Effects of rapid maxillary expansion on temporomandibular joints:A systematic review

Rapid maxillary expansion (RME) has been used as a routine clinical procedure, with the main objective of mechanically separating the palatine suture in young patients with maxillary transverse constriction, deep palatal vault, and or posterior crossbite. It has also been used in cases of crowding by increasing the perimeter of the arch and other situations where a transverse increase is necessary.^1,2 

After opening the median palatal suture, not only transverse but also vertical and anteroposterior changes occur. Clockwise rotation of the mandibular plane, resulting from a lower and posterior position of the mandible after RME, has been one of the most reported effects.^3,4  These changes may directly or indirectly affect other structures of the craniofacial complex to which the maxilla is interconnected, including the temporomandibular joint (TMJ).⁵ 

TMJ is referred to as the most complex articulation of the human organism, with its visualization difficult due to the anatomy and presence of neighboring structures. Thus, computed tomography (CT) is the examination of choice for the detailed study of changes in TMJ skeletal structures and magnetic resonance imaging (MRI) is the preferred diagnostic method for observation of the TMJ joint disc.⁶  Studies of the effects of RME on the TMJ are controversial. It is essential to synthesize the evidence and clarify the real impact of this intervention on the TMJ.

A previous systematic review summarized TMJ alterations after correction of posterior crossbites in growing patients.⁷  However, the authors included studies in which patients received other orthodontic procedures associated with RME. In one of them,⁸  there was intervention in the maxilla and mandible, including the McNamara expander and U-Bow type I activator. This may have increased the risk of bias since mandibular protrusion may have induced condylar remodeling in growing patients.⁹  This same study used ultrasound as the method of analysis of the TMJ, which does not give sharpness and precision in the assessment of articular structures.¹⁰ 

The current systematic review evaluated the impact of RME on condylar position, disc joint, joint space, and interarticular relationship in growing patients by means of CT or MRI. The studies included were those in which subjects underwent RME only.

This review was registered in the PROSPERO database under protocol CRD42017055790 (https://www.crd.york.ac.uk/PROSPERO/), in accordance with the PRISMA check-list of systematic reviews and meta-analyses.¹¹ 

The scientific literature was searched in May 2018 and alerts were received from the databases until October 2019. The search was conducted to identify articles reporting the effects of RME on the TMJ by means of MRI, CT, or cone beam tomography. A detailed search was conducted in the following electronic databases: PubMed, Scopus, Web of Science, Cochrane, Lilacs, Science Direct, OpenGrey, Google Scholar, and ClinicalTrials. An additional hand search was conducted to identify additional relevant publications. No restrictions were placed on the publication date or language. Specific search strategies were developed for each database (Appendix A). All relevant quotations were saved in the Endnote bibliographic reference manager (version X7, Thomson Reuters). The triage of articles by title and abstract; and full text, data extraction, and quality assessment were done independently by the first two authors. Disagreements were resolved through a consensus meeting and, where appropriate, consulted with a third author.

The eligibility criteria were defined based on the PICO research strategy for clinical practice based on scientific evidence.

1. Participants: growing patients with mixed or young permanent dentition and without symptomatology of temporomandibular dysfunction;

2. Intervention: RME;

3. Comparison: Evaluation before RME or growing patients receiving no treatment;

4. Outcome: Measurements in TMJ by means of CT or MRI; and

5. Types of studies: Randomized, non-randomized, prospective, or retrospective clinical trials.

1. Studies examining patients with cleft lip and or palate or any craniofacial anomalies;

2. Studies including cases with extraction of permanent teeth, previous orthodontic treatment, and any other type of interventions associated with RME; and

3. Case series, opinion articles, in vitro or animal studies, and literature reviews.

After reading the titles and abstracts, the articles that were not in accordance with the proposed theme were eliminated, the remaining articles were read for eligibility, and a final decision was made.

A data extraction table was developed with the following items: author's name, year of publication, type of malocclusion, and the need for expansion, sample size, age, expander device used, activation protocol, time measurement, and result of interest.

