Protocol and Registration

The protocol for this systematic review was registered on PROSPERO (CRD42015017287).

Eligibility Criteria

Information Sources and Search

The following electronic databases were systematically searched up to April 2015: Medline, Embase, and all evidence-based medicine reviews via OVID, Cochrane Library, Scopus, PubMed, and Web of Science. The used keywords included orthodontic anchorage, transpalatal arches (TPAs), or bar or bars. This search strategy was first designed for Medline (Appendix 1) and then adapted for the other databases. A partial gray literature search was performed using the Google Scholar search engine by looking over the first 100 listed hits. No restrictions were applied regarding the language or publication date.

Study Selection

During the selection phase, two reviewers (SD-B and MFNF) independently evaluated the titles and abstracts of the retrieved studies from the database searches using the inclusion criteria. In the second phase, the same reviewers performed assessment of the full-text articles. The reviewers resolved any discrepancies by discussion until consensus.

Data Collection Process

The data were first extracted according to standardized tables. Data was compared for accuracy, and any discrepancy was resolved through the reexamination of the original study until a consensus was reached.

Data Items

The variables extracted from each selected article included sample size, retraction method, type and material of TPA, anchorage device used in control groups, reference lines to which anterior teeth segment retraction and/or molar crown mesialization were measured, superimposition landmarks, percentage of mesial crown molar movement at the end of the anterior retraction phase, and the authors' conclusion.

Outcome

The primary outcome was the molar crown mesial movement during anterior teeth retraction.

Risk of Bias in Individual Studies and Quality of Evidence

Methodological quality appraisal was evaluated according to the Cochrane Collaboration's Risk of Bias tool¹⁶ for randomized clinical trials (RCTs). In case of non-RCTs, the Methodological Index for Non-randomized Trials (MINORS)¹⁷ was used. An additional summary of the certainty of the conclusions and strength of the evidence was developed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach (Table 1). The quality of evidence was assessed as high, moderate, low, or very low for the outcome mesial crown molar movement.

Table 1 GRADE Summary of Findings for Meta-Analyses on Molar Mesial Movement During Incisor Retraction^a

^a CI indicates confidence interval; RR, risk ratio.

^b Inconsistent study design.

^c Many of these studies did not consider blinding of the participants and outcome assessors.

View Fullscreen

Table 1 Extended

View Fullscreen

Data Synthesis

Data were pooled to provide an estimate of the effectiveness of the TPA using a random-effects model, given that there were more than three trials eligible for a quantitative analysis and considering the expected statistical heterogeneity.¹⁶ Random-effects models are preferred when significant differences are expected between patients and evaluation methods. The primary outcome was mesial crown molar movement (molar anchorage loss). For continuous outcomes, the mean difference with standard deviation and 95% confidence intervals were calculated. Clinical heterogeneity was examined by assessing the characteristics of the selected trials, including similarity between interventions, patients, phase of treatment in which intervention was applied, and outcome measures. Publication bias was examined for the trials to be included in a meta-analysis, using a funnel plot by visually assessing the degree of funnel plot asymmetry.¹⁸ Statistical heterogeneity across the studies was tested using the T² and the I² statistic, with guide for interpretation as follows: 0% to 30%, not important; 30% to 50%, moderate heterogeneity; 50% to 100%, considerable heterogeneity.^19,20 The pooled effect estimate was considered significant if P was <.05. A meta-analysis software (The Cochrane Collaboration's software Review Manager, RevMan) was used to perform data analyses.

Study Selection

A flowchart illustrating the selection of studies for this systematic review is presented in Figure 1. Twenty-five full texts were obtained for the second phase evaluation, of which 11 articles^21–31 were later excluded. The reasons for exclusion are listed in Appendix 2. Finally, 15 articles^{2,6,10–13,32–40} met the eligibility criteria. A summary of the key methodological data and study characteristics is presented in Tables 2 and 3.

View Figure

• Download as Powerpoint
• Download Figure

Table 2 Summary of Study Characteristics and Results of the Included Studies (TPA Only)^a

^a AP, anteroposterior; A-OLp position f the maxillary base; DUM-V, the vertical distance between the distal upper molar point and rotated SN plane; F, female; FH, Franfort horizontal; Is-OLp, position of maxillary incisor; M, male; Ms-OLp, position of maxillary molar; NS, nonsignificant; OLp, occlusal line perpendicular; PTV, pterygoid vertical plane; SS, stainless steel; TPA, transpalatal arch; U1, maxillary central incisor; U6, maxillary first molar; UIT-V, the vertical distance between the upper incisal tip and the rotated SN plane (SN′); Y, line perpendicular to Frankfort horizontal pass through sella.

^b Mesial molar movement is a calculated as percentage of the molar movement provided in the article.

View Fullscreen

Table 2 Extended

View Fullscreen

Table 3 Summary of Characteristics and Results of the Included Studies (Conventional Anchorage)^a

^a Mesial molar movement is a calculated percentage of the molar movement provided in the article. APO indicates line between A point on the maxilla and pogonion on the mandible; F, female; HG, headgear; Is-OLp, maxillary central incisor position; M, male; Ms-Olp, maxillary first permanent molar position; MXCI, maxillary central incisor; MXM1, maxillary first molar; NS, nonsignificant; PTV, pterygoid vertical line; TPA, transpalatal arch; U1, maxillary permanent incisors; U6, maxillary permanent molars.

View Fullscreen

Table 3 Extended

View Fullscreen

Risk of Bias Assessment

Methodological appraisal of the selected studies is presented in Tables 4 and 5. Nine of the included studies^{4,10–13,33–35,39} were RCTs, and all of them were considered to present high risk of bias.

