Impact of third molars on mandibular relapse in post-orthodontic patients: A meta-analysis.

BACKGROUND/
PURPOSE: Whether third molars contribute to or aggravate relapse, particularly in the mandibular dental arch, after orthodontic treatment remains controversial. Orthodontic clinicians vary widely in their practice regarding prophylactic third molar removal after orthodontic treatment. The present study systematically reviewed and meta-analyzed the available literature, and assessed the impact of third molar removal on the relapse of mandibular dental arch alignment after orthodontic treatment.
MATERIALS AND METHODS: Relevant literature was searched on online databases, namely Pubmed, Embase, and Cochrane. Outcomes of post-orthodontic mandibular relapse were evaluated in terms of the Little's irregularity index, intermolar width, and arch length. Statistical analysis was conducted using the Review Manager software (Version 5.3, The Cochrane Collaboration, Oxford, England).
RESULTS: Our initial search strategy yielded 360 citations, of which three retrospective studies were selected. The Little's irregularity index (weighted mean difference = 0.80, 95% confidence interval = 0.13-1.47, P = 0.02) differed significantly between the erupted third molar extraction group and agenesis third molar group; whereas the arch length and intermolar width did not. No outcome differed significantly between the impacted third molar extraction group and agenesis third molar group.
CONCLUSION: Removal of the mandibular third molars is recommended for alleviating or preventing long-term incisor irregularity.

Introduction

Whether third molars contribute to or aggravate the relapse of dental arch alignment, particularly in the mandibular dental arch, after orthodontic treatment remains controversial. Several studies have reported inconsistent findings. And orthodontic clinicians vary greatly in their practice regarding prophylactic removal of third molars after orthodontic treatment.1, 2, 3, 4, 5

Laskin surveyed more than 1300 American orthodontists and oral surgeons. And found that approximately 65% of them recommended removal of unerupted or impacted third molars in patients during or after orthodontic treatment. These orthodontists believed that unerupted or impacted third molars occasionally produce an anterior force that cause separation in the contact points and subsequent crowding of the mandibular incisors. Niedzielska reported that if sufficient space was unavailable for the third molars to erupt, they exert forces on the adjacent teeth, causing crowding. According to Tüfekçi, most US and Swedish orthodontists believe that the erupting lower third molars exert an anterior force and they ‘‘rarely’’ or ‘‘never’’ cause crowding of the dentition.

Some studies have reported contradictory opinions.4, 5 Bishara reviewed the literature extensively, and Sidlauskas and Trakiniene studied a group of 91 participants (average age, 21.01 ± 4.13 years); these studies have concluded that insufficient evidence is available to determine whether third molars are the only or even, the major etiologic factors affecting posttreatment changes in incisor alignment.

The present study systematically reviewed and meta-analyzed the available literature and assessed the impact of third molar removal on the relapse of mandibular dental arch after orthodontic treatment.

Materials and methods

Selection criteria

The inclusion criteria of the present study were as follows: (1) Studies evaluating the impact of third molar removal on the post-treatment changes in the mandibular dental arch alignment in post-orthodontic patients, (2) studies which clearly documented the inclusion and exclusion criteria of patient selection, (3) studies which clearly documented the operative techniques, including extracted teeth and treatment protocols, and (4) studies which clearly documented the post-treatment evaluation methods and criteria. Studies were excluded from the analysis if the outcomes of interest were unclear.

Search strategy

We systematically searched the literature for studies assessing lower incisor relapses in post-orthodontic patients, regardless of the third molar removal. Studies were identified by searching databases, namely Pubmed, Embase, and Cochrane, online. Key terms included in the search were a combination of: third molar, impaction, wisdom tooth, wisdom teeth, orthodontic, crowding, relapse, stability, and retention. The ‘‘related articles’’ function was used to broaden the search, and all abstracts, studies, and citations retrieved were reviewed. In addition, we attempted to identify other studies by hand-searching the reference sections of these papers and by contacting known experts in the field. No language restrictions were applied. The last search was conducted in July 2015.