The articles finally selected and included in this review were analyzed for risk of bias. For randomized controlled trials with a control group, the Cochrane Collaboration tool was used for systematic reviews of intervention.¹²  For studies without a control group, the National Heart, Lung, and Blood Institute (NHLBI) tool was used for Pre-Post studies.¹³ 

The level of evidence was calculated using the Grading of Recommendations, Assessment, Development and Evaluation Pro software (GRADEpro Guideline Development Tool, available online at www.gradepro.org).¹⁴ 

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram of the study selection process is presented in Figure 1. The established search strategy found the following number of records per database: PubMed (n = 1883), Scopus (n = 755), Web of Science (n = 689), Cochrane (n = 206), Lilacs (n = 408), Science direct (n = 168), OpenGrey (n = 0), and Google scholar (n = 194), totaling 4303 records. After deletion of duplicates, there were 2750 records for title and abstract reading of which 2735 were eliminated because they were not related to the topic or were case reports, editorials, letters, or literature reviews. Thus, the remaining 15 records were selected for full-text reading. Of these, one study was eliminated because it was a protocol proposal to be followed in future studies,¹⁵  one for evaluating pre- and post-intervention changes through gypsum models and bite registration,¹⁶  one for assessing condylar change by means of zonograms,¹⁷  and five studies were excluded since they examined patients who underwent slow maxillary expansion.^10,18–21  Therefore, seven articles from the systematic search were included for qualitative analysis;^22–28  as well as one record found by hand search (n = 1);²⁹  totaling eight articles included in this review (Table 1). A quantitative analysis was not feasible given the heterogeneity in the methodology of the selected articles.

The patients' ages ranged from 6.8 years²⁴  to 17 years.²⁸  The sample size ranged from 10 patients²⁵  to 39 patients.²⁷  Regarding the method used to measure changes, three articles used MRI, of which two evaluated the articular disc,^23,24  and the other one evaluated remodeling of the head and the condylar branch.²²  Five articles evaluated the condylar position and articular spaces by means of tomography: two by conventional tomography^25,26  and three by cone beam tomography.^27–29 

The standard protocol for RME employed by several authors^4,5  consisted of two quarter-turns of activation per day until overcorrection was reached. That protocol was used for six of the studies included.^{22–25,28,29}  Leonardi et al. (2012)²⁶  used the protocol of three quarter-turns per day while Melgaço et al. (2014)²⁷  achieve one full turn per day.

All of the studies carried out measurements prior to expansion. However, variability was observed in relation to the time of measurement fafter the activations. One study evaluated after 6 and 18 weeks,²²  two after 12 weeks,^25,27  one after 12 and 18 weeks,²⁴  one after 18 weeks,²³  two after 24 weeks,^28,29  and one immediately after the activations.²⁶ 

Arat et al. (2008)²²  observed that RME induced remodeling in the head and condylar branch. One article reported that RME modified joint space and condylar postion, promoting greater spatial symmetry between the crossbite and non-crossbite²⁵  side. Three articles found that the RME was able to change the individual condylar position^27,28,30  and joint space,^26,29  maintaining the symmetric relationship between the condyles.^26,27,29  There was no evaluation of intercondylar symmetry in the measures used by McLeod et al. (2016),²⁸  who observed no alteration of the condylar position in the glenoid fossa in individuals.

The two studies that analyzed the articular disc reported that RME did not alter the articular disc either in shape²⁴  or in position,^23,24  whether they were normally placed or had displacement prior to expansion.

In the study published by McLeod et al. (2016),²⁸  there was insufficient information about the process of random sequence generation, concealment of allocation, and blinding of outcome assessors. An imbalance was also observed in the number of patients included and the absence of a sample calculation. Four attempts over two months were made to contact the authors to obtain the necessary information; however, there was no response. Quality score was “risk of uncertain bias.”

No random sequence generation, allocation concealment, and an imbalance in the number of subjects between groups was observed in Ghoussoub et al. (2018).²⁹  Quality score was “high risk of bias.”

Blinding of participants or professionals would not affect the outcome; (Figure 2).

Arat et al. (2008) conducted a large study subdivided into three articles. In the present review, parts 2²²  and 3,²³  which analyzed the condylar remodeling and articular disc, respectively, were assessed for risk of bias in a single way and using the term “Arat et al. (2008).^22,23 ”

The NHLBI assessment revealed, in Table 2, that three articles had a “good” score indicating low bias risk^24,26,27  and three articles had a “fair” score indicating moderate risk of bias.^22,23,25  None of those scored a “bad” score that would indicate a high risk of bias. Only one study²⁷  presented an adequate methodology for all items on the checklist of the qualifier.

The primary outcome of each study as well as the quality assessment results are shown in Table 3.

The GRADE (Grading of Recommendations, Assessment, Development and Evaluation) evidence profile table is described in Table 4. The evidence for the outcomes evaluated ranged from low to very low quality, suggesting that the estimate reported can differ significantly from the measure evaluated.

