Changes in condylar position during orthodontic treatment depending on the amount of incisor retraction: a cone-beam computed tomography study

Orthodontic treatment may be associated with temporomandibular disorders through changes in the condylar position. This study aimed to evaluate changes in the condylar position among different amounts of maxillary incisor retraction during orthodontic treatment using cone-beam computed tomography images. Fifty-four participants were enrolled and divided into minimal (n = 14), moderate (n = 20), and maximal (n = 20) retraction groups based on the amount of incisor retraction (< 1, 1–6, and > 6 mm, respectively). Changes in condylar position before (T0) and after (T1) orthodontic treatment were assessed for the superior, anterior, posterior, and medial joint spaces (SJS, AJS, PJS, and MJS, respectively). Changes in joint spaces were compared between T0 and T1 in each group using paired t-tests and among the three groups using analysis of variance. Anterior movement of the condyle was observed in the maximal retraction group with a 0.2 mm decrease in ΔAJS and a 0.2 mm increase in ΔPJS, significantly greater than those in the minimal retraction group. The AJS and PJS showed statistically significant differences between T0 and T1 (P < 0.05) in the maximal retraction group. The condyle may show a statistically significant but clinically insignificant forward movement in the maximal incisor retraction group, whereas it was relatively stable in the minimal and moderate incisor retraction groups. More attention should be paid to the signs and symptoms of the condyle in patients with excessive incisor retraction during orthodontic treatment.


Introduction
Temporomandibular disorder (TMD) is characterized by several clinical problems involving the masticatory muscles, temporomandibular joint (TMJ), and associated structures [1]. TMJ is a small but complex structure surrounded by the mandibular condyle, articular eminence, and mandibular fossa in the temporal bone [2]. Divergent opinions exist regarding the relationship between condylar position and TMD. Condylar position plays an important role in the occurrence of TMD; a posteriorly positioned condyle is more frequently found in patients with TMD [3,4]. However, cone-beam computed tomographic (CBCT) assessment has contrastingly indicated that the condylar position had no relationship with TMD [5].
Orthodontic treatment may cause changes in the condylar position, implying that orthodontic treatment may be related to TMD. A small but statistically significant posterior movement of the condyle was noted after non-extraction orthodontic treatment [6]. Retraction of the anterior teeth following maxillary premolar extraction also leads to a 4-7% posterior movement of the condyle [7]. The decreased joint space resulting from condylar posterior movement may cause disc displacement and induce internal derangement, pain, or dysfunction [8,9]. Nevertheless, other reports claimed that the four-premolars extraction treatment was not related to the condylar posterior movement and did not increase the risk of TMD [6,10,11]. Furthermore, some studies have reported that orthodontic treatment neither causes TMD nor increases the risk of developing signs and symptoms of TMD [10][11][12][13].
Notably, few previous studies included patients under 19 years of age whose condyles might be undergoing growth [6,12]. Moreover, the condylar position was evaluated using two-dimensional (2D) images, in which the accuracy can be decreased by the superimposition of the bilateral structures of the TMJ [6,10]. Furthermore, different angles for each radiograph may not demonstrate the exact anatomical structure and cannot represent real changes in the condylar position. A study using CBCT images showed that the condyle significantly moved posteriorly after orthodontic treatment with extraction of the two maxillary premolars [14] although it only included participants treated with extraction of the maxillary teeth in Class II malocclusion without a control group. Other studies have indicated that in patients with Class II malocclusion, the condyle tends to be positioned anteriorly before and after orthodontic treatment. However, these studies did not specify the extent of tooth movement involved [15][16][17]. Another study reported no significant differences in joint space among patients with skeletal Class I, II, and III malocclusions [18].
Previous studies have reported inconsistent results regarding the number of tooth extractions and amount of anterior tooth retraction [6,7,10,11,14]. With the advancements in CBCT, changes in the mandibular condyle after orthodontic treatment can be observed using three-dimensional (3D) images more accurately. However, only a limited number of studies have been conducted specifically in patients with Class II malocclusion. Therefore, this study was conducted to verify the null hypothesis that the condylar position does not change according to the amount of retraction of the maxillary anterior teeth during orthodontic treatment in adults.
CBCT was used to evaluate changes in the condylar position during orthodontic treatment and compare these changes depending on the amount of maxillary incisor retraction.