Study selection and data extraction

Two reviewers independently extracted the following information from each study: first author, year of publication, study population characteristics, study design, inclusion and exclusion criteria, matching criteria, number of participants operated on using each technique, operative parameters, complications, and postoperative recovery. The individually recorded decisions of the two reviewers were compared and disagreements if any were resolved by a third reviewer.

Outcome measures

The following outcomes were used to evaluate the mandibular relapse in post-orthodontic patients during follow-up period: (1) Little's irregularity index, (2) intermolar width, and (3) arch length.

Statistical analysis

Statistical analysis was conducted using the Review Manager (RevMan) software (Version 5.3, The Cochrane Collaboration, Oxford, England). The meta-analysis of randomized controlled trials (RCTs) was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. Whenever necessary, standard deviations were estimated according to the reported confidence interval (CI) limits, standard error, and range values. The mean difference was calculated for continuous outcomes, and the weighted mean difference (WMD) was analyzed. The precision of an effect size was reported as a 95% CI. A pooled estimate of the mean difference was calculated using the DerSimonian and Laird random-effects model. The data were pooled only for studies that exhibited adequate clinical and methodological similarities. Statistical heterogeneity was assessed using the I2 test, with I2 quantifying the proportion of total outcome variability attributable to the variability among these studies.

Results

Characteristics of the selected trials

Fig. 1 depicts the flowchart of the selection procedure of the trial selection. Our initial search strategy yielded 360 citations. Of these, 335 were excluded because they either were not clinical trials or focused on a different condition. Thus, 25 full-text articles were assessed, of which 22 did not meet the eligibility criteria: one was not a clinical trial, thirteen studies were review articles,10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22 and seven lacked sufficient data.13, 23, 24, 25, 26, 27, 28 One RCT was excluded because the orthodontic procedure of the mandibular dental arch used in that study was different from the other studies that were included in the present meta-analysis. In total, three eligible trials were identified,30, 31, 32 all of which were retrospective studies. These studies were included in the meta-analysis and the measurable outcomes were pooled.

Figure 1

Flowchart of the trial selection.

Table 1 presents the characteristics and patient demographic data of each trial. These studies were published between 1974 and 1990 and had sample sizes between 35 and 97. All patients had finished their orthodontic treatment and most were treated with the edgewise technique. Premolar extraction or not during orthodontic treatment was not limited. The mean postretention time ranged from 9.3 to 13 years, with the mean age ranging from 26.6 years to 28 years 6 months. Both sets of ages were unavailable in Andreasen et al.

Table 1

Three clinical studies describing the posttreatment changes in the mandibular dental arch in post-orthodontic patients.

Study [citation]	Study design	Inclusion criteria	Number of patients in each group	Age	Follow-up	Outcomes
Kaplan, 197430	Retrospective	1. Orthodontically treated Caucasian patients	Erupted: 30Impacted: 20Agenesis: 25	26.6 years (postretention)	9.3 years	Little's irregularity indexArch lengthIntermolar width
Ades et al., 199031	Retrospective	1. All participants were Caucasian2. Participants free of all retention for at least 10 years	Erupted: 32Impacted: 14Absent: 17Extraction: 34	28 years 6 months (postretention)	13 years	Little's irregularity indexArch length
Andreasen et al., 198732	Retrospective	1. All the cases received the edgewise orthodontic treatment.	M + aP−b: 9M + P + c: 8M−dP−: 9M−P+: 9	Not mentioned	Not mentioned	Little's irregularity indexArch lengthIntermolar width

Molar present.

Premolar extracted.

Premolar present.

Molar extracted.

It existed some differences in these three controlled studies, including the postretention time and postretention age. The range of the mean postretention time is 9.3 to >10 years. And the mean postretention age was from 26.6 to 28.5 years. The differences cause some mild risk of bias, but it will not influence our results.

