 2.94, 95% CI = -3.78–9.74, P = 0.4). The maxillary protraction effect of the two treatment groups was similar.
For each measurement index of FM and FM/RME, we chose 3 relatively high-quality studies (score ≥4) to carry out the sensitivity analysis (shown in Figs
In the FM group, we chose 3 relatively high-quality studies (score ≥4) to carry out the sensitivity analysis. 1: Angle SNA. The FM treatment group presented a greater increase in SNA than the controls (MD = 1.75, 95% CI = 1.53–1.97, P < 0.00001). 2: Angle SNB. The FM treatment group presented a greater decrease in SNB than the untreated controls (MD = -1.54, 95% CI = -2.29–-0.80, P < 0.0001). 3: Angle ANB. The FM treatment group presented a greater increase in ANB than the controls (MD = 3.51, 95% CI = 2.99–4.03, P < 0.00001). 4: Angle SN/GoGn. The FM treatment group presented a greater increase in SN/GoGn than the controls (MD = 1.44, 95% CI = 0.21–2.68, P = 0.02). The sensitivity analysis results were consistent with previous results.
The same method was used in the FM/RME group; we also chose 3 relatively high-quality studies (score ≥ 4) to carry out the sensitivity analysis. The comparison was performed using five indices. 1: Angle SNA: The FM/RME treatment group presented a greater increase in SNA than the controls (MD = 1.38, 95% CI = 0.75–2.01, P < 0.0001). 2: Angle SNB. The FM/RME treatment group presented a greater decrease in SNB than the untreated controls (MD = -2.70, 95% CI = -3.34–-2.05, P < 0.00001). 3: Angle ANB. The FM/RME treatment group presented a greater increase in ANB than the controls (MD = 3.12, 95% CI = 1.43–4.81, P = 0.0003). 4: ANS-Me length. There were no significant differences in ANS-Me length between the FM/RME treatment group and untreated controls (MD = 1.43, 95% CI = -2.72–5.58, P = 0.50). The sensitivity analysis result was consistent with previous results.
A funnel plot was used to assess the publication bias of the literature. Symmetrical graphical funnel plots were obtained in all included studies (
These symmetrical plots indicate the absence of publication bias in the present meta-analysis.
Treating Class III malocclusion is currently considered one of the most challenging and complex parts of orthodontic practice. According to some surveys, the prevalence of Class III malocclusion is as high as 14% in Asian populations and approximately 1% to 5% in white populations[
If we want to gather adequate evidence to guide clinical practice and establish a standard of treatment, one approach is to treat patients with similar symptoms with the same treatment protocols. However, we cannot deny that definitive conclusions from any one trial are limited and should be treated casually, especially when they are based on studies with limited sample sizes[
Overall, our meta-analysis aimed to explain three things: firstly, whether the maxilla FM protractor is an effective treatment for skeletal Class III malocclusion; secondly, whether FM protractor and maxilla expansion causes forward displacement of the maxilla and the inhibition of mandibular growth; and thirdly, whether an early treatment group benefits more from maxilla protractor devices than does a parallel late treatment group. The results are as follows: for the first issue, a summary of the meta-analysis suggests that a maxillary protraction appliance is effective for correcting anterior crossbite with a retruded maxilla. The changes in SNA and ANB in the FM group with regard to anterior movement of the maxilla indicate similarity with findings reported in the previous literature [
The results of the last group suggest that early treatment was not more effective than was late FM therapy. An early treatment time involves the early mixed dentition period with the chronologic ages ranging from 7 to 10 years old, whereas a late treatment time is usually applied during late mixed dentition and early permanent teeth dentition from 11 to 14 years old. Evaluation of lateral cephalograms revealed no significant difference between the early treatment group and the late treatment group. In another words, the assumption that orthopedic forces on the maxilla and the mandible were more effective and advantageous during early treatment compared with later treatment was not supported (P > 0.05). The statistical findings offer the best evidence to address the third question. However, with the exception of the p value of SNB (P < 0.05), there is not enough evidence to represent the various angular changes in SNB between the two groups. Considering that the FM therapy patients are all either prepubertal or in puberty, it is expected that the natural growth of the mandible will result in the forward movement of point B and a natural change in SNB. In addition, younger children are expected to have more growth potential than do older children and a more significant change in SNB. As a result, our meta-analysis has provided the best evidence that early treatment is not more effective than late FM therapy. Although there were no differences in treatment time or clinical treatment effects between the early and late groups, the two groups were not entirely the same. In the early treatment patients, the maxillary sutures were still not fully fused at the chronologic ages of 7 to 10 years old. Therefore, the maxillary expansion and protraction effects still included some true skeletal decompensation with new bone deposition at the maxillary sutures. In other cases, the maxillary sutures completely closed after ten years of age. After closure of the maxillary sutures, the expansion of the maxilla lay in the eruption of the maxillary molar teeth and new bone deposition in the buccal side. Therefore, in the late treatment group, dental compensation played the leading role. This finding reminded us that true skeletal decompensation and maxilla expansion are considered possible only in younger children before maxillary suture closure. Although we found that that early treatment was not more effective than late FM therapy, we must admit that orthodontic interference only improves the clinical symptom with dental compensation rather than offering true skeletal decompensation after suture closure of the maxilla. True skeletal decompensation, which is considered possible in younger children, allows for good clinical stability, whereas dental compensation in older children is prone to relapse following appliance removal.
Nonetheless, this study has certain shortcomings similar to other articles due to the nature of meta-analysis. First, the numbers of relevant research articles and patients were not sufficiently large. In addition, some of the relevant studies were excluded from our analysis because of incomplete or overlapping data; consequently, our analysis may not provide a sufficient number of patients/cases. Moreover, not all sources of heterogeneity in the included studies could be addressed. Furthermore, certain methodological limitations exist because a meta-analysis is a retrospective study. Ultimately, we only provide evidence for the effectiveness of FM and FM/RME treatment using untreated controls. What we are also interested in is the difference between the FM- and FM/RME-treated groups. However, the newest relevant and eligible studies of FM groups and FM/RME controls are still not sufficient for conducting a meta-analysis. Given these results, additional research in this field is necessary, and our meta-analyses will continually improve.
In conclusion, the data of our meta-analysis supported that FM and FM/RME treatment are both effective clinical early treatment methods for skeletal Class III malocclusion. The statistical analysis thoroughly proved that late FM therapy could achieve relatively similar outcomes to early treatment therapy. Although real skeletal development is much better than dental compensation, we cannot abandon maxillary expansion and protraction treatment for early permanent dentition, which we treated as the late group. Current evidence of correlative research still needs to be greatly expanded due to the limited number of published articles in this field. Therefore, we still wish to conduct a large, detailed, randomized, well-designed, comprehensive, controlled trial with a long follow-up visit to confirm our recent research.
(DOC)
(DOC)