Participants
Participants were retrospectively enrolled among 1,035 patients who underwent comprehensive orthodontic treatment at a private clinic between March 2012 and June 2017. The inclusion criteria were as follows: (1) to exclude the effect caused by growth, the age of participants should be over 19 years; (2) the participants should have undergone orthodontic treatment with either non-extraction or four-premolar extraction; and (3) the participants should have available facial CBCT images taken before (T0) and after (T1) treatment. Participants were excluded if they had a history of orthodontic treatment, facial asymmetry (menton [Me] deviation > 4 mm) [19], any systemic disease, anterior crossbite before treatment, or presence of TMD, including asymptomatic condylar resorption. A total of 128 participants were initially enrolled in this study. Additionally, 74 participants were excluded based on the following exclusion criteria: movement of the maxillary central incisor (U1) > 1 mm in the non-extraction treatment group or movement of the central incisor < 1 mm in the premolar extraction treatment group (Fig. 1). Based on previous studies [20], the sample size was calculated as a minimum of 27 patients using the G-power program (G* Power 3.1.9.4, Dusseldorf, Germany) with effect size f = 0.43, α err prob = 0.05, and power = 0.95.
The participants were divided into three groups to compare the condylar movement depending on the amount of U1 retraction during orthodontic treatment as follows: 1) minimal retraction group or the control group treated with non-extraction treatment and demonstrated U1 movement < 1 mm (n = 14); 2) moderate retraction group, treated with four-premolars extraction treatment and showed 1 mm ≤ U1 retraction ≤ 6 mm (n = 20); and 3) maximal retraction group, treated with four-premolars extraction treatment and showed U1 retraction > 6 mm (n = 20) [21]. The movement of U1 was measured on lateral cephalogram images generated using CBCT. A single orthodontist treated all patients with the following treatment sequence: 0.018inch slot pre-adjusted edgewise appliance (Roth prescription) with a wire sequence of 0.014-inch nickel-titanium (NiTi), 0.016-inch NiTi, 0.018 × 0.018-inch copper-NiTi, 0.016 × 0.022-inch copper-NiTi, and 0.016 × 0.022-inch stainless steel; the orthodontist uses the same archwire sequence for all patients. En masse retraction was used to close the extraction space, and miniscrews were arranged in most extraction cases, depending on the anchorage requirement. In some cases, Class II elastics were used to complete orthodontic treatment during the finishing stage.
The study protocol was approved by the Institutional Review Board (IRB Approval No.:2-2017-0060) of Yonsei University Dental Hospital. Owing to the retrospective nature of the study, the review board waived the requirement for written informed consent from the participants.

Measurements
CBCT images (Pax zenith, Vatech Co., Gyeonggi-do, Korea) were imported as DICOM files with a 0.2 mm voxel size and reconstructed into 3D images using OnDemand3D software (CyberMed Inc., Seoul, Korea). The CBCT images were reoriented with three reference planes: 1) Frankfort horizontal plane (FHP) passing through the left orbital (Or) and bilateral porion (Po); 2) midsagittal plane (MSP) perpendicular to the FHP and passing through the sella (S) and nasion (N); and 3) coronal plane perpendicular to both the FHP and MSP and passing through N. T0 and T1 images were superimposed with the anterior cranial base through automatic voxel-based superimposition to make the images on the same coordinate system [22], and N was set to 0, 0, and 0.
Four linear parameters, including the superior, anterior, posterior, and medial joint spaces (SJS, AJS, PJS, and MJS, respectively), were measured by identifying the fossa and condylar points for each joint space (Table 1, Fig. 2a, b). The SJS, AJS, and PJS were measured in the sagittal plane, while the MJS was measured in the coronal plane. Joint spaces were measured on both the right and left sides, and each measurement was included. To exclude participants with changes in condylar position induced by condylar resorption during orthodontic treatment, the condylar height, ramus inclination, and ramus height were measured (   Table 1). The horizontal movement was measured along the MSP in reference to the vertical reference plane, whereas the vertical movement was measured along the FHP in reference to the horizontal reference plane.