The three controlled studies were included in this review “support” the cause-and-effect relationship between the eruption or impaction of the third molars and the development of anterior tooth crowding, but the relationship is weak. Because of the limitation in these trials, the definite relationship need more research to survey and make sure.

In Kaplan, the cases were classified according to the status of the third molars as both erupted to occlusal plane, impacted, and agenesis. By contrast, in Ades et al., the cases were classified according to the status of the third molars as erupted, impacted, agenesis, and extracted. In Andreasen et al., two groups were formed depending on the presence of third molars. Outcome measures were compared between the groups included in these studies.

Outcomes of erupted or agenesis molars and impacted or agenesis molars

For practical clinical directions, the extracted data were combined as follows: patients with erupted or agenesis third molars and patients with impacted or agenesis third molars.

Data on the posttreatment changes in the mandibular dental arch of post-orthodontic patients in the three retrospective studies were used in the meta-analysis.

Erupted or agenesis third molars

The data were classified into the erupted third molar group and the agenesis third molar group and subjected to a meta-analysis. The two groups exhibited a significant difference in the changes in the Little's irregularity index (WMD = 0.80, 95% CI = 0.13–1.47, P = 0.02, Fig. 2). However, the two groups did not differ significantly in terms of the changes in arch length and intermolar width (Figure 3, Figure 4).

Figure 2

Forest plot of the erupted third molar group compared with the agenesis third molar group. Outcome: Changes in Little's irregularity index.

Figure 3

Forest plot of the erupted third molar group compared with the agenesis third molar group. Outcome: Changes in arch length.

Figure 4

Forest plot of the erupted third molar group compared with the agenesis third molar group. Outcome: Changes in intermolar width.

Impacted or agenesis third molars

The impacted third molar group and the agenesis third molar group exhibited no significant difference for all outcomes (changes in the Little's irregularity index, arch length, and intermolar width; Figure 5, Figure 6, Figure 7, respectively).

Figure 5

Forest plot of the impacted third molar group compared with the agenesis third molar group. Outcome: Changes in Little's irregularity index.

Figure 6

Forest plot of the impacted third molar group compared with the agenesis third molar group. Outcome: Changes in arch length.

Figure 7

Forest plot of the impacted third molar group compared with the agenesis third molar group. Outcome: Changes in intermolar width.

Discussion

Zawawi et al. and Stanaitytė et al. surveyed and systematically reviewed the literature.

Stanaitytė et al. surveyed the literature by using the Medline database. In total, 223 relevant articles published between 1971 and 2011 were identified. However, only 21 articles corresponded to the selected criteria and were analyzed. The results were contradictory: Some researchers opined that lower third molars cause teeth crowding, whereas the others confirm controversy.

Zawawi et al. reviewed 12 clinical studies to evaluate the role of third molars in the development of crowding or relapse after orthodontic treatment in the mandibular dental arch and discovered that the third molars did not correlate with severe anterior tooth crowding in most studies. However, four studies reported a different outcome. A definitive conclusion on the role of third molars in the development of anterior tooth crowding could not be drawn. A high risk of bias was reported in most trials, and the outcomes were inconsistent. However, most studies did not support a cause–effect relationship. Therefore, in their opinion, third molar extraction to prevent anterior tooth crowding or post-orthodontic mandibular relapse was unjustified.

Because of these contradictory opinions, we surveyed the literature and conducted a meta-analysis to identify some clinical directions for orthodontic clinicians regarding third molars.