No previous systematic review evaluating changes in the TMJ after RME using only 3D images was found in the literature. In the present review, articles that evaluated changes after RME in the condyle, disc, and/or fossa by means of CT or MRI were included. All articles included in this review had a well-defined study objective, a sample consistent with the need for maxillary expansion, intervention with a well-described treatment protocol, and well-defined prespecified outcome measures.

For a significant difference between two groups of one standard deviation, 33 individuals would generate 80% power to detect this difference at a significance level of an alpha level of 0.05. Three studies fulfilled this requirement,^26–28  with a sample of 34, 39, and 37 individuals, respectively. Thus, to fill item 5 of the NHLBI qualifier, “NR” (not reported) was attributed to articles that did not report the sample size calculation, since it was not possible to determine if the sample would be representative of the population. However, if there were articles where the sample size was less than 10 patients in the interest group, the item would be classified as “No” because this small sample size would probably not be adequate to provide reliable results. This protocol was adopted by Lisboa et al. (2018)³⁰  in their review on the effects of mandibular advancement or retreat surgery on soft facial tissue.

Four articles^22,23,28,29  did not report the type of malocclusion of the patients. Three studies used only patients with Class I malocclusion,^25–27  while one article²⁴  used a sample with 11 patients with Class I malocclusion, 17 with Class II malocclusion (11 subdivision and six total), and two patients with Class III malocclusion. In all articles included in this review there was no report on modification or correction of malocclusion after RME. In a systematic review about sagittal changes after RME in Class II patients,³¹  it was concluded that the studies found had methodological deficiencies with controversial or non-relevant sagittal effects. Coskuner and Ciger (2015)¹⁸  verified that a Class II relationship can be corrected to a certain extent by slow expansion of the maxilla with a quadhelix; however, there was no significant effect on the TMJ.

The proper use of statistics minimized errors in reporting the results and interpreting findings. Based on the “Tutorial for statistical test tutorial,”³²  all of the studies used adequate tests for their analysis, with the exception of Arat et al. (2008)^22,23  who did not perform the Wilcoxon test suggested by the tutorial. In that study, a method of visual inspection was used instead, ignoring the inferential statistical analysis.

Of the eight articles included in this review, two^24,29  did not report whether there were no clinical signs and symptoms of temporomandibular dysfunction. Based on this review, it was observed that RME did not promote intercondylar impairment in patients with funtional unilateral posterior crossbite (FUPC) or maxillary constriction with bimaxillary posterior crossbite. Intercondylar asymmetry was improved to a symmetric position in patients with FUPC. Kasimoglu et al. (2007)³³  investigated the relationship between vertical asymmetries of the mandibular condyle with different occlusion types and concluded that growing patients with FUPC might be at risk for developing skeletal mandibular asymmetries in the future and the early correction of posterior crossbite can help practitioners to prevent skeletal asymmetries.

During the analysis of the results of the two studies that evaluated the articular disc,^23,24  it was concluded that RME was not able to change the articular disc even in patients who had pretreatment disc displacement. This was in agreement with Wadhawan et al. (2008)³⁴  who used the Twin-Block and Bionator orthopedic appliances in growing patients. Kinzinger et al. (2006),³⁵  used functional mandibular advancement devices and showed that, although the joints initially showed partial or total anterior displacement of the disc, a significant improvement could be achieved.

There were no studies regarding the effects of RME on the TMJ after the retention period. All studies evaluated the immediate effects, with the post-treatment evaluation being performed shortly after the active period or in the retention period (12 to 24 weeks after activation). Lagravère et al. (2006),³⁶  in their systematic review on the long-term skeletal effects of RME, showed that changes in the mandibular position were temporary. Garib et al. (2007)³⁷  evaluated the effects of RME and long-term orthodontic treatment. Three years after the end of treatment, the immediate effects of RME were not significant in the long term.

• RME in growing patients is able, in the short term, to modify the condyle-fossa relationship, not modifying the position or shape of the articular disc, but able to maintain or improve the intercondylar symmetry relationship.

Although the NHLBI score is low to moderate risk of bias, the clinical relevance of these review findings is limited by the Cochrane and GRADE scores.

Due to the methodological quality of the available evidence, RME cannot be indicated for the sole purpose of altering the TMJ individually or inter-articularly in growing patients requiring maxillary expansion.

There is a need for improved reliability of evidence on the subject, preferably through randomized controlled trials with long-term follow-up that cover post-retention stability and investigate a possible mitigation of the risk of developing future mandibular skeletal asymmetries in growing patients requiring maxillary expansion.