Statistical analysis
Statistical analyses were performed using Statistical Package for the Social Sciences software (version 20.0; IBM Corp., Armonk, NY, USA). All CBCT scans were digitized by a single examiner. Five samples in each group were randomly selected and re-digitized 2 weeks later to evaluate intraobserver reliability using the intraclass correlation coefficient (ICC). All ICCs were > 0.91, except for the MJS (0.87), which indicated good to excellent intra-observer reliability. The Shapiro-Wilk test was used to determine the distribution of measurements within each group. Statistical significance was set at P value < 0.05. Paired t-tests were used to evaluate the differences between T0 and T1 in each group. To compare the differences among the three groups, a one-way analysis of variance or the Kruskal-Wallis test was conducted depending on the normal distribution of the data, and post hoc tests were performed using the Scheffe and Mann-Whitney U tests.  The most lateral point of medial wall of mandibular fossa Medial joint space condyle point (MJSc) The most medial point of the condyle Measurement Superior joint space (SJS) † The distance from SJSf to SJSc Anterior joint space (AJS) † The distance from AJSf to AJSc Posterior joint space (PJS) † The distance from PJSf to PJSc Medial joint space (MJS)* The distance from MJSf to MJSc Ramus inclination Angle between FHP and ramus line, which is the tangent line of the posterior border of ramus Condyle height The distance from SJSc to the tangent line of the sigmoid notch which is perpendicular to the ramus line Ramus height The distance from Ar to Go Tooth movement ΔU1_H The amount of horizontal movement of U1 tip along the midsagittal plane in reference to the VRP ΔL1_H The amount of horizontal movement of L1 tip along the midsagittal plane in reference to the VRP ΔU6_H The amount of horizontal movement of U6 mesiobuccal cusp tip along the midsagittal plane in reference to the VRP ΔL6_H The amount of horizontal movement of L6 mesiobuccal cusp tip along the midsagittal plane in reference to the VRP ΔU1_V The amount of vertical movement of U1 tip along the midsagittal plane in reference to the VRP ΔL1_V The amount of vertical movement of L1 tip along the midsagittal plane with reference to HRP

Results
Based on the aforementioned criteria, we enrolled 14 participants in the minimal retraction group (ΔU1_H, -0.1 ± 0.7 mm), 20 in the moderate retraction group (ΔU1_H, -4.1 ± 1.4 mm), and 20 in the maximal retraction group (ΔU1_H, -8.0 ± 1.5 mm) ( Figs. 1 and 4). Although the age did not differ significantly among the three groups (P > 0.05), the treatment time of the minimal retraction group was significantly shorter than that of the other groups (P < 0.05; Table 2). The initial positions of the incisors (U1 and L1) and molars (U6 and L6) in the minimal retraction group indicated a significantly retrusive position compared with the other two groups (P < 0.05). The maximal retraction group had the highest OJ and ANB values. The SN-MP of the minimal retraction group was significantly smaller than those of the other two groups; however, the minimal retraction group had the largest PJS and longest PFH (P < 0.05). During orthodontic treatment, horizontal tooth movement, including ΔU1_H, ΔL1_H, ΔU6_H, ΔL6_H, ΔU1-SN, and ΔIMPA, showed significant differences among the three groups depending on the treatment objectives (P < 0.05; Table 3), whereas vertical tooth movement did not show any significant differences (P > 0.05).
The anterior condylar movement was observed in the maximal retraction group. The decrease in ΔAJS and increase in ΔPJS in the maximal retraction group (-0.2 ± 0.3 mm and 0.2 ± 0.3 mm, respectively) was significantly greater than changes in the minimal retraction group (P < 0.05, Table 3). Moreover, AJS and PJS showed statistically significant differences between T0 and T1 in the maximal retraction  Table 1 for definitions of the abbreviations and measurements

Table 3
Changes in cephalometric measurements and joint spaces before and after treatment in each group ANOVA with Scheffe post hoc test was used to compare three groups, while § variables were compared using the Kruskal Wallis test and Mann Whitney test for post hoc analysis because the data did not show normal distribution. Different superscript letters indicate statistically significant differences The abbreviations and definitions for each parameter are listed in Tables 1 and 2 Table 4). SJS and MJS showed no significant differences within each group or among the three groups during orthodontic treatment (P > 0.05). The decrease in OJ was significantly greater in the maximal retraction group than that in the other two groups (P < 0.05), although there were no statistically significant differences in OB, ANB, SN-MP, AFH, or PFH among the three groups (P > 0.05; Table 3). The condylar and ramus heights showed no significant differences between T0 and T1 in any groups (P > 0.05, Table 4). Ramus inclination showed a 1º decrease in the minimal retraction group during treatment (P < 0.05).