The meta-analysis reviewed three studies. In Kaplan, the cases were classified according to the status of the third molars as both erupted to occlusal plane, impacted, and agenesis. By contrast, in Ades et al., the cases were classified according to the status of the third molars as erupted, impacted, agenesis, and extracted. In Andreasen et al., two groups were formed depending on the presence of third molars. In our clinical practice, we come across patients with or without third molars. If the third molar was present, the need for extracting the third molars to prevent postretention crowding and the effect of impacted and erupted third molars were unknown. To offer a practically clinical reference, we combined the data to compare the populations with impacted or erupted third molar to those with the agenesis third molar. The impacted third molar group included patients in the impacted groups in Kaplan and Ades et al. The agenesis third molar group included patients in the agenesis group in Kaplan, patients in the agenesis and extracted group in Ades et al., and patients in the third molars missing groups in Andreasen et al. (not considering premolars). The erupted third molar group included patients from the erupted groups in Kaplan and Ades et al. and the third molars present group in Andreasen et al.

In Kaplan and Ades et al., no significant differences in the changes of all outcomes were observed between groups during the posttreatment period. Andreasen et al. study did not report data on comparison between the third molar subgroups; however, the researchers concluded that the absence of third molars did not increase the stability of the orthodontic results as evaluated by the changes in the incisor alignment, incisor angulation, or dental arch widths. As we pooled out the data from these studies and conducted a meta-analysis, different results were indicated as the sample numbers increased.

Overall, when the extracted data were classified into the erupted third molar group and the agenesis third molar group and subjected to a meta-analysis, a significant difference in the changes in the Little's irregularity index between the two groups were observed. However, the two groups did not significantly differ in terms of changes in the arch length and intermolar width. A comparison of the impacted third molar group and agenesis third molar group exhibited no significant differences for all outcomes (changes in the Little's irregularity index, arch length, and intermolar width). Further research is necessary to identify possible reasons and mechanisms.

Of the clinical studies we identified, the study by Harradine et al. was the only RCT. The researchers randomized 44 of 77 patients to have their third molars removed after completion of retention after their orthodontic treatment. This study was excluded from our meta-analysis because although all patients in this study had previously undergone the orthodontic treatment, all treatments offered in the mandibular dental arch simply comprised no treatment or premolar extractions only. More RCTs on this topic are necessary to offer stronger evidence.

There are several reasons might account for posttreatment relapse, such as growth related changes, muscular factors, periodontal ligament traction, bone adaptation, masticatory force, and third molars…, etc. On the basis of our results, third molars are not the major etiologic factors affecting posttreatment changes in incisor alignment. But after the meta-analysis we conducted, we still offer some clinical directions. While treating patients with erupted third molars, we suggest mandibular third molar removal for alleviating or preventing long-term incisor irregularity. However, all clinical studies selected were retrospective trials with risk of bias. In addition, the orthodontic treatment is complex. When treating a patient with or without third molar extraction before or during the orthodontic treatment, we recommend that additional clinical information be collected to obtain data on their strengths and weaknesses, such as patient age, potential development of pathogenesis, technical considerations of the surgical procedure, and long-term periodontal implications. Consultation and cooperation with other specialists are helpful.

RCTs are time-intensive studies but offer highly reliable evidence. Additional animal studies may provide possible explanations. Further investigations on the factors involved in postretention crowding are warranted.

In conclusion, our initial search strategy yielded 360 citations, of which three retrospective studies were selected. The Little's irregularity index (weighted mean difference = 0.80, 95% confidence interval = 0.13–1.47, P = 0.02) differed significantly between the erupted third molar extraction group and agenesis third molar group; whereas the arch length and intermolar width did not. No outcome differed significantly between the impacted third molar extraction group and agenesis third molar group.

We hypothesized that, compared to impacted third molars, erupted third molars might produce more anterior force and cause crowding of the mandibular incisors. On the basis of our results, while treating patients with erupted third molars, we suggest mandibular third molar removal for alleviating or preventing long-term incisor irregularity. However, additional clinical information should be collected, and consultation and cooperation with other specialists are still needed. Further investigations on the factors involved in postretention crowding are warranted.

Conflicts of interest

The authors have no conflicts of interest relevant to this article.