Table 4 Methodological Appraisal of the Selected Studies According to the Cochrane Risk of Bias Tool

View Fullscreen

Table 5 Methodological Appraisal of the Selected Studies According to the Methodological Index for Nonrandomized Studies (MINORS) Assessment Tool^a

^a The items are scored 0 (not reported), 1 (reported but inadequate), or 2 (reported and adequate), with the global ideal score being 16 for noncomparative studies and 24 for comparative studies.¹⁴

View Fullscreen

Six of the included studies^{2,11,32,36,38,40} were non-RCTs. Study scores ranged from 13 to 20 points out of 24. Significant limitations were identified for most of the studies, such as the retrospective enrollment of the sample^2,11,32,40 with nonconsecutive inclusion of patients^2,32,36 or unclear reports about inclusion criteria.^11,38,40

Study Characteristics

Effects of Interventions

Anchorage loss was significantly greater in the groups using TPA alone as an anchorage device, for a total of 158 individuals with a mean loss of anchorage ranging from 1.76 to 4.21 mm (Figure 2), which represents 27% to 54% of the mesial molar crown movement toward the extraction space. Groups using combined conventional anchorage devices also showed higher loss of molar crown anchorage as compared with skeletal anchorage. The conventional anchorage groups presented a mean mesial molar crown movement ranging from 1.26 to 4.28 mm (Figures 3 and 4) or approximately 20% to 40% of the extraction space. The TAD groups presented a mean loss ranging from 0.00 to 2.05 mm (0%–22%). Only one study² investigated the use of TPA compared with no anchorage, and the values revealed no significant differences between the two groups that reported a similar loss of anchorage of about 45%.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Eight clinical trials analyzing 308 patients and comparing TPA vs TADs were combined in a meta-analysis (Figures 2–4). The meta-analysis showed a statistically significant reduction in anchorage loss in the TADs group compared with TPA alone (mean difference [MD] 2.09 [95% confidence interval {CI} 1.80 to 2.38], I² = 51%, seven trials), TPA + headgear (MD 1.71 [95% CI 0.81 to 2.6], I² = 94%, four trials), and TPA + utility arch (MD 0.63 [95% CI 0.12 to 1.15], I² = 0%, three trials).

Certainty Levels and Strength of the Evidence

Based on the GRADE recommendations, the body of evidence reporting the mesial molar crown movement ranged from very low to high because of the limitations in the design and the high risk of bias in some of the included studies. Strong evidence is present among the studies that compared TAD against the use of TPA alone. Weak evidence supports the use of TPA even when it is paired with headgear or utility arch to retract either canines or anterior teeth when maximum anchorage is needed.

Summary of Evidence

In this review, RCTs and non-RCTs were selected to address the effectiveness of TPA in controlling the maxillary molars anchorage during retraction of anterior teeth in extraction cases. The studies included two categories: TPA sole use as an anchorage mean and TPA used as an adjunct with other conventional anchorage means.

It was suggested that the adjunctive use of TADs should be significantly favored over the sole use of TPA as an anchorage device during retraction when properly indicated. GRADE assessment tool application shows that there is high-quality evidence to support that claim.

The combined use of TPA and headgear did not enhance anchorage when compared with TADs. Even while retracting canines using only TPA and utility arch, adjunctive use of TADs resulted in better anchorage control. The studies considered in that matter varied from very low to low quality, mainly because of the lack of RCTs.

TPA was used in a selected number of clinical trials to test its anchorage ability. In one of the studies, Zablocki et al.² reported no difference in the molar mesial movement between the control group where no anchorage was planned and TPA-only anchorage group; thus, the TPA did not have any added value with regard to molar anchorage. A consistent finding from all RCTs^{6,10,13,33–34} implementing en masse retractions is that the TPA did not prevent molar mesial movement.

On the other hand, when TPA was used as the sole anchorage mean to retract canines in a two-step retraction technique, the two related studies^12,35 still failed to favor the use of TPA in preventing mesial movement of the molars. In a recent study, El-Bialy et al.⁴¹ concluded that TPA alone does not minimize anchorage loss when used with continuous arch mechanics, and they recommended not using the TPA.

The studies reporting the use of TPA as an adjunct anchorage mean with headgear or utility arch during anterior retraction again showed a consistent outcome in which molar anchorage loss was greater in the conventional anchorage group and the incisors were better controlled and more retracted with skeletal anchorage.

However, when only canine retraction was assessed in the combined TPA with other conventional anchorage means, it is suggested^11,38,39 that the anchorage achieved was equivalent to that of the skeletal anchorage, as there was no statistically significant difference in the mesial molar movement between both groups, although by the end of the retraction phase of the anterior teeth, skeletal anchorage showed better incisor anteroposterior control. Based on these findings, TPA would be recommended for canine retraction only and only if it were combined with other conventional means. Bearing in mind that canine retraction precedes anterior incisor retraction in most of the extraction cases, the orthodontist then would question the use of TPA as it becomes inconvenient and burdens the clinician with an additional unnecessary procedure. Finally, it has to be mentioned that the use of TPA for other purposes or in less demanding anchorage cases is not questioned in this systematic review.

Limitations

Among the included studies, failure to blind the patients and the clinician was a common, albeit inevitable, flaw. One has to admit that in such clinical trials, blinding the patient or the clinician to the appliances used is hardly achievable. Future studies should ensure that sequence generation and allocation concealment requirements are properly met to further reduce risk of bias. Among the included studies were also non-RCTs. Even though this type of study cannot avoid selection bias, use of consecutively treated patients could at least partially account for this bias.