Discussion
This study aimed to evaluate changes in condylar position among patients with different amounts of incisor retraction during orthodontic treatment using CBCT images. The present study used CBCT images acquired before and after treatment. To minimize radiation exposure, lateral cephalograms and panoramic radiographs were not obtained but were regenerated from the 3D CBCT images [23]. Statistically significant but clinically insignificant anterior condylar movement was observed in the maximal retraction group (Fig. 4), although the condylar position was stable in the minimal and moderate retraction groups.
Developments in temporary anchorage devices such as miniscrews have led to an expansion in the range of tooth movements, including retraction [21,24]. In the present study, miniscrews were used as anchorages in the maximal and most of the moderate retraction groups to minimize the anchor loss during space closure. Because of the miniscrews, the extraction space was used solely for incisor retraction while minimizing the mesial movement of the posterior teeth in the maximal retraction group. However, the palatal inclination of the maxillary incisors was greater in the maximal retraction group than that in the moderate retraction group, whereas there was no significant difference in the inclination of the mandibular incisors. This might be one of the reasons for the forward movement of the condyle [25].
The results of the present study indicated that the condyle was positioned 0.2 mm anteriorly when there was a large amount of anterior teeth retraction. Class II elastics used during the finishing stage would cause palatal inclination of the maxillary incisors and subsequent clockwise rotation of the mandible [26] which might be attributed to the forward movement of the condyle. Forward functional shift of the mandible to establish occlusion after treatment with maxillary premolar extraction was possible because the condylar movement was related to the displacement of the maxillary molar rather than the amount of incisor retraction [27]. However, the extent of the maxillary molar movement was more evident in the moderate retraction group than that in the maximal retraction group. Moreover, posterior movement of the condyle has also been reported after Class II camouflage treatment with extraction of the two maxillary premolars [14,28]. The masticatory muscles might cause mandibular retraction to achieve maximum occlusion. The contradictory findings regarding the condylar position following the retraction of the anterior teeth may be understood in light of the observed movement. This movement, measured at an average of 0.2 mm, represents a small change and thus may be considered clinically insignificant. The condylar position seems stable throughout orthodontic treatment, as the changes were reported to be less than 1 mm, as in the present study [14,29]. Although there were individual differences, the condyle showed slight forward or being in the centric positions in cases of non-extraction or Class II camouflage treatment with maxillary premolars extraction treatment [30]. Therefore, it has been claimed that the extraction treatment is not related to changes in the condylar position [6,29]. The present study also showed a stable condylar position in the minimal and moderate retraction groups.
Changes in condylar position, particularly in the posterior direction, may lead to disc displacement, often observed in patients with TMD [3,14,31,32]. The incidence rate of TMD in patients with anteriorly positioned condyles is less than that in those with posteriorly positioned condyles [3,9,33]. However, as skeletal and neuromuscular adaptive changes can be expected during orthodontic treatment, TMD would occur only when the amount of change exceeds the adaptive range [6].
According to the records of the patients enrolled in the present study, none reported any signs or symptoms of TMD during or after orthodontic treatment. This may be because all joint spaces were maintained within the optimal range of 1.5-4 mm [34], although the anterior joint space in the maximal retraction group showed statistically significant changes.
Visualization of complex anatomical structures may be unclear or limited to 2D images owing to the overlapping of adjacent or bilateral structures, which can be avoided in 3D images [14]. Moreover, 3D images allow accurate morphological assessment of the bony structures of the TMJ [34,35].
The strengths of this study lie in its methodology. Grouping based on the amount of U1 retraction and recording the amount of molar movement during orthodontic treatment, combined with CBCT imaging, allows for an accurate comparison of the impact of tooth movement on the condylar position during orthodontic treatment. Additionally, the evaluation of changes in condylar height, ramus inclination, and ramus height can reveal genuine alterations in condylar position, effectively excluding potential external influencing factors. This study had a few limitations. First, there was a lack of an untreated group for comparison, which could not be included for ethical reasons. Including the untreated group during the same period might have allowed for more accurate comparisons. Additionally, fewer participants were included in the minimal retraction group than that in the other groups. Furthermore, changes in disc position and classification of the molar relationship should be incorporated for further observation of the condylar position to assess changes during retention and evaluate the occurrence of TMD.